The direct answers, plain-language explanations and full technical assessments are collected here in the original order. Equations, evidence states, sources and unresolved inputs are visible for each question.
How to read this document
The reviewed R4 evidence and conditional results are retained. This page adds short editorial summaries; it does not add a new experiment, calculation or mission measurement. The original brief contained an unnumbered opening question and questions 1, 2, 4–22: these are the 22 scopes below. There is no invented question 3.
Answers are not the same as closed requirements. The largest remaining empirical gaps are the faint source population, true polar scene backgrounds, selected receiver performance at end of life, independent source-depth labels and a validated thermal-alternative red/blue bound. Each answer states its narrower gaps.
Formula text uses explicit conditions and local symbols. Detailed derivations, units, executable notebooks, raw retained arrays, residuals and tolerance policies are in the physicist verification companion. All answers and source records can be inspected without JavaScript.
Initial question · original scope initial
Transmission to nadir at 550 km versus overhead cloud optical depth; cloud-top/above-top source; Rayleigh versus particles.
Direct answer
There is no single received fraction determined by altitude and overlying cloud optical depth alone. The direct ray has an inverse-square geometric factor multiplied by exponential extinction; scattered light needs an angular, time-dependent transport calculation. In the dry reference atmosphere, Rayleigh scattering removes about 5–9% of the direct 337.1 nm beam above 15–18 km. A cloud beneath a cloud-top source can redirect extra light toward the satellite and add a delayed tail.
Physics 101
A photon is one packet of light. Imagine a flash sending packets in all directions. The instrument catches only the tiny fraction whose paths cross its entrance. Distance makes the same flash spread over a larger area. A clear line of sight is therefore not the same as catching every photon.
A cloud mostly changes directions. Some packets continue almost straight, some wander before escaping, and some leave through the sides or bottom. Optical depth measures how likely a packet is to encounter something along a path. An optical depth of 1 leaves about 37% of the original straight-going beam untouched; it does not mean that 63% of all light energy has disappeared.
The high atmosphere is thin, but ultraviolet light is readily scattered by gas molecules. Our reference calculation leaves about 92% of the direct 337 nm beam at 15 km and 95% at 18 km. This is large enough to include in a careful budget. Within a substantial cloud, ice-particle encounters are usually the larger effect.
At cloud top there may be no cloud above the flash. Light sent downward can still bounce through the cloud and return upward later. That gives a prompt component plus a delayed component. The answer therefore needs a cloud shape and a direction of observation, not just a single percentage labeled cloud transmission.
Full technical answer
The fraction is not determined by source altitude and overhead cloud optical depth alone. Specify whether the denominator is all photons emitted, upward escaping photons, photons per unit solid angle toward nadir, or photons incident on an aperture. For an isotropic point source and a small on-axis aperture, the direct vacuum fraction is A/(4 pi R^2). Multiply this direct term by exp[-tau_cloud - tau_gas - tau_aerosol] along the source-observer ray. The total, including redirected light, requires a time-dependent angular transport calculation. Scattering removes photons from a ray but does not necessarily destroy them.
At 337.1 nm the reviewed dry US Standard Atmosphere calculation gives vertical Rayleigh optical depth about 0.0888 at 15 km, 0.0759 at 16 km and 0.0555 at 18 km. Direct molecular transmissions are about 0.915, 0.927 and 0.946. The older CloudScat exponential 7.2 km atmosphere gives 0.921, 0.931 and 0.947 respectively. Thus Rayleigh is a several-percent direct-beam effect even above the source; the pressure/profile approximation is a smaller but recorded systematic. These factors exclude ozone, aerosols and diffuse photons scattered back into acceptance.
Cloud particles usually dominate scattering inside an optically substantial anvil. At the cloud top the overhead particle optical depth can vanish while molecules remain. An above-top source also illuminates the cloud beneath it: reflected photons can add a delayed component in the upward viewing direction. A directional increase relative to a no-cloud view is consistent with energy conservation. It must not be described as cloud transmission greater than the fraction of emitted photons.
Use beta_ext, single-scattering albedo and the angular phase function or matrix, together with physical cloud dimensions, source position and emission pattern. Identical optical depth with different physical distance or phase shape need not give the same delay. Ice crystals near anvil top are often nonspherical; Mie theory is exact for a homogeneous sphere, not for every ice habit. Ozone absorption and stratospheric aerosol cannot be assumed absent solely because the source is high.
For perspective, an overhead particle optical depth 0.01 is smaller than the reference molecular 0.055-0.089 column above 15-18 km; particle depths 1-10 are much larger. This compares direct-ray extinction only. Different angular scattering and cloud reflection still determine total received light.
Corrections to earlier assumptions
Direct-beam extinction is not total detected fluence.
Cloud-top does not mean vacuum; above-top does not mean no cloud contribution.
The 25 mm aperture used in examples is an illustrative design variable, not an established requirement.
Equations and their conditions
initial · expression 1
K_direct = A cos(theta) exp(-tau_LOS) / (4 pi R^2)
Fraction of isotropically emitted photons reaching a small planar aperture without collision.
Conditions: Point source; small aperture; stated line-of-sight extinction; no focusing or PDE included.
initial · expression 2
tau = integral beta_ext(s) ds; E_ph = h c / lambda
Optical depth is dimensionless. Photon energy at 337.1 nm is about 5.89e-19 J.
Conditions: Gas, aerosol and particle terms must use the same spectral and geometric conventions.
Notation is local to each expression. Here K_direct includes aperture area and is a dimensionless photon fraction. The verification companion separately defines transport kernels per unit area. Preserve the reference plane and units when comparing them.
What supports this answer
conditional calculation
Dry reference Rayleigh column and direct transmission; neither real polar weather nor total diffuse transmission.
Validated ice phase/absorption properties and actual aperture/field/time acceptance.
Related executed studies (4) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N01 · Clear-air direct Rayleigh transmission
executed conditional calculation not mission measurement
Recomputed the 337.1 nm molecular cross section and dry standard-atmosphere number column, with an independent pressure-column comparison and grid check. This is a direct no-collision factor, not total cloudy directional transmission.
Additional direct-ray table uses exact inverse-square geometry for 25 mm at 550 km, source 15 or 18 km, then multiplies the reference Rayleigh factor andexp(-tau_cloud). The 20% column is a separate illustrative detection probability. It is not used to infer an intrinsic source yield.
Calculation limitations
Direct/ballistic no-collision transmission only. Rayleigh-scattered photons are not all absorbed or necessarily outside the full instrument response.
Cloud attenuation can multiply this factor for the direct ray e^-tau_cloud only; total diffuse directional fluence requires radiative transfer and cannot be obtained by this exponential alone.
This is a clear dry reference atmosphere; no aerosol, ozone, real polar seasonal weather, source spectrum, or finite-band weighting included.
At/above cloudtop, tau_cloud_above=0 does not remove the delayed contribution of downward photons returning from the cloud below.
R4-N02 · An above-top reflector can increase nadir light and delay it
executed conditional calculation not mission measurement
Re-ran 108 analytic cases for an isotropic point above a flat instantaneous Lambert reflector. Independent radial integration and escape/threshold checks passed. In the surface limit, nadir fluence relative to vacuum is 1+2 rho while total upward escape is(1+rho)/2; directional gain therefore does not create photons.
The reflected waveform is a geometric convolution with a linear rise to an indefinite plateau. Its 10-90 rise is not an impulse arrival percentile or onset-to-peak duration.
No mappingofrhotooverheadcloudtau. h=0 meanslimitingjust-abovesurfaceposition, notanembeddedcloud-topscatteringboundary
R4-N03 · Depth and diffusion strength are not uniquely identifiable
executed conditional calculation not mission measurement
Re-ran 48 ideal half-space diffusion/absorption examples. The first-exit density is L/sqrt(4 pi D) t^(-3/2) exp[-L^2/(4 Dt)-gamma t]. Cases with identical L^2/D andgamma have the same entire kernel.
The computed leading 10-90 coefficient is 0.0743277 L^2/D, close to the published rounded 0.0742. The nonabsorbing mode is L^2/(6 D); their ratio is 0.44597, so the definitions are quantitatively different.
This diffusion density is a plane first-exit quantity integrated over transverse position, not finite-aperture directional radiance. Diffusion has noncausal early support and cannot replace transport near the ballistic front.
Model Dvaluesandtop 18 kmincounterexampleareconditional, notindependentheightobservations
Published 5.5 usmodel-basedsubsamplefitdoesnotprove general 5.5 usmeasurementprecisionat 10 usbinning
Exact L²/D whole-kernel degeneracy is a property of the parabolic diffusion PDE; finite-speed light transport can break it.
All parabolic kernels have nonzero pre-flight tails. Passing t 10>=L/c does not validate early-time radiative transport.
0.07432773536 versus the paper's 0.0742 is not an established published error; original Eq 6 bitmap was not retrieved.
R4-N11 · Independent cylinder transport and the unresolved rare-path test
executed conditional transport with incomplete GHM convergence
Executed 56 phase/geometry cells: 48 nonvacuum Monte Carlo cases plus 8 analytic vacuum cases. Production used 14.4 million histories in 288 independent batches. A homogeneous finite cylinder spans 14-17 km with 30 km radius; source heights 15, 16, 17, 18 km; observer 550 km and 60 degree circular field. Total vertical particle optical depth takes 0, 0.3, 1, 3, 10, 30, 45.
The scalar GHM bulk phase uses 10 micrometre effective radius and linearly interpolates 330/340 nm optical-model data at the 337.1 numerical coordinate. Its derived g is 0.7709447; the comparison HG has the same g. SSA is assumed 1. This is one ice optical model, not measured event microphysics; no polarization, gas absorption, Rayleigh, aerosol, ground or real 3D heterogeneity is present.
Packet flights use exact cylinder intersections, physical path lengths and a nadir next-event estimator with an analytic direct term. Source-proposal importance weights preserve the stated isotropic emission. Production had zero truncated histories, zero negative-delay violations and zero reported 12 ms kernel overflow; maximum weighted energy residual was 2.91e-15 per packet. These checks do not certify directional variance convergence.
All 24 nonzero GHM rows remain flagged for nonuniform point-nadir convergence; 10 have estimated fluence standard error above 5%. At source 15 km, total optical depth 10, the production GHM gain 1.666+/-0.118 changes to 2.004+/-0.099 with a larger-packet control. A source-cone control gives 1.805+/-0.132. These are estimator diagnostics, not confidence bounds on a physical cloud. Rare contributions can escape a small-batch SEM estimate. No physical GHM/HG rise-time bias is inferred.
Two additional exactly matched ideal HG controls were compared to retained CloudScat results. Their g=0.87098, source 16 km, 12-18 km cylinder and 20 km radius differ from the production grid. Directional gains agree within 0.41 and 0.69 combined batch standard error at overlying optical depth 3 and 10. Timing postprocessing is shared, so this is not an independent timing-estimator validation. GHM and real-ice propagation remain unvalidated.
Retained nonpassing evidence includes a deliberately inadequate 10-scatter cap, packet/proposal/radius sensitivity, and a vacuum-metadata repair. Early code/serialization attempts are logged. Later diagnostic code changes are distinguished from the executed production snapshot; failed controls were not deleted.
Calculation limitations
Conditional model comparison, not a universal transmission curve or inversion.
Equal g does not fix higher phase-function moments. This grid supplies conditional estimates with MC errors; the current GHM rare-path controls do not establish a physical timing bias.
Two matched homogeneous-HG controls from the new engine were compared with retained CloudScat outputs and agree within their combined Monte Carlo errors. This independent-code check does not validate tabulated GHM, atmospheric transport or a real-cloud classifier; own analytic tests are labeled separately.
The native processed phase grid is assumed. Interpolation, ice habit and size-distribution uncertainties are not represented by MC error.
No actual event-brightness population, absolute rate, current calibration, flight noise or rise-time classifier is provided.
The tabulated GHM point-nadir rare-path estimator is not uniformly converged. Representative increased-packet and source-proposal controls can differ more than naive finite-batch SEM suggests. Do not infer a physical GHM/HG rise-time bias from unconverged ratios. All original batches and controls are retained.
GHM point-nadir rare-path convergence and nonlinear timing uncertainty
UV ice habit, roughness, size-distribution and interpolation representativeness
Real ice absorption and atmospheric Rayleigh/ozone/aerosol paths
3D cloud structure and source angular/temporal/spectral distribution
Observed noisy onset, eventpopulation classifier and flight detection rates
Batch SEM is an estimate, not a guaranteed 95% interval. Source and radius controls may share random streams; do not assign independent-difference significance to those comparisons.
Aperture transfer has unitsm^-2 per emitted photon. Multiply by actual aperture area for photon fraction, and only then by appropriate PDE for detected events. Directional gain above 1 is not a photon probability above 1.
Sources for this question (6)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Which radiative-transfer calculation/tool should be used, and alternatives to CloudScat?
Direct answer
Use CloudScat.jl as the geometry-specific starting point, with independent analytic and Monte Carlo checks. MCARaTS is the strongest identified alternative for an independent transient comparison, but its required build and matched 337 nm run were not completed. The executed independent cylinder model and two matched ideal-HG controls test limited implementation cases; they do not validate real ice or a source-depth inversion.
Physics 101
A radiative-transfer program follows the possible journeys of light through air and cloud. In a Monte Carlo calculation, the computer samples many random journeys and estimates how often light reaches the instrument and how late it arrives.
CloudScat is a sensible starting point because its designers had this lightning-to-satellite problem in mind. A second program is valuable as a cross-check, but only if the two programs describe the same experiment. Comparing all escaping light from one program with only nadir light from another would give a disagreement even if both were correct.
More simulated photons make the computer estimate less noisy. They cannot tell us whether a chosen cloud really has the assumed crystal shapes or whether the flash emitted the assumed pulse. This is the difference between numerical precision and physical accuracy.
Some alternatives are excellent at atmospheric spectra or images but do not directly produce the arrival-time history needed here. The useful choice is the tool whose documented source, geometry, phase function and timing can be checked. The report labels a model as executed only when a retained receipt and results support that statement.
The new revision also checked two simple matched cases against CloudScat. They agree within their estimated simulation uncertainty. That is useful progress, but the harder tabulated-ice cases still have unresolved sampling uncertainty.
Full technical answer
CloudScat.jl remains a practical primary tool because it was developed specifically for internal lightning emission viewed from space, and it has actually been installed and exercised in the earlier project. Its outputs are conditional on the supplied optical properties, geometry and source waveform. A large packet count reduces Monte Carlo sampling error; it does not validate those physical assumptions.
The most promising independent alternative identified in the campaign is MCAR aTS: the retrieved 0.10.3 source contains direct-source local estimates and path-length histograms. Its source and interfaces were inspected, but no compiler/dependency build or matched 337 nm execution was completed in that campaign. An isotropic source cannot be represented by an invalid 360 degree full-cone parameter if the code only permits 180 degrees; two opposite half-power hemispheres are one possible valid construction.
Other relevant tools solve different parts of the problem. SMART-G and nonspherical-ice lookup tables can provide realistic steady radiative transfer and particle properties. Public libRadtran includes a restricted subset of MYSTIC functionality; access to the required 3D version must be verified. MSCART/McRALI and published transport models provide useful comparisons, while general transient renderers such as mitransient require careful unit, source, phase and optical-path audits. ARTDECO is useful for atmospheric optical properties and steady calculations, not automatically a drop-in transient lightning solver.
The fresh revision calculation, if its execution receipt is present in the numerical appendix, is an independently written idealized Num Py transport diagnostic. It is not an independent real-ice validation or a replacement for a qualified production code. Before comparing models, match source normalization, boundaries, detector geometry, direct light, time origin and phase convention. Benchmark vacuum, Beer-Lambert direct light, isotropic single scattering, energy balance and convergence before interpreting cloud-depth curves.
The fresh independent cylinder implementation now has two executed matched ideal HG cross-code controls against CloudScat, with fluence differences 0.41 and 0.69 combined estimated SE. This is a limited new implementation result. It does not validate tabulated GHM, realice, timinganalysis or source-depth inversion.
Corrections to earlier assumptions
Available documentation, downloaded code, an installed executable and a validated calculation are different states.
Do not advertise all named atmospheric tools as equivalent transient 3D solvers.
Equations and their conditions
A1 · expression 1
y(t) = integral s(t-u) K(u; geometry, optics) du
Separate source emission history s from the propagation impulse response K.
Conditions: Linear transport and fixed medium; detector response is a further convolution.
What supports this answer
published model and prior execution
CloudScat geometry-specific methods and earlier project execution.
A matched independent transient benchmark using the same real-ice properties and detector estimator.
A validated event-specific source/cloud model.
Related executed studies (1) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N11 · Independent cylinder transport and the unresolved rare-path test
executed conditional transport with incomplete GHM convergence
Executed 56 phase/geometry cells: 48 nonvacuum Monte Carlo cases plus 8 analytic vacuum cases. Production used 14.4 million histories in 288 independent batches. A homogeneous finite cylinder spans 14-17 km with 30 km radius; source heights 15, 16, 17, 18 km; observer 550 km and 60 degree circular field. Total vertical particle optical depth takes 0, 0.3, 1, 3, 10, 30, 45.
The scalar GHM bulk phase uses 10 micrometre effective radius and linearly interpolates 330/340 nm optical-model data at the 337.1 numerical coordinate. Its derived g is 0.7709447; the comparison HG has the same g. SSA is assumed 1. This is one ice optical model, not measured event microphysics; no polarization, gas absorption, Rayleigh, aerosol, ground or real 3D heterogeneity is present.
Packet flights use exact cylinder intersections, physical path lengths and a nadir next-event estimator with an analytic direct term. Source-proposal importance weights preserve the stated isotropic emission. Production had zero truncated histories, zero negative-delay violations and zero reported 12 ms kernel overflow; maximum weighted energy residual was 2.91e-15 per packet. These checks do not certify directional variance convergence.
All 24 nonzero GHM rows remain flagged for nonuniform point-nadir convergence; 10 have estimated fluence standard error above 5%. At source 15 km, total optical depth 10, the production GHM gain 1.666+/-0.118 changes to 2.004+/-0.099 with a larger-packet control. A source-cone control gives 1.805+/-0.132. These are estimator diagnostics, not confidence bounds on a physical cloud. Rare contributions can escape a small-batch SEM estimate. No physical GHM/HG rise-time bias is inferred.
Two additional exactly matched ideal HG controls were compared to retained CloudScat results. Their g=0.87098, source 16 km, 12-18 km cylinder and 20 km radius differ from the production grid. Directional gains agree within 0.41 and 0.69 combined batch standard error at overlying optical depth 3 and 10. Timing postprocessing is shared, so this is not an independent timing-estimator validation. GHM and real-ice propagation remain unvalidated.
Retained nonpassing evidence includes a deliberately inadequate 10-scatter cap, packet/proposal/radius sensitivity, and a vacuum-metadata repair. Early code/serialization attempts are logged. Later diagnostic code changes are distinguished from the executed production snapshot; failed controls were not deleted.
Calculation limitations
Conditional model comparison, not a universal transmission curve or inversion.
Equal g does not fix higher phase-function moments. This grid supplies conditional estimates with MC errors; the current GHM rare-path controls do not establish a physical timing bias.
Two matched homogeneous-HG controls from the new engine were compared with retained CloudScat outputs and agree within their combined Monte Carlo errors. This independent-code check does not validate tabulated GHM, atmospheric transport or a real-cloud classifier; own analytic tests are labeled separately.
The native processed phase grid is assumed. Interpolation, ice habit and size-distribution uncertainties are not represented by MC error.
No actual event-brightness population, absolute rate, current calibration, flight noise or rise-time classifier is provided.
The tabulated GHM point-nadir rare-path estimator is not uniformly converged. Representative increased-packet and source-proposal controls can differ more than naive finite-batch SEM suggests. Do not infer a physical GHM/HG rise-time bias from unconverged ratios. All original batches and controls are retained.
GHM point-nadir rare-path convergence and nonlinear timing uncertainty
UV ice habit, roughness, size-distribution and interpolation representativeness
Real ice absorption and atmospheric Rayleigh/ozone/aerosol paths
3D cloud structure and source angular/temporal/spectral distribution
Observed noisy onset, eventpopulation classifier and flight detection rates
Batch SEM is an estimate, not a guaranteed 95% interval. Source and radius controls may share random streams; do not assign independent-difference significance to those comparisons.
Aperture transfer has unitsm^-2 per emitted photon. Multiply by actual aperture area for photon fraction, and only then by appropriate PDE for detected events. Directional gain above 1 is not a photon probability above 1.
Sources for this question (7)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Multiple scattering and temporal profile; robustness of 30-80 us versus>200 us rise-time discrimination.
Direct answer
Rise time contains information about photon paths, but it is not a unique depth measurement. The observed edge combines source history, cloud scattering, electronics, sampling and noise. The proposed fast/slow separation is still a hypothesis requiring independent depth labels. At 10 kS/s the target fast edge is under-resolved; even 50 kS/s supplies few samples and may have very few leading-edge photons.
Physics 101
A cloud is a light-delay machine: wandering photons travel farther than direct photons. Deeper flashes often have more opportunity to wander. That makes rise time a useful clue about depth.
But it is a clue shared by several causes. A slow source can look slow even near cloud top. A deep source with a favorable path can show an early component. Electronics can smooth the pulse, and a detector that misses closely spaced photons can distort its shape. Knowing the pulse lasted milliseconds does not guarantee that enough photons arrived during its first 30 microseconds to measure that rise.
We must also agree on what rise time means. Time from first detectable light to the peak is different from time between 10% and 90% of the peak. Comparing those two definitions as though they were the same would create a false physical conclusion.
Sampling is like taking photographs of a moving hand. One photograph every 100 microseconds cannot directly show the detailed motion during 30 microseconds. Faster samples help, but a useful test must include random alignment of the pulse with the sample clock, actual photon noise, the analogue filter and the fitting rule. The correct next step is to inject plausible shallow and deep waveforms through the complete model and measure the confusion between classes.
One new test detected all 1000 example pulses yet assigned a sufficiently separated fast/slow label to only 21 of them. That is not a failed detector: it shows that recognizing a flash and knowing its leading-edge shape require different amounts of information. The test used invented but explicit pulse families, so it does not certify cloud-depth discrimination.
Full technical answer
Multiple scattering can lengthen the optical path and produce a delayed rise and tail. It can therefore carry information about source depth. It does not supply a universal one-to-one map from rise time to depth. Physical depth, extinction, angular scattering, cloud boundaries and the intrinsic emission history all enter the observed waveform.
The proposed 30-80 microsecond versus greater-than 200 microsecond separation is a hypothesis to validate, not a demonstrated classifier at 550 km. The literature uses different rise definitions, including onset-to-peak and 10-90% amplitude. They cannot be interchanged. The measured signal is the source history convolved with cloud transport, detector/electronics response and sampling/integration. A source can keep exciting nitrogen after its earliest light; a long tail need not be solely propagation.
The universal greater-than 3 ms source duration is also unsupported: published shallow blue-flash examples include 200-910 microsecond durations, and multipulse events can contain renewed source activity. Nitrogen excited-state lifetimes and collisional quenching are fast, while continuous/repeated excitation can make a longer optical event. A fit assuming an instantaneous source does not independently prove that assumption.
At 10 kS/s the 100 microsecond sample spacing under-resolves a 30-80 microsecond edge. At 50 kS/s, 20 microsecond spacing gives only 1.5-4 samples. Model fitting can extract information between samples only with a valid pulse/noise/phase model. At 50 kS/s the Nyquist frequency is 25 kHz; an analogue 50 kHz bandwidth cannot be treated as unaliased baseband. Photon arrivals on the leading edge, not merely total event photons, determine timing information. Dead time and saturation can also make an observed rise artificially short.
The fresh 81,000 trial receiver-morphology study makes the detect-versus-classify distinction concrete. With 50 kS/s, background 50/s and 100 plateau-equivalent PE, true 30 us pulses were detected in 1000/1000 trials, but only 21 were accepted by the declared fast/slow score-margin rule. With zero errors among those 21, the upper endpoint of an individual two-sided exact 95% error interval is still about 16.1%. With 1000 scale PE, the accepted fractions improve but remain source-shape/background dependent. These are synthetic receiver-input shapes, not independent cloud-depth labels.
Corrections to earlier assumptions
Keep intrinsic, cloud-propagated and sampled rise times separate.
Keep onset-to-peak and 10-90% definitions separate.
A tail or a fast leading edge alone does not uniquely identify depth or discharge class.
Equations and their conditions
A2 · expression 1
y = sample{ s * K_cloud * h_receiver } + noise
The star denotes convolution; every stage can alter timing.
Conditions: Linear response before any explicitly modeled dead time or saturation.
A2 · expression 2
delta_t = (L_path - R_straight) / c
Extra path length gives propagation delay relative to the direct ray.
Conditions: Bulk transport at approximately vacuum light speed; source-to-observer flight time is not rise time.
What supports this answer
measured examples and model assumptions
Published source durations and fitted propagation examples have distinct definitions and selection.
Consistent rise definitions and independently labeled source depths/classes.
Intrinsic pulse families, realistic cloud phases/geometry and calibrated receiver response/noise.
Injection-recovery over actual leading-edge counts, background, sample phase and count losses.
Related executed studies (6) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N02 · An above-top reflector can increase nadir light and delay it
executed conditional calculation not mission measurement
Re-ran 108 analytic cases for an isotropic point above a flat instantaneous Lambert reflector. Independent radial integration and escape/threshold checks passed. In the surface limit, nadir fluence relative to vacuum is 1+2 rho while total upward escape is(1+rho)/2; directional gain therefore does not create photons.
The reflected waveform is a geometric convolution with a linear rise to an indefinite plateau. Its 10-90 rise is not an impulse arrival percentile or onset-to-peak duration.
No mappingofrhotooverheadcloudtau. h=0 meanslimitingjust-abovesurfaceposition, notanembeddedcloud-topscatteringboundary
R4-N03 · Depth and diffusion strength are not uniquely identifiable
executed conditional calculation not mission measurement
Re-ran 48 ideal half-space diffusion/absorption examples. The first-exit density is L/sqrt(4 pi D) t^(-3/2) exp[-L^2/(4 Dt)-gamma t]. Cases with identical L^2/D andgamma have the same entire kernel.
The computed leading 10-90 coefficient is 0.0743277 L^2/D, close to the published rounded 0.0742. The nonabsorbing mode is L^2/(6 D); their ratio is 0.44597, so the definitions are quantitatively different.
This diffusion density is a plane first-exit quantity integrated over transverse position, not finite-aperture directional radiance. Diffusion has noncausal early support and cannot replace transport near the ballistic front.
Model Dvaluesandtop 18 kmincounterexampleareconditional, notindependentheightobservations
Published 5.5 usmodel-basedsubsamplefitdoesnotprove general 5.5 usmeasurementprecisionat 10 usbinning
Exact L²/D whole-kernel degeneracy is a property of the parabolic diffusion PDE; finite-speed light transport can break it.
All parabolic kernels have nonzero pre-flight tails. Passing t 10>=L/c does not validate early-time radiative transport.
0.07432773536 versus the paper's 0.0742 is not an established published error; original Eq 6 bitmap was not retrieved.
R4-N04 · Raw dead time can compress an apparent rise
executed conditional calculation not mission measurement
Re-ran 84 nonparalyzable quasistatic mean-response cases. For a linear ramp, raw normalized output becomes f(1+x)/(1+xf), x=S tau/(1+b tau). Its 10-90 rise divided by the true linear-ramp rise is(1+x)/(1+x+0.09 x^2).
This diagnostic concerns raw uncorrected expectation. It does not demonstrate bias in a published corrected MMIA waveform, does not restore individual photons and does not produce a valid Poisson likelihood after correction.
Calculation limitations
Only RAW uncorrectedmeanwaveformshape; publishedcorrected MMIA waveformscannotbeaccusedofthisbiasfromthisdiagnostic
R4-N09 · Detection is easier than trusted morphology classification
executed conditional calculation not mission measurement
Executed 81,000 synthetic signal trials across 81 scenarios: receiver-input 10-90 rises 30, 80, 200 us; plateau-equivalent 3 ms scales 100, 1000, 10000 PE; background 50, 10000, 1e6/s; sampling 10, 50, 100 kS/s. Amplitude, onset and background are fitted using a finite 10 ms prewindow and explicit gates.
A predefined profile-log-likelihood margin ln 19 selects accepted classifications. It is not automatically a calibrated 95% posterior probability or a 95% correctness guarantee. Rejected events, wrong accepted labels and interval uncertainty are all reported.
The plateau-equivalent scale is rate times 3 ms. Actual source totals are smaller because the linear ramp uses part of the finite 3 ms pulse. These are receiver-input morphology trials, not source-depth/class validation. Count-loss occupancy flags are diagnostics only; the generator remains ideal independent Poisson.
Independent profile-objective comparisons and exact seeded reruns passed. The known-template oracle has extra information and is an optimistic benchmark, not an achievable estimator promise.
Calculation limitations
This is a conditional synthetic receiver-morphology study, not a flight sensitivity/completeness validation.
The classes mean 30/80/200 us 10-to 90% rises at a hypothetical linear receiver input after transport; they do not label source depth.
Source brightness, cloud transfer, line spectrum, detector PDE and source occurrence population are outside this count-level calculation.
The practical code has an external coarse onset bracket and a specified event gate. Blind stream triggering and a mission false-alarm rate are unvalidated.
A fixed ln 19 profile-score margin is not calibrated 95% confidence. Always show accepted fraction, rejected fraction and the error interval together.
The 3-class fast-subtype distinction is harder than the grouped fast-vs-slow test; these are separate decisions.
Only exact model-matched templates and constant backgrounds are generated. Real independently labeled optical/radio controls are still needed.
Post-counting electronic filtering, detector excess noise, afterpulses, saturation and deadtime are not simulated.17 ns occupancy is an illustrative hardware flag, not Poisson restoration.
Known-background red limits are invalid if a noisy finite baseline is silently substituted. The provided alternative uses a conservative Bonferroni construction.
Red limits constrain expected registered counts in a specified window, not thermal-source absence. Selection, red-window timing, throughput and population ratios remain physical inputs.
A useful mission event rate is NOT COMPUTABLE without an absolute received-fluence distribution, completeness and exposure definition.
No empirical wide-FOV event rate is inferred from the conditional geometric/population proxy; source geography and coverage remain unknown.
R4-N11 · Independent cylinder transport and the unresolved rare-path test
executed conditional transport with incomplete GHM convergence
Executed 56 phase/geometry cells: 48 nonvacuum Monte Carlo cases plus 8 analytic vacuum cases. Production used 14.4 million histories in 288 independent batches. A homogeneous finite cylinder spans 14-17 km with 30 km radius; source heights 15, 16, 17, 18 km; observer 550 km and 60 degree circular field. Total vertical particle optical depth takes 0, 0.3, 1, 3, 10, 30, 45.
The scalar GHM bulk phase uses 10 micrometre effective radius and linearly interpolates 330/340 nm optical-model data at the 337.1 numerical coordinate. Its derived g is 0.7709447; the comparison HG has the same g. SSA is assumed 1. This is one ice optical model, not measured event microphysics; no polarization, gas absorption, Rayleigh, aerosol, ground or real 3D heterogeneity is present.
Packet flights use exact cylinder intersections, physical path lengths and a nadir next-event estimator with an analytic direct term. Source-proposal importance weights preserve the stated isotropic emission. Production had zero truncated histories, zero negative-delay violations and zero reported 12 ms kernel overflow; maximum weighted energy residual was 2.91e-15 per packet. These checks do not certify directional variance convergence.
All 24 nonzero GHM rows remain flagged for nonuniform point-nadir convergence; 10 have estimated fluence standard error above 5%. At source 15 km, total optical depth 10, the production GHM gain 1.666+/-0.118 changes to 2.004+/-0.099 with a larger-packet control. A source-cone control gives 1.805+/-0.132. These are estimator diagnostics, not confidence bounds on a physical cloud. Rare contributions can escape a small-batch SEM estimate. No physical GHM/HG rise-time bias is inferred.
Two additional exactly matched ideal HG controls were compared to retained CloudScat results. Their g=0.87098, source 16 km, 12-18 km cylinder and 20 km radius differ from the production grid. Directional gains agree within 0.41 and 0.69 combined batch standard error at overlying optical depth 3 and 10. Timing postprocessing is shared, so this is not an independent timing-estimator validation. GHM and real-ice propagation remain unvalidated.
Retained nonpassing evidence includes a deliberately inadequate 10-scatter cap, packet/proposal/radius sensitivity, and a vacuum-metadata repair. Early code/serialization attempts are logged. Later diagnostic code changes are distinguished from the executed production snapshot; failed controls were not deleted.
Calculation limitations
Conditional model comparison, not a universal transmission curve or inversion.
Equal g does not fix higher phase-function moments. This grid supplies conditional estimates with MC errors; the current GHM rare-path controls do not establish a physical timing bias.
Two matched homogeneous-HG controls from the new engine were compared with retained CloudScat outputs and agree within their combined Monte Carlo errors. This independent-code check does not validate tabulated GHM, atmospheric transport or a real-cloud classifier; own analytic tests are labeled separately.
The native processed phase grid is assumed. Interpolation, ice habit and size-distribution uncertainties are not represented by MC error.
No actual event-brightness population, absolute rate, current calibration, flight noise or rise-time classifier is provided.
The tabulated GHM point-nadir rare-path estimator is not uniformly converged. Representative increased-packet and source-proposal controls can differ more than naive finite-batch SEM suggests. Do not infer a physical GHM/HG rise-time bias from unconverged ratios. All original batches and controls are retained.
GHM point-nadir rare-path convergence and nonlinear timing uncertainty
UV ice habit, roughness, size-distribution and interpolation representativeness
Real ice absorption and atmospheric Rayleigh/ozone/aerosol paths
3D cloud structure and source angular/temporal/spectral distribution
Observed noisy onset, eventpopulation classifier and flight detection rates
Batch SEM is an estimate, not a guaranteed 95% interval. Source and radius controls may share random streams; do not assign independent-difference significance to those comparisons.
Aperture transfer has unitsm^-2 per emitted photon. Multiply by actual aperture area for photon fraction, and only then by appropriate PDE for detected events. Directional gain above 1 is not a photon probability above 1.
R4-N12 · Conditional transported pulse timing and electronics
executed conditional HG waveform diagnostic not depth classifier
Convolved each retained kernel with assumed intrinsic 10-90 rises 30, 80 or 200 us in an exponential-decay source family. Applied no electronics or an ideal first-order 10/50 kHz low-pass. These cutoff frequencies are not ADC sample rates or validated antialias designs.
The table shows only HG examples to avoid presenting unconverged GHM differences as physics. A source 2 km below the modeled top can still have a 30 us input broadened to about 48 us in one HG case. This is a conditional counterexample to a universal geometric-depth threshold, not evidence that an actual deep NBE will always look fast.
Both 10-90 rise and known-model-onset-to-peak are retained separately. Model curves have known onset and no photon noise. Near-window and bin-placement checks have limited scope; they do not validate observed onset, unknown intrinsic pulses or population classification.
Calculation limitations
HG is a comparison phase model, not an event-specific ice measurement. GHM timing remains numerically unconverged.
Assumed source decay 3 ms is not a universal observed duration. Output normalized to its own peak hides amplitude differences intentionally.
Known-model onset, 10-90 rise, arrival quantiles and noisy inferred onset are different quantities.
No detector dead time, correlated noise, uncertain baseline or sampling-phase likelihood included in this transport convolution.
Sources for this question (8)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
A verified total polar, cloud-viewing 337 nm night-background radiance is still missing. The 1.77 × 10⁻⁹ W m⁻² sr⁻¹ nm⁻¹ example is a processed ground pseudo-continuum reference, not total orbital background. Airglow, aurora, moonlight and reflected stellar/zodiacal illumination must be combined with their own geometry and the actual spectral-angular response.
Physics 101
Background is all the light arriving while no target flash is present. It matters even if we subtract its average: the number of background photons fluctuates. A brighter steady background therefore makes a faint flash harder to recognize.
Different instruments measure different kinds of brightness. A wide ultraviolet channel combines many wavelengths. A spectrometer separates them. A ground instrument looks upward through a different atmosphere from a satellite looking down at cloud. Dividing a broad measurement by 100 nm silently assumes the light is evenly distributed across that range; atmospheric emission often is not.
We recovered useful real reference spectra. One ground-based mean gives about 1.8 billionths of a watt per square metre, steradian and nanometre around our 4 nm band after the authors' processing. That is a scale example, not a promised polar background. The starlight map is also useful, but the cloud reflects a whole hemisphere of celestial light with uneven brightness.
The missing result is the distribution of total background for the actual field, orbit, cloud and filter. A noise budget should show separate quiet, moonlit and auroral scenarios with clear assumptions, then replace those scenarios with representative measurements when available.
Full technical answer
No single verified polar, cloud-viewing 337 nm background radiance was established. The needed quantity is a spectral and angular scene radiance folded with the actual receiver response. Quiet airglow, aurora, reflected moonlight, zodiacal illumination and direct/diffuse starlight have different geometries and cannot be replaced by one broadband number divided by a chosen width.
A useful bounded example is the published Noll et al.(2024) mean pseudo-continuum derived from 10, 633 selected Paranal spectra. Folding that table with a unit top-hat 335.1-339.1 nm band gives about 7.08e-9 W m^-2 sr^-1 integrated, or 1.77e-9 W m^-2 sr^-1 nm^-1 averaged over 4 nm. This is a processed low-latitude ground reference: strong resolved lines and modeled extraterrestrial components were removed, observations are long integrations, and extinction/layer geometry were corrected. It is not total sky, a quiet polar distribution or a satellite-cloud measurement.
The Bianchi(2024) direct-plus-diffuse starlight cube gives about 8.23e-10 W m^-2 sr^-1 nm^-1 as an all-sky mean interpolated near 337.1 nm. Its roughly 13 nm spectral bracketing is too coarse to certify a 4 nm stellar spectrum. An Earth-filled nadir instrument generally receives this component after atmospheric/cloud reflection, not as a direct celestial view. The campaign evaluated hypothetical cloud-normal hemispheres, not mission orbit/date illumination.
Public RAIDS near-UV spectra were decoded and their count/calibration relation checked privately. Their limb geometry, sequential wavelength scan, preflight calibration, additive backgrounds and publication terms prevent treating selected rows as clean 4 nm nadir radiance. TEMPO and Mini-EUSO give additional measured context, but their processing, broad responses and viewing geometry differ. These are constraints and research routes, not interchangeable background measurements.
Corrections to earlier assumptions
Noll pseudo-continuum is not total night sky.
Broad UV orbital counts are not a measured 4 nm 337 nm polar radiance.
Limb spectra and cloud-reflected nadir brightness require different transfer.
Equations and their conditions
B4 · expression 1
b = integral A_eff(lambda,Omega) L_lambda(Omega) lambda/(h c) d lambda d Omega
Background detected photon rate is a spectral-angular integral.
Conditions: A_eff includes projected collecting area, optical throughput and photon detection probability exactly once.
B4 · expression 2
1 R = 10^10/(4 pi) photons m^-2 sr^-1 s^-1
Rayleigh is an emission radiance unit, unrelated to Rayleigh scattering as a mechanism.
Conditions: For a spectral value, retain whether it is pernm or per Angstrom.
What supports this answer
published processed measurement
Noll mean continuum table can be folded conditionally; its selection/subtractions limit mission transfer.
A total 337 nm scene distribution with measured components and geometry at high latitudes.
Actual filter/field response, cloud directional reflection and gas transmission.
Related executed studies (1) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N07 · Actual reference continuum and starlight data folds
executed conditional calculation not mission measurement
Re-ran Noll processed-mean spectral folds and Bianchi coarse-spectral sky-cube hemisphere examples. Source data are real/model-published inputs, but the target geometry and ideal filters are conditional.
The Noll extraction removes strong lines and modeled celestial components. Bianchi covers stellar components, not Moon, zodiacal or airglow, and its spectralgrid near 337 nm is much coarser than 4 nm. Neither is a total polar scene distribution.
Calculation limitations
Mean 10633 selected 2009-2019 Cerro Paranal spectra, >=10 min exposure; Moonabovehorizonwithillumination>50% excluded. Not a polar or microsecond distribution.
Continuum extracted by median/percentile filters; strong resolved lines removed, unresolved O2 UV bands retained partly. Extraterrestrial components modeled/subtracted; extinction and van Rhijn 90 km corrected. Not total night sky or total 777 line background.
Header only says wavelengthnm. Official ESO says Xshooter pipeline wavelength in air; authors could have subsequently converted, not explicitly verified here. Fold centre 337.1 uses table convention; photonenergyuses nominal wavelength with <0.03% air/vac sensitivity.
4 component NMF is statistical; labelledcomponents must not automatically be treated as puremolecular emissions.
Totalmeans direct+interstellar-dust-scattered starlight; no zodiacal, moon, airglow. Opticalphotometry+stellar-atmosphere/exctinction extrapolation, not a narrow 337 measurement.
99 logbins span~0.1-4.3 um: around 337 bracket~13 nm wide. Cannot infer actual 4 nm spectralfeatures or own covariance from tabulated broad-model errors.
Hypotheticalsix cloudnormaldirections, noactualorbit/date/Sun/moon/cloudprofile. Not globalminimum/maximum or occurrenceweighteddistribution.
Earthfilled nadir does not see this external field directly; need hemisphericillumination and cloud BRDF/atmosphericscattering. Lambertalbedo.8 is an explicit example only.
Cubeis HEALPix NESTED order 6 in celestial RA/Dec despite paper figures plottedin Galacticcoordinates. Do not reinterpret cube in Galacticcoordinates.
Sources for this question (7)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Auroral 337 nm 4 nm-band brightness and shot-noise floor during moderate/disturbed conditions.
Direct answer
Aurora emits in the same nitrogen band, so spectral filtering cannot distinguish those photons from the target band. It raises shot noise even after its mean is subtracted, and some auroral structure is faster than seconds. Historical brightness examples and conditional count-load calculations are available; a defensible moderate/disturbed-Kp-to-nadir-337-nm distribution is not.
Physics 101
An auroral 337 nm photon and a discharge 337 nm photon are identical to the detector. Their different origins do not give the filter a way to distinguish them.
A slowly changing glow can be estimated from nearby times and subtracted. But photon arrivals are random: a subtraction cannot remove the fluctuations in how many photons happened to arrive during the flash. This is why aurora can raise the noise floor even when its changing shape is easy to recognize.
An index such as Kp summarizes geomagnetic activity over a broad region and time. It does not tell us exactly whether our field contains a bright narrow arc or a dim patch, or what electron energies produced the emission. That is why a single background number for moderate conditions would claim more knowledge than we have.
We can still make useful engineering calculations: choose explicit received brightnesses, compute the resulting background counts and measure how many flash photons are needed. Those calculations answer what happens if the background is a stated value. They do not tell us how often that value occurs on the proposed orbit. Fast auroral structure and particle artifacts also need controls beyond a simple slow-baseline subtraction.
Full technical answer
Auroral nitrogen emission falls in the same 337 nm band as the target, so a filter cannot remove it selectively. It raises photon shot noise and may add structured temporal variations. High-latitude operation makes this an important scenario, but no defensible universal conversion from moderate/disturbed Kp to 4 nm nadir radiance was established.
A selected historical spectrum reports a 337 nm band area 2270 +/-290 R with an additional 25% calibration-scale uncertainty. Nearby readings from another historical photometer were about 596-696 R. These differ in observable, weighting, field and calibration; no authoritative correction or universal auroral range was derived. The numbers are selected examples, not a distribution or a statement that every 4 nm filter captures the full reported band area. The numerical appendix may use imposed received radiances to show shot-noise consequences, while labeling them as scenarios.
GLOW contains useful 337/777 emission outputs, with brightness depending on imposed precipitating-particle spectra and flux, atmosphere and viewing geometry. TRE x-ATM's paper-level wavelength capability and the current public API fields must be distinguished: the inspected API did not expose arbitrary 337/777 outputs. OVATION flux/power categories may supply conditional precipitation inputs; they are not optical radiance measurements. Proton/hydrogen excitation and secondary electrons can also matter.
Aurora should not be assumed perfectly constant on every relevant timescale. Primary observations demonstrate subsecond structure and flickering. They do not establish a 337 nm microsecond noise spectrum for this wide field. Even a perfectly steady mean leaves Poisson shot noise; baseline subtraction removes its mean, not that random variance. Geomagnetic classification alone does not specify arc filling fraction, latitude/local time, cloud reflection or the receiver's background.
Corrections to earlier assumptions
Same-wavelength auroral photons cannot be filtered away from the target.
Slow mean variability does not eliminate shot noise, and aurora is not universally slow.
An imposed precipitation/model case is not a measured Kp-conditioned sky distribution.
Equations and their conditions
B5 · expression 1
Var(N_background) = b T
Ideal Poisson background counts have variance equal to their mean.
Conditions: Independent arrivals, known live time and no correlated detector noise; estimating the baseline adds variance.
B5 · expression 2
L_band = R_band [10^10/(4 pi)] h c / lambda
At 337.1 nm, 1 R corresponds to about 4.69e-10 W m^-2 sr^-1 integrated.
Conditions: Band-integrated Rayleigh, not R pernm; source spectral capture must be specified.
What supports this answer
selected measurement
Historical band-area and photometer intensities are examples with material observable/calibration differences.
A calibrated 4 nm 337 nm radiance/time distribution for the actual high-latitude field and scene.
Precipitation spectra, arc filling, cloud transfer and actual detector end-of-life noise.
Related executed studies (1) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N05 · Auroral/background amplitude scenarios and exact Poisson thresholds
executed conditional calculation not mission measurement
Re-ran 192 ideal detector-count scenarios with uniform received band-equivalent brightness 15, 100, 1000 or 2270 R, eta 0.2 and 50 darkcounts/s. A 2270 R amplitude is a selected historical scale, not a received uniform 4 nm polar measurement.
Exact Poisson tails set critical counts and 95% detection power. A hypothetical operational budget assumes 50000 starts/s, 86400 s and 3 families for 0.1 false trigger per day; actual correlated/overlapping searches and nuisance parameters need their own allocation.
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Lunar phase/elevation at which cloud background dominates detector noise.
Direct answer
Express the moonlight limit as a received radiance or count-rate crossover first. There is no universal phase/elevation threshold: it depends on the receiver, cloud reflection, lunar irradiance, geometry and the estimator. Several lunar standards do not demonstrate their quoted accuracy at 337 nm, so a validated UV lunar model is still needed.
Physics 101
Moonlight is reflected sunlight. How much reaches our detector depends on how brightly the Moon illuminates the cloud, how the cloud sends that light upward, and how much sky the instrument accepts.
A large field collects more background from many directions. A wider filter accepts more background wavelengths. A warmer detector may generate more false internal counts. Changing any of these can change the Moon condition at which background becomes the main limitation.
The phrase background limited means the random fluctuation in arriving background light exceeds the detector's own fluctuation. It does not mean the background makes detection impossible: a sufficiently strong flash can still stand out. It also does not identify one lunar phase without a design and a scene model.
We found good lunar calibration work, but precision demonstrated at visible wavelengths must not be copied to 337 nm. The practical next step is to calculate several explicit moonlit-cloud scenarios and identify the maximum received background the receiver can tolerate, then test how often the orbit encounters it.
Full technical answer
There is no universal lunar phase or elevation at which photon background overtakes detector noise. The crossover depends on aperture, field, filter equivalent bandwidth, optical response, detector temperature/dark statistics and estimator duration. Moon elevation at the cloud and lunar phase/libration affect incident irradiance; cloud reflectance and atmospheric paths determine what reaches the satellite.
For a Lambertian example L_lambda = rho E_Moon, lambda cos(z_Moon)/pi, with transmission factors and a real directional reflectance model added as needed. This is a conditional scene model, not a measured anvil BRDF or universal lunar spectrum. A useful crossover compares variance after the same estimator: background photon variance against dark/electronic/correlated variance, not mean optical power against amplifier current without conversion.
The calibration literature has important wavelength limits. ROLO's fitted observations begin around 350 nm. Air-LUSI measured 350-980 nm and demonstrated its strongest stated <=1% result over 450-970 nm, not 337 nm. LIME's actual available observed bands do not establish a 337 nm absolute reference merely because a broader model description is given. A proposed 320 nm-capable LUSI system is not a demonstrated 0.5%337 nm lunar standard. GOME/LROC and LESSR offer UV routes but require actual wavelength, phase, absolute-scale and data-access qualification.
The current design should therefore carry lunar scenarios with explicit uncertain 337 nm irradiance and cloud transfer, and report a crossover in received radiance/count rate. Convert that into phase/elevation only after choosing and validating a UV lunar model and receiver. Cloud, Moon and orbital night selection are correlated; multiplying independent marginal duty-cycle fractions is generally unjustified.
Corrections to earlier assumptions
Do not transfer visible-band lunar absolute accuracy to 337 nm.
Background-limited is a comparison of noise variances for a specified receiver and estimator.
Equations and their conditions
B6 · expression 1
L_Moon,lambda = rho E_Moon,lambda cos(z_Moon) / pi
Ideal Lambertian reflected spectral radiance.
Conditions: Homogeneous diffuse reflector; specified albedo and transmission, no claim of anvil angular behavior.
B6 · expression 2
background limited when V_b > V_detector
Compare variance after the same integration/filter/estimator.
Conditions: Include baseline-estimation covariance and detector correlations.
What supports this answer
published calibration with wavelength limits
ROLO, LIME andair-LUSI references have specific demonstrated bands and geometry.
Validated 337 nm lunar irradiance versus phase/libration and cloud illumination geometry.
Chosen receiver noise/temperature, spectral-angular response and cloud reflection.
Related executed studies (1) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N05 · Auroral/background amplitude scenarios and exact Poisson thresholds
executed conditional calculation not mission measurement
Re-ran 192 ideal detector-count scenarios with uniform received band-equivalent brightness 15, 100, 1000 or 2270 R, eta 0.2 and 50 darkcounts/s. A 2270 R amplitude is a selected historical scale, not a received uniform 4 nm polar measurement.
Exact Poisson tails set critical counts and 95% detection power. A hypothetical operational budget assumes 50000 starts/s, 86400 s and 3 families for 0.1 false trigger per day; actual correlated/overlapping searches and nuisance parameters need their own allocation.
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Detector choice at 337 nm, large area and 10-50 kHz; claimed photodiode/TIA floor.
Direct answer
Compare a flight-like PMT counting/analogue chain with a specified semiconductor chain under equal aperture-photon injections. A universal silicon-photodiode floor of 1 nA does not follow from one module’s gain-bandwidth product. PMTs offer area and gain; SiPMs/APDs require dark-noise, capacitance and radiation treatment; CMOS must demonstrate the required accepted-area cadence and summed read noise.
Physics 101
The detector is the light-to-electricity converter. A photodiode releases a small electrical charge when it absorbs light. A PMT, SiPM or APD multiplies that charge internally, making it easier for later electronics to see.
Internal multiplication is useful, but it is not free sensitivity. The detector can create false signals, multiplication can fluctuate, and a busy detector can miss photons. Making a solid-state sensor larger usually adds false counts and capacitance as well as useful area.
The earlier one-nanoampere floor came from treating one amplifier specification as a law of nature. Better amplifier arrangements exist. The remaining question is whether a real low-noise amplifier still works well with the chosen large detector and in the available power, temperature and radiation environment.
A camera has another trade-off: thousands or millions of small pixels must be read and added. A low-noise pixel does not imply a low-noise full camera at very high speed. The fair comparison uses the same collected photons, field, pulse, temperature and background for each complete receiver. That measurement has not yet selected a winner.
Full technical answer
The proposed silicon-photodiode exclusion is not a universal law. Transimpedance gain, amplifier open-loop gain, detector/input capacitance, feedback compensation and multiple gain stages determine bandwidth and stability. A quoted product for one module cannot set a physical ceiling for every architecture. A 2026 bench study reports 1 Gohm total gain with 30 kHz output filtering and 3.898 pA RMS open/shielded-input noise; this refutes a universal 1 nA floor but does not establish our detector-coupled 337 nm sensitivity or flight suitability.
A PMT offers large receiving area and internal multiplication with low electronic readout burden; its costs include HV, magnetic sensitivity, bright-light protection, gain/collection calibration and count/charge dynamic range. SiPM s offer solid-state gain and modular area, but dark counts, capacitance, optical crosstalk, afterpulsing, recovery and radiation damage grow important as area increases. Linear APDs provide gain without Geiger holdoff but require excess-noise, capacitance, bias/temperature and amplifier treatment. PIN and UV-selective SiC photodiodes remain candidates to evaluate, not devices already shown adequate.
A back-illuminated CMOS camera has many pixels. Its per-pixel read noise and short minimum exposure do not mean the whole active area can be summed at 50 kHz with the same noise. An inspected GSENSE400BSI specification gives 48 full frames/s, 1.6 electrons typical temporal noise per pixel and peak QE at 570 nm. Neither exact 337 nm QE nor fast full-area flight operation follows. ROI readout changes area/field unless the optics are redesigned.
Choose the detector together with the optics, background, required timing and gain range. For example, a 6 x 6 mm SiPM may have high PDE but cannot receive every ray from a 25 mm aperture over a 60-90 degree cone in air without collection loss. A small red APD has the same geometric issue. The preferred baseline is therefore a flight-like PMT counting/analogue comparison with at least one specified semiconductor chain, followed by equal aperture-photon injection tests; it is not a class-wide hardware purchase decision.
Corrections to earlier assumptions
A vendor gain-bandwidth product is not a universal detector limit.
Peak QE at 405, 450 or 570 nm is not measured 337 nm response.
Per-pixel noise and minimum exposure do not specify a 50 kHz large-area photometer.
Equations and their conditions
C7 · expression 1
V_out(f) = Z_T(f) I_photo(f)
Transimpedance is a frequency-dependent circuit response, not merely a gain number.
Conditions: Actual capacitance, stability, layout and feedback must be retained.
What supports this answer
bench measurement
A published high-gain TIA counterexample rejects the universal floor without qualifying this receiver.
Selected sensor/package 337 nm PDE or responsivity, actual sensitive area/capacitance and bias/temperature-dependent dark PSD or correlated-count distribution
Complete readout noise, impulse response, power and permitted readout mode; for CMOS, measured full accepted-area cadence and summed read noise, not minimum exposure alone
Next inputs in the reviewed register
Required aperture/field, temperature, sample rate in S/s, analog bandwidth and estimator ENBW as separate quantities
Related executed studies (2) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N08 · Optical/electronic design algebra with explicit assumptions
executed conditional calculation not mission measurement
Fresh design suite contains 36 etendue bounds, 372 hypothetical Gaussian source/filter cases, 900 sample-phase examples, 9 AC-droop rows, 58 OD allocations, 72 dark-only NEP scenarios and 192 browser-budget references. Forty-eight independent Gaussian quadrature checks agreed at about 4.4e-16 maximum absolute difference.
These are algebra and sampling diagnostics. The Gaussian source is not N2 line reconstruction; dark-only NEP omits sky, amplifier, correlations and radiation; ideal etendue is a necessary bound, not a ray-traced layout.
Calculation limitations
No flight filter optimum, no measured electronics noise, no hardware qualification and no source-population inference.
R4-N12 · Conditional transported pulse timing and electronics
executed conditional HG waveform diagnostic not depth classifier
Convolved each retained kernel with assumed intrinsic 10-90 rises 30, 80 or 200 us in an exponential-decay source family. Applied no electronics or an ideal first-order 10/50 kHz low-pass. These cutoff frequencies are not ADC sample rates or validated antialias designs.
The table shows only HG examples to avoid presenting unconverged GHM differences as physics. A source 2 km below the modeled top can still have a 30 us input broadened to about 48 us in one HG case. This is a conditional counterexample to a universal geometric-depth threshold, not evidence that an actual deep NBE will always look fast.
Both 10-90 rise and known-model-onset-to-peak are retained separately. Model curves have known onset and no photon noise. Near-window and bin-placement checks have limited scope; they do not validate observed onset, unknown intrinsic pulses or population classification.
Calculation limitations
HG is a comparison phase model, not an event-specific ice measurement. GHM timing remains numerically unconverged.
Assumed source decay 3 ms is not a universal observed duration. Output normalized to its own peak hides amplitude differences intentionally.
Known-model onset, 10-90 rise, arrival quantiles and noisy inferred onset are different quantities.
No detector dead time, correlated noise, uncertain baseline or sampling-phase likelihood included in this transport convolution.
Sources for this question (7)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
NEP and aperture photons per event for candidate detectors.
Direct answer
There is no defensible single operating NEP for each detector class at this bandwidth. Use the selected chain’s measured noise spectrum or count statistics, its actual 337 nm response and the estimator. Convert the required detected signal to aperture photons through the full spectral-angular efficiency. Detection and accurate rise-time estimation have different photon requirements.
Physics 101
NEP is a way to express detector noise as an equivalent amount of incoming light. It is useful only when we also know the wavelength, bandwidth and measurement method.
Think of listening for a short sound. Noise quoted at one audio frequency is not the same as the total noise heard through a broad microphone. A long average and a quick measurement also have different noise. Optical pulse measurements work the same way.
For a photon counter, an even clearer question is: with this background and these false internal counts, how many extra detected photons are needed to recognize a flash reliably? Then we ask how many aperture photons were required to produce those detected photons. If efficiency is 20%, 500 incoming photons produce 100 detected photons on average, but this is an example conversion, not an established detection threshold.
A proper answer also states how often noise alone is allowed to trigger. Searching continuously creates many opportunities for an accidental pattern. The new statistical studies make those assumptions explicit rather than turning a convenient noise-density number into an unsupported photons-per-event promise.
Full technical answer
NEP is input optical power that produces signal comparable to noise under a stated convention. A spectral NEP in W/sqrt(Hz) is not a bandwidth-integrated power, event energy or photon threshold. A spot value at 1 kHz cannot be multiplied blindly by sqrt(50000) when the noise spectrum, analogue filter and estimator weighting are not flat. Count-mode detectors may be better described by DCR, PDE, correlated-count statistics and a likelihood.
Use measured input-current PSD divided by actual spectral responsivity for a linear chain, then integrate through the transfer function. For APDs, distinguish unmultiplied surface leakage from multiplied bulk current: the Hamamatsu technical model gives 2 q[(I_L+I_dg)M^2 F+I_ds]B. Multiplying an already multiplied catalogue total current by M^2 can double-count gain. SiPM crosstalk/afterpulses and PMT multiplication variance require their own treatment.
The fresh design appendix gives dark-only examples using explicitly assumed dark-count, photon-efficiency and excess-noise scenarios. They exclude amplifier and optical background and are not promised operating NEPs. Likewise the preliminary SUR500x table gives 18 fW/sqrt(Hz) at 350 nm, gain M=100, not a qualified 337 nm receiver value. The actual temperature and noise-frequency conditions must be supplied before using it for a flight requirement.
Translate a validated detected-signal requirement S into aperture photons with N_ap=S/eta only when eta is the appropriate spectral/angle/time-weighted efficiency. Fluence is N_ap/A for uniform illumination. Detection probability and false-alarm control determine S; a generic 5 sigma rule can be wrong at low counts or after searching billions of windows. A pulse with many total counts may still lack enough leading-edge counts to estimate its rise. No single NEP table by detector class can answer that combined detection-and-timing question.
For the fresh linear-ramp morphology example, 100 plateau-equivalent PE across 3 ms with a 30 us 10-90 rise means 99.375 actual expected source PE but only 0.4 expected source photoelectrons in the first 30 us. An integrated detection can therefore be secure while rise estimation remains weak.
Corrections to earlier assumptions
Separate spot noise density, integrated RMS, dark current, DCR and detection probability.
Do not apply photocathode gain or APD multiplication twice.
Aperture photons and registered counts use different reference planes.
Equations and their conditions
C8 · expression 1
sigma_P^2 = integral NEP(f)^2 |H_est(f)|^2 df
Input-equivalent power variance for a specified linear estimator.
Conditions: Consistent one-sided PSD convention and responsivity; include correlated terms as needed.
C8 · expression 2
N_ap = S_required / eta; Phi_ph = N_ap / A
Convert detected signal to incident photons and photon fluence.
Conditions: Efficiency must match source spectrum, field, timing and live time.
What supports this answer
device noise model
APD surface/bulk noise and multiplication require the correct current convention.
Actual receiver spectral noise/count statistics, gain/excess noise and target-band throughput at beginning/end of life
Injection-recovery versus aperture photon fluence under relevant backgrounds, integration windows, pulse shapes and false-trigger rules
Next inputs in the reviewed register
Reference plane for photons, spectral versus integrated NEP, filter/estimator transfer and correlated-noise treatment
Related executed studies (4) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N05 · Auroral/background amplitude scenarios and exact Poisson thresholds
executed conditional calculation not mission measurement
Re-ran 192 ideal detector-count scenarios with uniform received band-equivalent brightness 15, 100, 1000 or 2270 R, eta 0.2 and 50 darkcounts/s. A 2270 R amplitude is a selected historical scale, not a received uniform 4 nm polar measurement.
Exact Poisson tails set critical counts and 95% detection power. A hypothetical operational budget assumes 50000 starts/s, 86400 s and 3 families for 0.1 false trigger per day; actual correlated/overlapping searches and nuisance parameters need their own allocation.
R4-N08 · Optical/electronic design algebra with explicit assumptions
executed conditional calculation not mission measurement
Fresh design suite contains 36 etendue bounds, 372 hypothetical Gaussian source/filter cases, 900 sample-phase examples, 9 AC-droop rows, 58 OD allocations, 72 dark-only NEP scenarios and 192 browser-budget references. Forty-eight independent Gaussian quadrature checks agreed at about 4.4e-16 maximum absolute difference.
These are algebra and sampling diagnostics. The Gaussian source is not N2 line reconstruction; dark-only NEP omits sky, amplifier, correlations and radiation; ideal etendue is a necessary bound, not a ray-traced layout.
Calculation limitations
No flight filter optimum, no measured electronics noise, no hardware qualification and no source-population inference.
R4-N09 · Detection is easier than trusted morphology classification
executed conditional calculation not mission measurement
Executed 81,000 synthetic signal trials across 81 scenarios: receiver-input 10-90 rises 30, 80, 200 us; plateau-equivalent 3 ms scales 100, 1000, 10000 PE; background 50, 10000, 1e6/s; sampling 10, 50, 100 kS/s. Amplitude, onset and background are fitted using a finite 10 ms prewindow and explicit gates.
A predefined profile-log-likelihood margin ln 19 selects accepted classifications. It is not automatically a calibrated 95% posterior probability or a 95% correctness guarantee. Rejected events, wrong accepted labels and interval uncertainty are all reported.
The plateau-equivalent scale is rate times 3 ms. Actual source totals are smaller because the linear ramp uses part of the finite 3 ms pulse. These are receiver-input morphology trials, not source-depth/class validation. Count-loss occupancy flags are diagnostics only; the generator remains ideal independent Poisson.
Independent profile-objective comparisons and exact seeded reruns passed. The known-template oracle has extra information and is an optimistic benchmark, not an achievable estimator promise.
Calculation limitations
This is a conditional synthetic receiver-morphology study, not a flight sensitivity/completeness validation.
The classes mean 30/80/200 us 10-to 90% rises at a hypothetical linear receiver input after transport; they do not label source depth.
Source brightness, cloud transfer, line spectrum, detector PDE and source occurrence population are outside this count-level calculation.
The practical code has an external coarse onset bracket and a specified event gate. Blind stream triggering and a mission false-alarm rate are unvalidated.
A fixed ln 19 profile-score margin is not calibrated 95% confidence. Always show accepted fraction, rejected fraction and the error interval together.
The 3-class fast-subtype distinction is harder than the grouped fast-vs-slow test; these are separate decisions.
Only exact model-matched templates and constant backgrounds are generated. Real independently labeled optical/radio controls are still needed.
Post-counting electronic filtering, detector excess noise, afterpulses, saturation and deadtime are not simulated.17 ns occupancy is an illustrative hardware flag, not Poisson restoration.
Known-background red limits are invalid if a noisy finite baseline is silently substituted. The provided alternative uses a conservative Bonferroni construction.
Red limits constrain expected registered counts in a specified window, not thermal-source absence. Selection, red-window timing, throughput and population ratios remain physical inputs.
A useful mission event rate is NOT COMPUTABLE without an absolute received-fluence distribution, completeness and exposure definition.
No empirical wide-FOV event rate is inferred from the conditional geometric/population proxy; source geography and coverage remain unknown.
R4-N10 · Exact counts, finite-baseline limits and search control
executed conditional calculation not mission measurement
Executed 63 deterministic count-inference controls and separate coverage/null Monte Carlo diagnostics with 200000 trials per configuration. The mathematical model is independent Poisson counts with stationary background and known on/off exposure ratio.
Known-zero-background zero counts gives the 95% upper mean 2.9957. The specified conservative finite-baseline construction with zero on and zero off counts gives 3.6889 under equal alpha splitting. This difference is a procedure/model change, not conflicting physics.
Individual discrete confidence procedures may be conservative. The retained initial individual 99.9% Monte Carlo discrepancy was not hidden or rerolled; the declared simultaneous 99% family audit overlaps the analytic coverage enclosures. Monte Carlo intervals diagnose implementation; coverage follows from the stated mathematical construction.
Unconditional coverage need not remain 95% after selecting only non-detections or accepted classes. Per-test alpha is not continuous-search false alarm. Dead time, overdispersion, unknown calibration and varying background require more models.
Calculation limitations
Registered counts in nonoverlapping on/off windows are independent Poisson variables, before any deadtime/censoring issue.
Source expectation s>=0 is confined to on window; background stationary with common rate and known exposure ratio.
Known-background upper bound requires externally known fixed b_on, not a finite-baseline plug-in estimate.
Bonferroni alpha split is fixed before observing counts; total nominal alpha=.05, default equal split.
Limits cover expected detected source counts in the selected window, not emitted photons, current, peak rate, or source class.
Per-test alpha is not a search false-alarm probability. A family bound needs the actual number of evaluated tests; operational overlapping-window trial count is unspecified.
Rate times assumed 17 ns deadtime is only a diagnostic flag in the parent analysis. Flagging does not restore ideal Poisson counts; no Bernoulli-bin detector model is imposed here.
Applying limits only after data-dependent selection or nondetection can change conditional coverage; unconditional bounds do not automatically survive that conditioning.
Unknown backgrounds, drift, overdispersion, overlapping on/off windows, optical calibration and detector artefacts require separate models.
Sources for this question (6)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
DC coupling is feasible, including grounded-anode PMT architectures. For an analogue envelope, a short AC-coupling time constant causes droop and baseline errors. Counting short electrical single-photon pulses can preserve a millisecond arrival-rate envelope even when individual pulses are AC shaped; dead time, gain threshold and overload recovery then need explicit tests.
Physics 101
DC coupling lets a sustained input remain a sustained output. It can preserve both the fast start and the following milliseconds of a flash, provided the amplifier does not saturate or drift too much.
An AC-coupled analogue signal tends to forget a sustained level. The output jumps at the start, then slides back toward zero even while the light remains on. This is the droop that concerned you, and it is a real design issue.
Counting individual photons works differently. Each photon produces a very short electrical pulse. We can shape that pulse, count it, and ask how the number per time bin changes during the optical event. Shaping each tiny pulse does not necessarily erase the longer envelope made by many arrivals.
The important test is the complete measured response. A detector can appear quiet because it is overloaded, not because light is absent. We therefore need tests that switch between dark, bright and faint illumination and verify both amplitude and recovery.
Full technical answer
DC coupling is feasible and does not require a new physical detector principle. A PMT with negative cathode HV and its anode at ground can drive a DC-coupled analogue current/voltage chain. ET Enterprises explicitly describes this topology; the named HV 2520 CN base supports AC or DC output coupling. These are commercial architecture facts, not measured flight noise or external-vacuum qualification.
AC coupling has two distinct meanings. If a capacitor high-pass-filters the millisecond analogue optical envelope, a flat pulse droops and the falling edge produces a baseline excursion. For an ideal single-pole high-pass, less-than 5% droop over 3 ms requires tau>58.5 ms, or cutoff below 2.72 Hz. Real multi-stage responses require measurement. A short coupling constant can bias amplitude and an onset/threshold estimator even when the initial sharp edge remains.
In a photon-counting PMT architecture, each detected photon produces a nanosecond electrical pulse. That individual pulse may be AC shaped and discriminated; accumulating its occurrences in 10 microsecond bins can preserve a millisecond photon-arrival-rate envelope. The key limitations then are threshold/gain stability, baseline movement, pile-up, dead time and overload recovery, not the high-pass time constant applied to the optical envelope.
MMIA demonstrates such counting heritage but not unlimited linear rate. At bright-end rates a mean dead-time correction becomes uncertain and does not recreate lost arrival times or a Poisson process. Test faint-to-bright sequences, >3 ms plateaus and 30 microsecond edges through the actual chain at random sample phase and temperature. Any design needing simultaneous faint and very bright events may require parallel gains, attenuation or a dedicated analogue path.
Corrections to earlier assumptions
Envelope AC coupling and per-photon pulse shaping are different operations.
DC feasibility does not certify complete-chain noise, stability or vacuum compatibility.
Equations and their conditions
C9 · expression 1
plateau(t)/plateau(0) = exp(-t/tau_HP)
Ideal high-pass step-response droop.
Conditions: Single-pole analogue envelope model; not a photon-counter count-rate response.
C9 · expression 2
m = r/(1+r tau_d)
A nonparalyzable counter mean-response example.
Conditions: Stationary/quasistatic Poisson arrivals and known fixed dead time; not an exact noisy-waveform correction.
What supports this answer
manufacturer architecture
Grounded-anode negative-HV coupling and named AC/DC base are documented.
Full-chain optical impulse/step and>3 ms pulse response across faint/background/overload conditions
Droop, pedestal, saturation, recovery, discrimination/dead time and pulse-phase bias with actual ADC integration/filtering
Next inputs in the reviewed register
DC analog envelope versus AC-shaped individual-photon counting architecture and exact coupling schematic
Related executed studies (3) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N04 · Raw dead time can compress an apparent rise
executed conditional calculation not mission measurement
Re-ran 84 nonparalyzable quasistatic mean-response cases. For a linear ramp, raw normalized output becomes f(1+x)/(1+xf), x=S tau/(1+b tau). Its 10-90 rise divided by the true linear-ramp rise is(1+x)/(1+x+0.09 x^2).
This diagnostic concerns raw uncorrected expectation. It does not demonstrate bias in a published corrected MMIA waveform, does not restore individual photons and does not produce a valid Poisson likelihood after correction.
Calculation limitations
Only RAW uncorrectedmeanwaveformshape; publishedcorrected MMIA waveformscannotbeaccusedofthisbiasfromthisdiagnostic
R4-N08 · Optical/electronic design algebra with explicit assumptions
executed conditional calculation not mission measurement
Fresh design suite contains 36 etendue bounds, 372 hypothetical Gaussian source/filter cases, 900 sample-phase examples, 9 AC-droop rows, 58 OD allocations, 72 dark-only NEP scenarios and 192 browser-budget references. Forty-eight independent Gaussian quadrature checks agreed at about 4.4e-16 maximum absolute difference.
These are algebra and sampling diagnostics. The Gaussian source is not N2 line reconstruction; dark-only NEP omits sky, amplifier, correlations and radiation; ideal etendue is a necessary bound, not a ray-traced layout.
Calculation limitations
No flight filter optimum, no measured electronics noise, no hardware qualification and no source-population inference.
R4-N12 · Conditional transported pulse timing and electronics
executed conditional HG waveform diagnostic not depth classifier
Convolved each retained kernel with assumed intrinsic 10-90 rises 30, 80 or 200 us in an exponential-decay source family. Applied no electronics or an ideal first-order 10/50 kHz low-pass. These cutoff frequencies are not ADC sample rates or validated antialias designs.
The table shows only HG examples to avoid presenting unconverged GHM differences as physics. A source 2 km below the modeled top can still have a 30 us input broadened to about 48 us in one HG case. This is a conditional counterexample to a universal geometric-depth threshold, not evidence that an actual deep NBE will always look fast.
Both 10-90 rise and known-model-onset-to-peak are retained separately. Model curves have known onset and no photon noise. Near-window and bin-placement checks have limited scope; they do not validate observed onset, unknown intrinsic pulses or population classification.
Calculation limitations
HG is a comparison phase model, not an event-specific ice measurement. GHM timing remains numerically unconverged.
Assumed source decay 3 ms is not a universal observed duration. Output normalized to its own peak hides amplitude differences intentionally.
Known-model onset, 10-90 rise, arrival quantiles and noisy inferred onset are different quantities.
No detector dead time, correlated noise, uncertain baseline or sampling-phase likelihood included in this transport convolution.
Sources for this question (5)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
2-3 year 550 km polar/SAA radiation effects on dark current and gain.
Direct answer
A 2–3 year degradation coefficient cannot be assigned from altitude and detector type alone. Use the actual orbit, shielding, temperature, bias and detector lot, and separate ionizing dose, displacement damage, single-event effects and prompt particle artifacts. Existing flight and irradiation measurements constrain risks but do not establish a universal radiation-tolerance winner.
Physics 101
Radiation damage is not one thing. Some particles leave lasting defects that create false counts. Others make a momentary pulse or upset an electronic state. A detector can keep working electrically while becoming too noisy for a faint measurement.
Silicon detectors often develop more dark counts after energetic particles damage their crystal. Cooling can reduce some of those counts. Heating may repair some defects, but this is a tested procedure with limits, not a free recovery button. PMTs have different vulnerabilities rather than complete immunity.
A satellite at lower altitude, another inclination or a different place inside its spacecraft can receive a different radiation exposure. That is why an observed increase on one mission cannot simply be multiplied by three years for ours.
There is also a classification hazard: a particle can hit only the UV detector or make a window flash. The red channel can remain quiet. We must measure these signatures and use valid health, live-time and recovery checks, rather than assuming every blue-only pulse came from a thunderstorm.
Full technical answer
A reliable 2-3 year forecast at 550 km polar inclination requires the chosen orbit epoch, shielding CAD, temperatures, bias states and detector lot. Total ionizing dose, displacement damage, single-event upsets/latch-up, prompt charged-particle pulses and cumulative optical/gain aging are separate mechanisms. On/off operation may affect some bias-related damage but does not block penetrating displacement damage.
Measured heritage is informative but cannot supply a universal ranking. GRID-02 reports 93-110 microampere/year dark-current growth per 6 x 6 mm SiPM at 5 C in a 457 x 467 km orbit. Spoo Qy-1 follows specific SAP500 APDs and custom quenching for 600 days while altitude fell 418 to 378 km at 51.6 degrees. Its count growth, geometry and temperature are not a 550 km polar coefficient. CMOS radiation tests show hot/noisy pixels and random-telegraph tails; long-exposure test structures do not determine 20 microsecond full-camera noise.
Proton-tested APDs can retain timing/PDE while dark counts grow severely; cooling and annealing may help dark counts but can alter afterpulsing or exceed a safe thermal/power budget. In one passive-quench bench experiment, extreme activity caused over-saturation: registered counts fell while true activity increased. PMTs avoid the same silicon bulk-dark-current mechanism but still have HV electronics, window/cathode/MCP aging, magnetic response and prompt particle/light artifacts. Small PMT sample studies do not establish a class-wide radiation winner.
Prompt false events deserve their own budget. TUS observed fast track-like flashes in UV channels, with simulations implicating filter interactions and an anti-trigger reducing that class. Ice Cube measured correlated millisecond internal-light bursts in a very different glass/gel assembly. These examples do not predict this instrument's rate, but show why a UV-only event or millisecond duration is not a universal astrophysical veto. Test the actual head, filters and electronics for particle signatures, correlated noise, saturation and recovery.
Corrections to earlier assumptions
Operational lifetime is not preserved faint-photon sensitivity.
Prompt hits are distinct from cumulative radiation degradation.
Cooling, annealing and detector-class ranking remain device/circuit/mission specific.
Selected device/lot pre/post proton, TID and SEE tests at intended temperatures, including PDE/gain, noise PSD/DCR, correlated noise and annealing limits
Prompt particle/window/coating/electronic pulse shapes and false-event/veto response of the assembled head
Next inputs in the reviewed register
Orbit epoch/inclination/altitude history and shielding CAD/materials; thermal/power and bias/protection states
Related executed studies (0) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
No question-specific numerical study is assigned in the reviewed register.
Sources for this question (8)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
European suppliers for flight or near-flight detectors.
Direct answer
ET Enterprises and Photek are credible European starting points for heritage-based discussions, with different product strengths. Exact 337/777 nm response and qualification must be established for the delivered build. A European sales presence, a supplier’s unrelated qualification or one custom flight assembly does not certify another detector module.
Physics 101
A supplier having built space hardware is useful evidence. It does not mean every part in its catalogue is ready to fly.
We found two particularly relevant routes: ET Enterprises for the PMT approach closest to MMIA, and Photek for custom intensified detectors with independently documented space use. The next step is to specify exactly which tube, window, base, electronics and tests are required.
A hermetic package keeps its interior sealed. That says less about whether the attached circuit, cables, insulation and cooling work in external vacuum. Likewise, an ESA certificate can cover a narrow manufacturing process rather than the optical detector we want.
Country of origin needs equal care. A European office or brand does not show where the sensor wafer or assembled part was made. If European manufacture is a mission requirement, obtain a declaration for the selected part and lot. The report therefore keeps supplier leads separate from accepted flight hardware.
Full technical answer
ET Enterprises is a reasonable first supplier to evaluate for closest MMIA blue-photometer heritage; Photek is credible for custom MCP/intensified space assemblies. This is a shortlist to investigate, not a declaration that either offers a catalogue receiver qualified for this mission. Manufacturing country, corporate/sales location, delivered qualification and exact wavelength performance are separate procurement attributes.
ET9111B/WB bialkali and 9113B/WB S20 families have documented 22 mm active diameter in 25 mm-class outer tubes. Their M selection offers supplied spectral-response calibration; flying-lead and custom-selection options exist. Neither the peak QE nor a typical curve guarantees exact 337.1/777.4 nm PDE. MMIA's paper names the supplier but not the catalogue model, so these families must not be labeled the actual flight tube. Manufacturer reduced-pressure cap precautions do not qualify the complete HV base/module.
The independent UVIT flight paper identifies Photek UK custom intensified CMOS detectors, including an S20/silica VIS channel covering 320-550 nm, with assembled-payload thermal-vacuum/vibration tests and in-orbit images. This does not transfer its qualification to all current MCP-PMT configurations or establish exact 337 nm efficiency, 777 nm response or 50 kS/s operation.
First Sensor's inspected ESCC certificate 379 A covers ATH assembly/test processes for hermetic n-channel power MOSFETs; it is not APD optical-product approval. Its commercial red-sensitive APD arrays remain candidates requiring their own evidence. Laser Components' preliminary SUR 500 x UV APD is a small-area candidate, while the independent SAP 500 flight study concerns another family/custom circuit. The company identifies APD manufacturing in Arizona. Excelitas identifies a Canadian sensor facility; European sales access alone is not European manufacture. No complete off-the-shelf European-manufactured 337/777 receiver with the required full qualification was verified in the bounded search; this is not proof of market absence.
Corrections to earlier assumptions
Do not identify ET catalogue candidates as the unpublished MMIA model.
First Sensor MOSFET process approval is not APD certification.
Hermetic packaging, supplier heritage and European sales presence are not complete flight qualification or lot origin.
Equations and their conditions
No numerical formula is assigned to this supplier/procurement question in the reviewed source. The relevant checks are the delivered part, spectral calibration and accepted qualification scope.
What supports this answer
manufacturer options
ET exact families, calibrated selections and base/vacuum guidance are documented.
Part/lot-specific spectral calibration at 337/777 and accepted qualification/acceptance results for the actual detector, base, HV and readout build
Next inputs in the reviewed register
Manufacturer/wafer/assembly country-of-origin declaration if European manufacture is required; screening/traceability and complete external-vacuum/materials/HV qualification scope
Related executed studies (0) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
No question-specific numerical study is assigned in the reviewed register.
Sources for this question (11)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Maximise nonimaging photometer etendue within roughly 1U.
Direct answer
Design the entrance aperture, field, detector area and filter together using étendue conservation. Wide-angle light cannot be concentrated without limit onto a tiny high-efficiency detector. No collector type has yet been demonstrated best inside the proposed 1U volume; a packaged ray-traced design and measured field response are still needed.
Physics 101
The entrance aperture is the opening that collects light. The detector is the surface that turns that light into charge. They are not automatically the same size.
A wide field accepts rays from many directions. There is a limit to how much of that light passive optics can squeeze onto a small receiving surface. A funnel can concentrate a narrow set of directions strongly, or accept a broad set with less concentration; it cannot give both without paying elsewhere.
This is why a tiny detector with excellent photon efficiency may still be a poor match to a large, wide-field photometer. A larger detector array can help, but its internal noise and electronics also grow. The optical and detector choices must be made together.
The spacecraft volume matters as well. MMIA's small entrance and large photocathode show one successful arrangement, not proof that a 25 mm entrance with the same narrow filter fits into 1U. We have useful physical lower bounds, but still need an actual optical layout and test.
Full technical answer
Maximize useful accepted phase space, not aperture diameter alone. For a circular cone in air, G=A pi sin^2(theta_half). An ideal lossless concentrator into a flat detector accepting a hemisphere has A_detector >= A_entrance sin^2(theta_half). A 25 mm circular entrance therefore requires at least 122.7 mm^2 detector area for 60 degree full field or 245.4 mm^2 for 90 degrees. These are necessary bounds, not a realizable design or guarantees against aberration, vignetting and losses.
A small high-PDE SPAD or 0.5 mm APD cannot accept that entire 25 mm/wide-field input through air. A mosaic can provide more area, but brings summed dark/electronic/correlated noise, dead space and readout power. Immersion can change the refractive-index factor only with a specified physical coupling; inserting silicon's bulk index for an air-gapped package is invalid.
MMIA's custom relay combines a nominal 1 mm UV entrance with a 25 mm photocathode. Its 337/777 field is an 80 degree square diagonal; circular-cone bounds cannot be applied without integrating that field. The reviewed square-field phase-space integral is 0.9631 sr of projected solid angle. A 25 mm entrance with every ray at one flat filter constrained within 10 degrees requires at least 79.71 mm circular clear filter in the ideal air/lossless model. Space for lenses, baffles, detector, HV, thermal paths and electronics is additional.
A reflective collector avoids chromatic lens absorption but still faces obscuration, coating, field and concentration constraints. A nonimaging concentrator can be efficient within a defined angular acceptance; a bare detector plus baffle is simple but sacrifices aperture or field/stray rejection. None was proven best within 1U. A ray-traced packaged layout and measured acceptance are required. Enlarging aperture raises both source and diffuse-background counts; in a background-limited fixed-field limit, signal-to-noise grows with diameter, not area.
Corrections to earlier assumptions
The 25 mm MMIA photocathode is not its UV collecting aperture.
Conservation of etendue cannot be bypassed by nonimaging optics or high PDE.
Circular 60/90 degree and square 80 degree-diagonal fields require different integrals.
Equations and their conditions
D12 · expression 1
G = n^2 integral_A integral_Omega cos(theta) dOmega dA
Optical etendue is conserved by ideal passive lossless optics.
Conditions: Actual refractive coupling and field/pupil acceptance must be specified.
D12 · expression 2
A_det >= A_ent sin^2(theta_half) (air-to-air)
Ideal concentration into a detector hemisphere.
Conditions: Circular cone, flat projected areas; losses and package restrictions can only reduce acceptance.
Ray-traced optical phase space with actual stops, clear areas, refractive coupling, obscuration and detector active area; 1U mass/volume/thermal/power layout
Related executed studies (2) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N06 · Field size versus an unknown source-brightness law
executed conditional calculation not mission measurement
Re-ran 216 spherical-geometry sensitivity cases with an assumed cumulative source number N(>yield) proportional to yield^(-alpha). Alpha and normalization are not measured or fitted.
The wider field raises covered area but also background, off-axis range and projection losses. The 90/60 relative yield can change sign of preference as alpha changes. The weighted exposure area uses an arbitrary 1 e 20 photon pivot only for units; it is not an assumed event output or an actual effective mission exposure.
Bright-enough population WEIGHT proxy, not an orbital event rate or a full detection-probability integral.
R4-N08 · Optical/electronic design algebra with explicit assumptions
executed conditional calculation not mission measurement
Fresh design suite contains 36 etendue bounds, 372 hypothetical Gaussian source/filter cases, 900 sample-phase examples, 9 AC-droop rows, 58 OD allocations, 72 dark-only NEP scenarios and 192 browser-budget references. Forty-eight independent Gaussian quadrature checks agreed at about 4.4e-16 maximum absolute difference.
These are algebra and sampling diagnostics. The Gaussian source is not N2 line reconstruction; dark-only NEP omits sky, amplifier, correlations and radiation; ideal etendue is a necessary bound, not a ray-traced layout.
Calculation limitations
No flight filter optimum, no measured electronics noise, no hardware qualification and no source-population inference.
Sources for this question (3)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Maintain narrowband filter response over 60-90 degree full field.
Direct answer
A wide external field can coexist with a narrow filter only through a carefully designed relay and sufficiently controlled ray angles. Telecentric chief rays do not remove each point’s finite cone; graded coatings do not automatically correct every ray. Evaluate the assembled wavelength–field–pupil response before choosing a larger filter, smaller field or wider passband.
Physics 101
An interference filter selects color using layers of material. Light arriving at a slant sees those layers differently, so the transmitted color shifts.
A wide external field does not force the filter itself to see that same wide angle: lenses can rearrange the rays. But doing so uses optical area and volume. Even when the center ray is well controlled, each bundle still contains a spread of directions.
A graded coating is like tuning different parts of the filter differently. It can help if we know which rays go where, but it cannot make every direction identical. A wider passband is easier to tolerate geometrically, while collecting more unwanted photons.
The useful measurement is therefore a map: which colors pass at every field position, temperature and polarization through the assembled head? That map also affects the blue/red ratio. A source at the edge may otherwise appear fainter or redder simply because the instrument response changed.
Full technical answer
A thin-film interference filter generally shifts toward shorter wavelengths as incidence increases, with polarization splitting and a finite cone broadening the effective response. The common effective-index formula lambda(theta)=lambda 0 sqrt[1-sin^2(theta)/n_eff^2] is a useful limited-angle estimate. It is not a substitute for measured 337 nm coating data over field, pupil, temperature and polarization.
Telecentricity can make chief-ray angles similar across field, but does not remove the marginal-ray cone. Radially graded coatings can compensate a predictable spatial mean angle; they do not automatically undo the range of angles crossing each point. A larger filter/pupil can reduce the angular spread at a cost in size. Widening the band can retain more source light but also admits more background and possible spectral contaminants.
MMIA uses a relay and internal interference filter with telecentric illumination, followed by a square field stop and distribution optics. Its published CAD drawing is not an optical prescription: filter diameter, F-number and actual marginal-ray angles cannot be measured from an unscaled illustration. Fig 6 supplies one curve per channel without a full angle/temperature setup. Section 6.3 explicitly notes an angle correction C(theta) but does not provide its map.
For the proposed wide field, compute the entire wavelength-angle response, then fold source and background spectra. The appropriate design choice may be a larger relay, a smaller field, a wider filter, a larger receiving area or another architecture. A nominal 4 nm/80 degree heritage specification does not decide among them. The fresh design appendix's angle/width numbers are idealized sensitivity cases, not an accepted coating design.
Corrections to earlier assumptions
Telecentric does not mean every ray is normal to the filter.
Do not infer a PHOT cone or filter diameter from the camera F-number or unscaled CAD.
Filter transmission map versus wavelength, actual ray angle/cone, polarization, temperature and vacuum for the selected coating
Full-field assembled spectral response including pupil weighting and tolerances
Next inputs in the reviewed register
Optical prescription and clear filter diameter; actual ray distribution and any radial grading registration
Related executed studies (1) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N08 · Optical/electronic design algebra with explicit assumptions
executed conditional calculation not mission measurement
Fresh design suite contains 36 etendue bounds, 372 hypothetical Gaussian source/filter cases, 900 sample-phase examples, 9 AC-droop rows, 58 OD allocations, 72 dark-only NEP scenarios and 192 browser-budget references. Forty-eight independent Gaussian quadrature checks agreed at about 4.4e-16 maximum absolute difference.
These are algebra and sampling diagnostics. The Gaussian source is not N2 line reconstruction; dark-only NEP omits sky, amplifier, correlations and radiation; ideal etendue is a necessary bound, not a ray-traced layout.
Calculation limitations
No flight filter optimum, no measured electronics noise, no hardware qualification and no source-population inference.
Sources for this question (3)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Is 4 nm FWHM optimal for the N2(0, 0)band and background?
Direct answer
Treat 4 nm as a heritage starting point, not a proven optimum. Nitrogen emission is a structured molecular band. Widening helps if useful source photons were being lost, but increases background after the source is already captured. The optimization still needs a representative source spectrum and measured angular/thermal filter response, in one air/vacuum wavelength convention.
Physics 101
The 337 nm emission is a group of closely spaced colors produced by the same nitrogen transition system. A filter catches some fraction of that group.
Making the filter narrower can reject background, but it can also reject the signal we want. Making it wider can catch more signal until almost all useful light is included; after that, extra width mainly adds unwanted photons. The best choice depends on both curves together.
The shape matters as well as the quoted width. A rounded Gaussian filter and a rectangular 4 nm filter do not pass exactly the same amount of a smooth background. The angle-dependent color shift can matter as much as a small change in nominal width.
We should not optimize hardware using a line spectrum whose positions or populations have not been checked. The new report keeps synthetic spectral examples separate from verified molecular data and leaves the final band choice conditional on source and assembled-filter measurements.
Full technical answer
Four nanometres is a reasonable heritage starting point, not an established SNR optimum. N2 second-positive(0, 0) emission is a rovibronic band with structured line positions and populations, not a delta function at a single label 337.1 nm. Capture depends on population/temperature assumptions, band-head registration, filter shape and angular/thermal changes. More emitted wavelengths are useful only when the background and competing emissions are also included.
The campaign narrowed public primary line-list/database routes and found material coordinate/population issues. Conditional line reconstructions using lower-state energy/J assumptions were explored but do not establish an NBE rotational/vibrational population or a flight optimum. This revision does not retain an earlier unvalidated molecular-line reconstruction as a validated spectrum. A Gaussian synthetic source in the fresh design algebra is explicitly a hypothetical shape, not a reconstructed N2 band.
For flat background, its variance often grows approximately with equivalent filter width. If all useful source photons are already captured, further widening degrades background-limited SNR approximately as 1/sqrt(width). If a narrow filter loses source photons, widening can instead help. The optimum maximizes the actual detection/timing objective, including filter transmission, PDE, scene spectrum and false alarms. A 4 nm-FWHM Gaussian of unit peak has equivalent width about 4.26 nm, not 4 nm.
Use one air/vacuum convention for source lines, solar/stellar spectra and filter measurements. Around 337 nm the coordinate difference is about 0.1 nm; near 777 nm about 0.21 nm. This coordinate conversion is distinct from a physical coating shift caused by vacuum or temperature. The Noll table wavelength convention is not fully verified at the final processed level, so its narrow folds retain that qualification.
Corrections to earlier assumptions
A nominal band-head wavelength is not a complete emission spectrum.
FWHM is not equivalent noise bandwidth.
Unvalidated line-list reconstruction and assumed NBE temperatures cannot select the flight filter.
Equations and their conditions
D14 · expression 1
S(w)=integral Phi_source,lambda T_w(lambda) eta(lambda) d lambda
Detected signal depends on spectral capture, not width alone.
Conditions: Also integrate field, pupil and temperature where necessary.
D14 · expression 2
Delta_lambda_eq = integral T(lambda) d lambda / T_peak
Representative source-band spectra or validated physical spectral model constrained by source conditions
Scene/background spectra and actual assembled filter/detector response with stable wavelength registration
Next inputs in the reviewed register
Source air/vacuum convention and population/pressure/temperature assumptions; FWHM versus equivalent noise bandwidth; defined detection objective
Related executed studies (2) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N07 · Actual reference continuum and starlight data folds
executed conditional calculation not mission measurement
Re-ran Noll processed-mean spectral folds and Bianchi coarse-spectral sky-cube hemisphere examples. Source data are real/model-published inputs, but the target geometry and ideal filters are conditional.
The Noll extraction removes strong lines and modeled celestial components. Bianchi covers stellar components, not Moon, zodiacal or airglow, and its spectralgrid near 337 nm is much coarser than 4 nm. Neither is a total polar scene distribution.
Calculation limitations
Mean 10633 selected 2009-2019 Cerro Paranal spectra, >=10 min exposure; Moonabovehorizonwithillumination>50% excluded. Not a polar or microsecond distribution.
Continuum extracted by median/percentile filters; strong resolved lines removed, unresolved O2 UV bands retained partly. Extraterrestrial components modeled/subtracted; extinction and van Rhijn 90 km corrected. Not total night sky or total 777 line background.
Header only says wavelengthnm. Official ESO says Xshooter pipeline wavelength in air; authors could have subsequently converted, not explicitly verified here. Fold centre 337.1 uses table convention; photonenergyuses nominal wavelength with <0.03% air/vac sensitivity.
4 component NMF is statistical; labelledcomponents must not automatically be treated as puremolecular emissions.
Totalmeans direct+interstellar-dust-scattered starlight; no zodiacal, moon, airglow. Opticalphotometry+stellar-atmosphere/exctinction extrapolation, not a narrow 337 measurement.
99 logbins span~0.1-4.3 um: around 337 bracket~13 nm wide. Cannot infer actual 4 nm spectralfeatures or own covariance from tabulated broad-model errors.
Hypotheticalsix cloudnormaldirections, noactualorbit/date/Sun/moon/cloudprofile. Not globalminimum/maximum or occurrenceweighteddistribution.
Earthfilled nadir does not see this external field directly; need hemisphericillumination and cloud BRDF/atmosphericscattering. Lambertalbedo.8 is an explicit example only.
Cubeis HEALPix NESTED order 6 in celestial RA/Dec despite paper figures plottedin Galacticcoordinates. Do not reinterpret cube in Galacticcoordinates.
R4-N08 · Optical/electronic design algebra with explicit assumptions
executed conditional calculation not mission measurement
Fresh design suite contains 36 etendue bounds, 372 hypothetical Gaussian source/filter cases, 900 sample-phase examples, 9 AC-droop rows, 58 OD allocations, 72 dark-only NEP scenarios and 192 browser-budget references. Forty-eight independent Gaussian quadrature checks agreed at about 4.4e-16 maximum absolute difference.
These are algebra and sampling diagnostics. The Gaussian source is not N2 line reconstruction; dark-only NEP omits sky, amplifier, correlations and radiation; ideal etendue is a necessary bound, not a ray-traced layout.
Calculation limitations
No flight filter optimum, no measured electronics noise, no hardware qualification and no source-population inference.
Sources for this question (7)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Required optical density and rejection wavelength range.
Direct answer
Derive blocking from an allowed detected leakage rate across the detector’s entire sensitive range. A single OD number without its wavelength interval, angle and measurement floor is insufficient. The conditional OD algebra is available; the mission requirement still needs scene spectra and qualified full-range measurements of the assembled filter stack.
Physics 101
A narrow filter is useful only if it also blocks unwanted colors strongly enough. The detector may respond to a much broader range than the nominal science band.
Optical density is a logarithmic blocking measure. OD 6 means one part in a million transmission, but only where that value is actually guaranteed. A guarantee near 337 nm says nothing by itself about a leak near 800 nm.
A tiny leak can matter when the unwanted light is much brighter or occupies many more wavelengths. We therefore compare the total leaked photon rate with the useful band and its noise allowance. The correct answer is a blocking curve plus a wavelength interval, not one impressive number.
A laboratory scan also needs enough dynamic range to verify the claim. Seeing no leak on an insensitive measurement does not prove the leak is absent. The assembled head must be tested because light can scatter around the intended filter path or be generated inside materials.
Full technical answer
Specify a source-weighted leakage allocation rather than a single unqualified OD number. Compute in-band and out-of-band detected photon rates using the full scene spectrum, filter transmission and detector response. A detector still sensitive in the visible/NIR can receive substantial leakage from red nightglow, moonlight, twilight or bright lightning even when the 337 nm band is narrow.
The inspected Alluxa 337 nm example gives OD 6 over 291-331 nm and 343-380 nm on its listed specification. That does not demonstrate OD 6 from the UV through silicon's near-IR cutoff. A data-sheet plot with limited axes also cannot supply a guaranteed floor below the plot dynamic range. Real blocking depends on angle, polarization, temperature, manufacturing defects, pinholes and aging.
If a uniform out-of-band transmission t_leak=10^-OD is an explicit assumption and C_out is the unblocked out-of-band detected rate, requiring leakage below fraction f of in-band rate C_in gives OD >= log10[C_out/(f C_in)]. This algebra becomes a useful numerical requirement only after C_out, C_in and the guaranteed wavelength range are known. Spectrally structured leaks require integration rather than one effective OD.
Test the combined filter stack with sufficient dynamic range to support the required rejection, including fluorescence/scintillation and scattered-light paths. A blue-only artifact generated within a window or detector can bypass an external spectral rejection argument. The fresh OD scenarios show conditional allocations, not supplier acceptance or complete mission spectra.
Corrections to earlier assumptions
Limited-interval OD 6 is not broadband OD 6.
A display plot or unverified average blocking claim is not an acceptance measurement.
Equations and their conditions
D15 · expression 1
C_leak = integral_out L_lambda A_eff,unblocked T_leak lambda/(h c) d lambda dOmega
Leakage is a detector-weighted integral.
Conditions: Scene spectrum and angular/coating response must be specified.
D15 · expression 2
OD >= log10[C_out/(f C_in)]
Uniform-leak allocation example.
Conditions: Constant leakage over a stated interval; required fraction f is a design choice.
What supports this answer
manufacturer specification
Exact filter guaranteed intervals limit a quoted OD.
Dynamic-range-qualified leak/blocking scan through detector cutoff, including pinholes, angle/polarization/thermal changes and induced fluorescence
Scene spectra sufficient to bound out-of-band photoelectrons and stray light under permitted night/Moon/twilight conditions
Next inputs in the reviewed register
Allowed leakage fraction/noise allocation and guaranteed wavelength interval for each coating/blocker
Related executed studies (1) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N08 · Optical/electronic design algebra with explicit assumptions
executed conditional calculation not mission measurement
Fresh design suite contains 36 etendue bounds, 372 hypothetical Gaussian source/filter cases, 900 sample-phase examples, 9 AC-droop rows, 58 OD allocations, 72 dark-only NEP scenarios and 192 browser-budget references. Forty-eight independent Gaussian quadrature checks agreed at about 4.4e-16 maximum absolute difference.
These are algebra and sampling diagnostics. The Gaussian source is not N2 line reconstruction; dark-only NEP omits sky, amplifier, correlations and radiation; ideal etendue is a necessary bound, not a ray-traced layout.
Calculation limitations
No flight filter optimum, no measured electronics noise, no hardware qualification and no source-population inference.
Sources for this question (6)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
777.4 nm null-channel detector and comparable photon thresholds.
Direct answer
A silicon APD is plausible for 777.4 nm, but it is not the only option. Compare both channels through detection probability versus incident aperture photons, including field, spectrum, time window, noise and recovery. A red nondetection is an upper limit rather than proof of zero emission. A universal minimum red/blue ratio for every thermal alternative remains unestablished.
Physics 101
The red channel is a test of a competing explanation. To trust a negative result, we must first show that the test was capable of seeing the red light in question.
A quiet detector does not mean zero emitted light. It means the data put an upper limit on how much light could have been present. Even a perfectly working detector can miss photons by chance, and a noisy or insensitive channel has a weaker limit.
There are further ways to get a misleading null. The red pulse may arrive later, the source may lie where red throughput is poor, or the detector may be overloaded and failing to count. A particle can also produce only a blue-channel artifact.
The fair blue/red comparison therefore uses calibrated photons at the entrance and measured probabilities of detection. It also needs physical examples of ordinary lightning and target events. Better red sensitivity reduces one source of mistakes, but does not create a universal law saying all thermal discharges must have a particular minimum red ratio.
Full technical answer
A silicon APD is a plausible red-channel option, not a forced solution. GaAs photocathode PMTs can outperform ordinary multialkali QE near 800 nm, though receiving area, cooling power, dark noise and qualification matter. An inspected H 16722-50 specification gives 15% typical QE at 800 nm with a 5 mm cathode; this is not exact 777.4 nm PDE or a complete low-power system. A tiny high-PDE APD still faces the wide-field etendue requirement.
Comparable sensitivity must mean comparable detection probability versus incident aperture photons for specified source spectra, time windows, field positions and backgrounds. Equal output voltage, nominal QE, NEP or event counts does not ensure that. MMIA's nominal red aperture is 2.3 mm versus 1 mm in UV, but its Table 5 total efficiencies 1.6% versus 13.6% give red/UV A times PDE about 0.622, not automatically equal thresholds. Other aggregate calibration values differ and should not be silently substituted.
A red non-detection must be an upper confidence limit with calibrated live time, bias, noise, saturation, recovery and acceptance. Zero observed counts with exactly known zero background gives an ideal 95% mean-signal upper limit 2.996 counts; actual background, dead time and nuisance parameters change the procedure. An overloaded passive APD can register fewer counts as true activity grows. A narrow simultaneous window can also miss delayed thermal emission; Leal et al.(2026) selected direct-view return strokes have 22.7-58.3 microsecond red-minus-blue peak offsets, not proof of red absence at blue onset or for a 3 ms window.
Most importantly, no universal positive 777/337 lower bound for every weak thermal alternative was established. Blue-rich components can occur within ordinary lightning, and a UV-only particle artifact can pass a red-null rule. The criterion can be valuable but needs independently labeled controls and joint sensitivity/false-classification tests; it is not unique proof that no thermal channel formed.
Corrections to earlier assumptions
Red non-detection is a confidence statement, not zero emission.
Red sensitivity must be compared in photon terms through the whole instrument.
A blue-first peak does not by itself demonstrate full-event red absence.
Equations and their conditions
E16 · expression 1
P(N=0|S)=exp(-S); S_95=-ln(0.05)=2.996
Ideal zero-count upper limit.
Conditions: Known zero background, unit-valid live time and independent Poisson counts; not a general ratio confidence procedure.
E16 · expression 2
epsilon_false_null = P(red accepted as null | physical alternative,full chain)
Measure the relevant classification error directly.
Conditions: Include spectrum, delay, field, background, health and selection.
What supports this answer
manufacturer option
GaAs PMTs are an alternative to low-red-QE multialkali and APDs.
Both full-chain photon-scale spectral/field response, background/noise, live time and impulse response, including dead time, saturation and recovery
Red-channel upper-limit coverage and joint false-negative/false-classification performance using calibrated injections and representative source controls
Next inputs in the reviewed register
Physically justified red search window, health/bias criteria, confidence rule and source-family spectral/delay assumptions
Related executed studies (3) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N04 · Raw dead time can compress an apparent rise
executed conditional calculation not mission measurement
Re-ran 84 nonparalyzable quasistatic mean-response cases. For a linear ramp, raw normalized output becomes f(1+x)/(1+xf), x=S tau/(1+b tau). Its 10-90 rise divided by the true linear-ramp rise is(1+x)/(1+x+0.09 x^2).
This diagnostic concerns raw uncorrected expectation. It does not demonstrate bias in a published corrected MMIA waveform, does not restore individual photons and does not produce a valid Poisson likelihood after correction.
Calculation limitations
Only RAW uncorrectedmeanwaveformshape; publishedcorrected MMIA waveformscannotbeaccusedofthisbiasfromthisdiagnostic
R4-N09 · Detection is easier than trusted morphology classification
executed conditional calculation not mission measurement
Executed 81,000 synthetic signal trials across 81 scenarios: receiver-input 10-90 rises 30, 80, 200 us; plateau-equivalent 3 ms scales 100, 1000, 10000 PE; background 50, 10000, 1e6/s; sampling 10, 50, 100 kS/s. Amplitude, onset and background are fitted using a finite 10 ms prewindow and explicit gates.
A predefined profile-log-likelihood margin ln 19 selects accepted classifications. It is not automatically a calibrated 95% posterior probability or a 95% correctness guarantee. Rejected events, wrong accepted labels and interval uncertainty are all reported.
The plateau-equivalent scale is rate times 3 ms. Actual source totals are smaller because the linear ramp uses part of the finite 3 ms pulse. These are receiver-input morphology trials, not source-depth/class validation. Count-loss occupancy flags are diagnostics only; the generator remains ideal independent Poisson.
Independent profile-objective comparisons and exact seeded reruns passed. The known-template oracle has extra information and is an optimistic benchmark, not an achievable estimator promise.
Calculation limitations
This is a conditional synthetic receiver-morphology study, not a flight sensitivity/completeness validation.
The classes mean 30/80/200 us 10-to 90% rises at a hypothetical linear receiver input after transport; they do not label source depth.
Source brightness, cloud transfer, line spectrum, detector PDE and source occurrence population are outside this count-level calculation.
The practical code has an external coarse onset bracket and a specified event gate. Blind stream triggering and a mission false-alarm rate are unvalidated.
A fixed ln 19 profile-score margin is not calibrated 95% confidence. Always show accepted fraction, rejected fraction and the error interval together.
The 3-class fast-subtype distinction is harder than the grouped fast-vs-slow test; these are separate decisions.
Only exact model-matched templates and constant backgrounds are generated. Real independently labeled optical/radio controls are still needed.
Post-counting electronic filtering, detector excess noise, afterpulses, saturation and deadtime are not simulated.17 ns occupancy is an illustrative hardware flag, not Poisson restoration.
Known-background red limits are invalid if a noisy finite baseline is silently substituted. The provided alternative uses a conservative Bonferroni construction.
Red limits constrain expected registered counts in a specified window, not thermal-source absence. Selection, red-window timing, throughput and population ratios remain physical inputs.
A useful mission event rate is NOT COMPUTABLE without an absolute received-fluence distribution, completeness and exposure definition.
No empirical wide-FOV event rate is inferred from the conditional geometric/population proxy; source geography and coverage remain unknown.
R4-N10 · Exact counts, finite-baseline limits and search control
executed conditional calculation not mission measurement
Executed 63 deterministic count-inference controls and separate coverage/null Monte Carlo diagnostics with 200000 trials per configuration. The mathematical model is independent Poisson counts with stationary background and known on/off exposure ratio.
Known-zero-background zero counts gives the 95% upper mean 2.9957. The specified conservative finite-baseline construction with zero on and zero off counts gives 3.6889 under equal alpha splitting. This difference is a procedure/model change, not conflicting physics.
Individual discrete confidence procedures may be conservative. The retained initial individual 99.9% Monte Carlo discrepancy was not hidden or rerolled; the declared simultaneous 99% family audit overlaps the analytic coverage enclosures. Monte Carlo intervals diagnose implementation; coverage follows from the stated mathematical construction.
Unconditional coverage need not remain 95% after selecting only non-detections or accepted classes. Per-test alpha is not continuous-search false alarm. Dead time, overdispersion, unknown calibration and varying background require more models.
Calculation limitations
Registered counts in nonoverlapping on/off windows are independent Poisson variables, before any deadtime/censoring issue.
Source expectation s>=0 is confined to on window; background stationary with common rate and known exposure ratio.
Known-background upper bound requires externally known fixed b_on, not a finite-baseline plug-in estimate.
Bonferroni alpha split is fixed before observing counts; total nominal alpha=.05, default equal split.
Limits cover expected detected source counts in the selected window, not emitted photons, current, peak rate, or source class.
Per-test alpha is not a search false-alarm probability. A family bound needs the actual number of evaluated tests; operational overlapping-window trial count is unspecified.
Rate times assumed 17 ns deadtime is only a diagnostic flag in the parent analysis. Flagging does not restore ideal Poisson counts; no Bernoulli-bin detector model is imposed here.
Applying limits only after data-dependent selection or nondetection can change conditional coverage; unconditional bounds do not automatically survive that conditioning.
Unknown backgrounds, drift, overdispersion, overlapping on/off windows, optical calibration and detector artefacts require separate models.
Sources for this question (7)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Cross-calibration of different detector technologies for a trustworthy ratio.
Direct answer
Use a common traceable photon-flux reference plane and calibrate both complete optical chains across wavelength, field, temperature and intensity. Separately measure multiplication gain, photon efficiency, timing and nonlinearity, then propagate shared uncertainty and covariance. Equal electrical gain or nominal QE does not make a trustworthy photon ratio.
Physics 101
A ratio can look stable even when both channels are wrong in the same way, or change because the channels drift differently. Calibration connects each electrical result to a known number of incoming photons.
We should illuminate both channels with known colors and pulses, through the real entrance path. We then check how much response changes with field position, temperature and brightness. An independent light monitor tells us whether the test source changed while we were measuring.
Timing matters too. If the blue and red electronics respond at different speeds, the ratio at one instant can be wrong even when their integrated light measurements are correct. The instrument must measure those relative delays.
Using the same reference can reduce some uncertainty, but not all. A reference star's uncertainty at one wavelength is not automatically its uncertainty at another, and a wide nonimaging field changes its integrated background when the spacecraft slews. The calibration plan therefore needs both photon counting and geometry.
Full technical answer
Cross-calibrate the channels at a common physical reference plane using traceable spectral photon flux. Measure the complete spectral, field, pupil, polarization, temperature and intensity response, not just electrical gain. Use overlapping calibrated attenuation ranges and an independent monitor so that changing source intensity is not mistaken for detector gain or efficiency. The actual science-band filter and entrance optics must be in the path for end-to-end response.
Different technologies need different internal diagnostics: PMT charge/gain distributions, counting discriminator efficiency, APD multiplication/excess noise and SiPM correlated avalanches are not equivalent. Separate gain calibration from absolute incident-photon efficiency. A common pulse source can reveal relative delay, impulse response, dead time and recovery, which matter when an instantaneous 337/777 ratio is formed from different waveforms.
Propagate shared uncertainties and covariance. A common source calibration can cancel part of a ratio uncertainty, but wavelength-specific reference errors, field mismatch and spectral assumptions may not cancel. For a nonimaging instrument with unknown source angle, use an independently localized source or a conservative field-dependent response envelope. A central on-axis calibration cannot certify every location.
CALSPEC provides useful stellar spectra covering both target bands. Its normalization accuracy near 555.75 nm vacuum must not be treated as identical absolute accuracy at 337 nm. On-orbit stellar transfer also needs a view of the star: nominal nadir pointing in a 60-90 degree field at 550 km is Earth-filled. Maneuver constraints, scene backgrounds and response maps are part of the calibration design.
Corrections to earlier assumptions
Equal electrical gains are not equal photon response.
Single-photoelectron gain calibration does not certify absolute PDE.
Shared calibration uncertainty may cancel partly; wavelength/field-specific uncertainty remains.
Conditions: Counts must be background/live-time corrected with valid likelihood; integrate actual spectra and time responses. Near a red null use an upper limit, not division by zero.
Traceable common-source calibration across each channel spectrum, field, pulse fluence and operating temperature with overlapping attenuation ranges
Relative delay/IRF and systematic covariance, plus repeated stability checks and independent reference monitoring
Next inputs in the reviewed register
Photon reference planes and air/vacuum coordinate conventions; ratio definition for unequal time responses and selection rules
Related executed studies (1) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N10 · Exact counts, finite-baseline limits and search control
executed conditional calculation not mission measurement
Executed 63 deterministic count-inference controls and separate coverage/null Monte Carlo diagnostics with 200000 trials per configuration. The mathematical model is independent Poisson counts with stationary background and known on/off exposure ratio.
Known-zero-background zero counts gives the 95% upper mean 2.9957. The specified conservative finite-baseline construction with zero on and zero off counts gives 3.6889 under equal alpha splitting. This difference is a procedure/model change, not conflicting physics.
Individual discrete confidence procedures may be conservative. The retained initial individual 99.9% Monte Carlo discrepancy was not hidden or rerolled; the declared simultaneous 99% family audit overlaps the analytic coverage enclosures. Monte Carlo intervals diagnose implementation; coverage follows from the stated mathematical construction.
Unconditional coverage need not remain 95% after selecting only non-detections or accepted classes. Per-test alpha is not continuous-search false alarm. Dead time, overdispersion, unknown calibration and varying background require more models.
Calculation limitations
Registered counts in nonoverlapping on/off windows are independent Poisson variables, before any deadtime/censoring issue.
Source expectation s>=0 is confined to on window; background stationary with common rate and known exposure ratio.
Known-background upper bound requires externally known fixed b_on, not a finite-baseline plug-in estimate.
Bonferroni alpha split is fixed before observing counts; total nominal alpha=.05, default equal split.
Limits cover expected detected source counts in the selected window, not emitted photons, current, peak rate, or source class.
Per-test alpha is not a search false-alarm probability. A family bound needs the actual number of evaluated tests; operational overlapping-window trial count is unspecified.
Rate times assumed 17 ns deadtime is only a diagnostic flag in the parent analysis. Flagging does not restore ideal Poisson counts; no Bernoulli-bin detector model is imposed here.
Applying limits only after data-dependent selection or nondetection can change conditional coverage; unconditional bounds do not automatically survive that conditioning.
Unknown backgrounds, drift, overdispersion, overlapping on/off windows, optical calibration and detector artefacts require separate models.
Sources for this question (5)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
On-orbit gain drift, absolute radiometric stability and peak brightness versus current.
Direct answer
Combine monitored optical references, full-path external calibration where possible and electrical injection. One internal LED reading confounds source brightness, photon efficiency and multiplication gain; a bypass path cannot measure entrance-optics aging. Absolute optical calibration is also insufficient by itself to turn brightness into discharge current, which needs an independently constrained source/transport model.
Physics 101
Calibration asks whether the same number of incoming photons still gives the same reported answer years later. The detector, electronics, filter and entrance optics can all change.
An internal lamp is valuable, but it can also age. If its reading drops, we need to distinguish a dimmer lamp from a less efficient detector or lower gain. A monitor helps, while its own stability must be checked. A light path that skips the front optics cannot tell us whether those optics became dirty.
Individual photoelectron pulses can reveal multiplication gain under suitable tests. That is different from asking what fraction of incoming photons produces a pulse. We need both quantities to trust absolute brightness.
Even perfectly calibrated brightness does not automatically tell us discharge current. Light is the result of how particles were excited and how the cloud redirected it. Different source sizes, fields and histories can produce similar observed brightness. The route to current is therefore a tested physical model tied to independent observations, not a universal conversion factor.
Full technical answer
Use several complementary calibration observables. An internal monitored LED can track repeatability and some gain changes; a full-path external reference can constrain entrance optics/filter/PDE drift; electrical injection checks readout gain, timing and offset but not photon efficiency. Reference-source and monitor drift themselves need redundancy or independent constraints. A reference bypassing the entrance optics cannot measure their contamination or transmission loss.
A single mean LED reading measures the product of LED photons, detection efficiency and multiplication gain. Under a validated Poisson photoelectron model, externally triggered illuminated and blank charge distributions, including no-signal pulses, can identify multiplication gain and variance separately from mean intensity. They still identify occupancy=photons times efficiency, not its two factors. Source intensity fluctuations, SiPM correlations, afterpulses, saturation and threshold truncation require expanded models.
Flight examples illustrate architecture rather than a Cube Sat accuracy guarantee. ISUAL's mainly-red internal LED response and HV corrections do not directly measure 337 nm PDE aging. OMPS working/reference diffuser schedules support long-term response tracking over a defined UV range, but their accuracy does not transfer to this design. CALSPEC stars offer another route if spacecraft attitude and full-path access permit it; on/off slews of a wide field change Galactic/stellar and terrestrial backgrounds. A favorable Poisson star-count budget does not prove percent-level absolute calibration.
Finally, absolute brightness is not a unique current meter. Photon production depends on electron energy distribution, reduced field E/N, excitation volume, quenching and source history; cloud transport and view geometry then alter the observed peak. Current, current moment and radio-inferred source parameters are different quantities. AIRFLY's measured 5.61 photons/MeV high-energy fluorescence yield is not a streamer photons-per-ampere calibration. A current relation requires a kinetic/radio/transport forward model and independent measurements.
Corrections to earlier assumptions
Internal LED mean response is a product, not a unique gain/PDE measurement.
Published visible or reference-wavelength stability is not automatically 337 nm absolute stability.
Absolute photons do not alone identify discharge current.
Equations and their conditions
F18 · expression 1
E[Q_light]-E[Q_blank] = Phi epsilon q_bar
Mean charge combines incident photons, photoelectron efficiency and mean multiplication charge.
Conditions: Linear independent-photoelectron model; source fluctuations and backgrounds require treatment.
F18 · expression 2
Var(Q_light)-Var(Q_blank) = mu q_bar^2 F_q; mu=Phi epsilon
Charge fluctuations and zero occupancy can help estimate gain.
Conditions: Acquire all triggered pulses; no unknown truncation or unmodeled correlations.
F18 · expression 3
brightness = transport{ excitation(E/N, n_e, N, volume, time, quenching) }
A source-current inference is an inverse problem with additional physical parameters.
Conditions: Independent validation required; current and current moment must be distinguished.
What supports this answer
calibration identifiability method
Charge-distribution and occupancy methods separate some gains while retaining photon-efficiency ambiguity.
Independent optical reference/monitor stability and full-path throughput/PDE drift at science wavelengths, with separate multiplication-gain and offset diagnostics
If using stars, actual attitude/path/field response and differential-background covariance; if inferring current, independent current/radio-optical/source-condition calibration
Next inputs in the reviewed register
Observable calibration paths, schedules, celestial visibility, full occupancy/blank acquisition and source-current/transport forward model
Related executed studies (0) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
No question-specific numerical study is assigned in the reviewed register.
Sources for this question (9)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Ground calibration/characterisation and facilities.
Direct answer
Before flight, characterize photon response, spectrum and blocking, field/pupil maps, noise statistics, timing, linearity, saturation and recovery, then repeat relevant calibration after environmental tests. Documented NPL/PTB/NIST/ESA capabilities are leads rather than reserved access or an achieved calibration. This study claims no institute affiliation or facility access.
Physics 101
A useful ground test asks a known question: we send a measured amount of light through the real entrance and check the reported answer. We repeat it for different colors, directions, temperatures and brightnesses.
Testing only the middle of the field leaves an important hole. This instrument does not form an image, so a flash may arrive from an unknown direction. We need to know how much the response changes across the whole field, and whether the two channels change together.
We also need to test mistakes: light too bright for the detector, a flash during recovery, false bursts in darkness, and a faint pulse on a bright background. Environmental tests should be followed by calibration to see whether sensitivity changed.
Specialist laboratories can provide traceable standards and relevant facilities. Their existence is not a promise that a particular service, uncertainty or booking is available. This report identifies routes and required measurements; it does not claim an institute of its own.
Full technical answer
Ground characterization should establish the delivered receiver's photon-scale response and uncertainty, not merely confirm that it produces an output. Measure spectral response across both passbands and the full detector-sensitive rejection range; field/pupil/polarization maps; temperature dependence; absolute throughput/PDE; linearity, dynamic range and saturation/recovery. Include cross-channel timing, clock accuracy, analogue/ADC impulse response and trigger selection.
Noise measurements must use the actual detector, bias, cabling, HV enclosure and estimator. Acquire dark distributions/PSD, afterpulses/correlated avalanches, rare bursts and particle/bright-light artifacts, not only RMS. Injection tests should span faint leading edges, strong backgrounds, and pulses exceeding 3 ms, with onset randomized against the clock. Calibrate attenuation and retain no-signal triggers where required by the statistical estimator.
Qualification adds vibration/shock, thermal-vacuum, HV behavior, materials/outgassing, contamination, daylight/sun protection, radiation and recovery tests for the selected assembly. Repeating calibration before and after these tests separates survival from maintained science performance. A reference detector or power standard must have a traceability chain and uncertainties appropriate to 337/777 nm and the receiving geometry.
NPL, PTB, NIST and ESA optical facilities provide documented relevant capabilities, but spectral coverage alone is not an accepted service for this exact large-field, pulsed receiver. The campaign establishes no institute affiliation or facility access on behalf of this assistant. Obtain an agreed scope, uncertainty budget, reference planes and acceptance criteria from the chosen laboratory before treating a calibration as available or achieved.
Corrections to earlier assumptions
Facility capability is not an agreed calibration uncertainty or booking.
Surviving environmental tests is not the same as preserving calibrated sensitivity.
Equations and their conditions
F19 · expression 1
measurement uncertainty includes reference,geometry,spectrum,temperature,repeatability,linearity and drift
Repeatability alone is not the absolute uncertainty.
Conditions: Retain covariance for the two-channel ratio and selection effects for faint pulses.
What supports this answer
documented facility capability
Relevant traceable calibration and optical test routes exist with configuration-specific limits.
Configuration-specific traceable spectral/angular/pulsed receiver calibration with stated uncertainty
Noise/linearity/trigger/timing/bright-light recovery, environmental qualification and contamination/radiation tests for selected build
Next inputs in the reviewed register
Facility-agreed spectral range, aperture/FOV, vacuum/pulse capabilities, access and accepted test plan
Related executed studies (0) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
No question-specific numerical study is assigned in the reviewed register.
Sources for this question (6)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Laboratory 337 nm source with 30 us rise and>3 ms duration.
Direct answer
Use a near-UV LED and a sufficiently fast current driver as the practical programmable testbed candidate, with optical timing measured by an independent fast reference. A 340 nm LED is not a monochromatic 337.1 nm source, and the previously cited slow LED driver cannot certify a 30 µs rise. A nitrogen laser matches the wavelength better but does not naturally reproduce the required millisecond envelope.
Physics 101
The easiest programmable source is probably a UV LED driven by an adjustable current pulse. But the light pulse must be measured directly: an electrical command is not proof that the LED followed it perfectly.
A 340 nm LED emits a spread of colors around its label. Some of its light passes a 337 nm filter and some does not. That is acceptable for certain tests if we measure the spectrum and calibrate how many useful photons entered the instrument.
A 337 nm nitrogen laser matches the color more closely, but its normal short pulses do not automatically match the desired long optical waveform. Color, duration, rise time and stability are separate requirements.
A bright reference photodiode can watch the test source while calibrated attenuation makes the instrument input very faint. We should test many pulse shapes and brightnesses, including overload and recovery. This gives a real measurement of where the instrument detects, classifies or fails.
Full technical answer
A near-UV LED with a suitably fast current driver is the most practical programmable end-to-end source to investigate. The emitted light, not the electrical drive command, must demonstrate the desired 30 microsecond rise and greater-than 3 ms duration at the selected operating current and temperature. A fast reference photodiode and oscilloscope can verify bright-source timing without claiming that a PIN receiver detects the faint orbital signal.
The inspected M 340 L 5 specifies 340 nm with 10 nm FWHM, so it is not a monochromatic 337.1 nm source. It can test the integrated band response if its spectrum and the instrument filter capture are calibrated. The previously cited LEDD 1 B transition values, about 51 microseconds rising and 79 microseconds falling under its test conditions, are too slow to certify a 30 microsecond optical rise. Use a driver whose loaded optical performance is measured rather than assume its modulation setting is sufficient.
A nitrogen 337 nm laser provides a closer wavelength but typically short individual pulses; the color match does not reproduce a millisecond plateau. Pulse trains, modulation or fluorescent sources introduce additional timing/statistical/afterglow assumptions. An actual discharge can supply relevant spectra and complexity but is harder to use as a stable calibrated waveform standard.
Build the testbed with independently monitored output, traceable attenuation, light-tight baffling and calibrated coupling to the entrance. Test several intrinsic shapes, including shallow/slow ramps, multipulses and bright-to-faint sequences. The goal is to characterize detection, red upper limits and timing over the required photon range, not only display one attractive oscilloscope trace.
Corrections to earlier assumptions
A nominal wavelength or driver bandwidth does not certify the emitted optical waveform.
The cited LEDD 1 B example is not a demonstrated 30 microsecond source.
Equations and their conditions
F20 · expression 1
N_band = integral E_pulse,lambda lambda/(h c) T(lambda) d lambda
Calibrate useful photons through the actual spectral response.
Conditions: Measured LED spectrum, attenuation and pulse energy; no arbitrary full-width multiplication.
What supports this answer
manufacturer test source specs
M340L5 and LEDD 1 B give specific spectrum and transition limitations.
Optical waveform measured by a sufficiently fast reference at actual drive/temperature: 30 us rise, >3 ms duration, stability and recovery
Spectrum/filter capture and calibrated attenuation from faint through overload, with stray-light control
Next inputs in the reviewed register
Actual LED/source and fast driver/load; target 337 band equivalence and reference detector calibration
Related executed studies (0) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
No question-specific numerical study is assigned in the reviewed register.
Sources for this question (4)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
Can 2-4 kg/5-10 W deliver a useful event rate at 550 km?
Direct answer
A CubeSat-class optical head is physically plausible, but a useful orbital detection rate has not been demonstrated. Calibrated selected-event fluences now constrain examples, while the faint population and completeness remain missing. Larger apertures also create bright-event saturation and recovery problems. An accepted packaged optical/power/thermal design, end-of-life sensitivity and joint orbit/population model are still required.
Physics 101
The size and power constraints do not violate an obvious law of physics. Small instruments have measured related optical events in space. That makes the mission worth designing, but it does not yet tell us how many scientifically useful events it will collect.
We now have real observed brightness examples. They show that some events could be very easy to see with a larger aperture, possibly too bright for a simple photon counter. They do not reveal how many much fainter events are missing from the published samples.
The rate question needs two pieces: how often the desired physical events occur in the places and conditions we can observe, and how often our receiver recognizes them correctly. A bigger field sees more territory but also more background and unrelated activity. A larger aperture collects more signal and more background, and may overload on bright events.
The next decision should use a measured receiver threshold and dynamic range, then combine it with a calibrated source/exposure sample. We can present conditional capability curves now. We cannot responsibly promise a useful polar event rate or declare a particular aperture necessary from a selected dozen events.
Full technical answer
A 2-4 kg, 5-10 W optical head is physically plausible given flight photometer heritage, but the requested useful detection rate is not yet demonstrated. Sensitivity, classification, bright-event survival and orbital yield are distinct requirements. A component power number excludes HV, cooling, ADC/FPGA, heaters, safing and spacecraft interfaces unless explicitly included. No packaged 25 mm/wide-field 1U optical design or end-of-life total power budget has been accepted.
The campaign recovered absolute observed 337 nm examples and public plotted waveforms. For one selected Bai et al.(2023) event, baseline-subtracted finite-window integration from-0.2 to 3 ms gives about 7.53e-9 J/m^2 at the original spacecraft geometry, or 1.28e10 photons/m^2 near 337.1 nm. Its selection favored a fast, intense event. It is not intrinsic source energy, a faint completeness limit, or a 550 km measurement. Do not apply cloud attenuation to an observed spacecraft fluence a second time; an inverse-square range adjustment alone is conditional on unchanged directional source/cloud geometry.
Published selected peaks also expose the bright-end problem. The reviewed source ledger's 25 mm/20% examples imply roughly 0.33-24 GH z photoelectron arrival rates for selected peaks, far beyond a simple 17 ns counter's linear regime. Enlarging aperture does not exclusively improve faint sensitivity; it can create saturation and bright-light recovery requirements. Parallel gain/attenuation or another dynamic-range architecture may be needed.
Useful rate requires the joint source-fluence, class, altitude, depth, cloud and geographic/time distribution, including non-detections and calibrated selection. ASIM event cohorts overlap, and particle-rejection algorithm changes alter high-latitude samples. Global selected BLUE rates, radio NBP fractions and the spacecraft footprint cannot be multiplied into a justified polar yield without their common exposure/selection. Night at the cloud, shadow at the emitting layer, spacecraft eclipse, HV enabled state, valid telemetry and observing duty cycle are distinct. The fresh field-of-view population calculation explores unknown slope assumptions; it does not fit a luminosity function or predict events/day.
Corrections to earlier assumptions
Observed spacecraft fluence is not pre-cloud source output and must not be attenuated twice.
Selected bright events cannot define the faint population or unbiased orbital rate.
Larger aperture also raises background and bright-end saturation.
Equations and their conditions
G21 · expression 1
N_useful = integral population(x) exposure(x) P_detect(x) P_classify_correct(x) dx
Rate depends on joint conditions, not independent scalar fractions.
Conditions: x includes geometry, time, cloud, class, brightness, background and receiver state.
G21 · expression 2
S = Phi_ap A eta
Observed aperture-plane photon fluence can be replayed into a hypothetical receiver.
Conditions: No additional cloud attenuation on already observed data; range/angle changes need explicit assumptions.
What supports this answer
selected calibrated observation
BATH figure arrays support a finite-window observed fluence with published calibration and original geometry.
Faint source/dual-band population with calibrated completeness and non-detections.
Packaged optical/electronic/thermal/power design, field response, backgrounds and end-of-life injection-recovery.
Actual orbit/time/exposure, health/safing, telemetry and false-classification controls.
Related executed studies (2) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
R4-N06 · Field size versus an unknown source-brightness law
executed conditional calculation not mission measurement
Re-ran 216 spherical-geometry sensitivity cases with an assumed cumulative source number N(>yield) proportional to yield^(-alpha). Alpha and normalization are not measured or fitted.
The wider field raises covered area but also background, off-axis range and projection losses. The 90/60 relative yield can change sign of preference as alpha changes. The weighted exposure area uses an arbitrary 1 e 20 photon pivot only for units; it is not an assumed event output or an actual effective mission exposure.
Bright-enough population WEIGHT proxy, not an orbital event rate or a full detection-probability integral.
R4-N09 · Detection is easier than trusted morphology classification
executed conditional calculation not mission measurement
Executed 81,000 synthetic signal trials across 81 scenarios: receiver-input 10-90 rises 30, 80, 200 us; plateau-equivalent 3 ms scales 100, 1000, 10000 PE; background 50, 10000, 1e6/s; sampling 10, 50, 100 kS/s. Amplitude, onset and background are fitted using a finite 10 ms prewindow and explicit gates.
A predefined profile-log-likelihood margin ln 19 selects accepted classifications. It is not automatically a calibrated 95% posterior probability or a 95% correctness guarantee. Rejected events, wrong accepted labels and interval uncertainty are all reported.
The plateau-equivalent scale is rate times 3 ms. Actual source totals are smaller because the linear ramp uses part of the finite 3 ms pulse. These are receiver-input morphology trials, not source-depth/class validation. Count-loss occupancy flags are diagnostics only; the generator remains ideal independent Poisson.
Independent profile-objective comparisons and exact seeded reruns passed. The known-template oracle has extra information and is an optimistic benchmark, not an achievable estimator promise.
Calculation limitations
This is a conditional synthetic receiver-morphology study, not a flight sensitivity/completeness validation.
The classes mean 30/80/200 us 10-to 90% rises at a hypothetical linear receiver input after transport; they do not label source depth.
Source brightness, cloud transfer, line spectrum, detector PDE and source occurrence population are outside this count-level calculation.
The practical code has an external coarse onset bracket and a specified event gate. Blind stream triggering and a mission false-alarm rate are unvalidated.
A fixed ln 19 profile-score margin is not calibrated 95% confidence. Always show accepted fraction, rejected fraction and the error interval together.
The 3-class fast-subtype distinction is harder than the grouped fast-vs-slow test; these are separate decisions.
Only exact model-matched templates and constant backgrounds are generated. Real independently labeled optical/radio controls are still needed.
Post-counting electronic filtering, detector excess noise, afterpulses, saturation and deadtime are not simulated.17 ns occupancy is an illustrative hardware flag, not Poisson restoration.
Known-background red limits are invalid if a noisy finite baseline is silently substituted. The provided alternative uses a conservative Bonferroni construction.
Red limits constrain expected registered counts in a specified window, not thermal-source absence. Selection, red-window timing, throughput and population ratios remain physical inputs.
A useful mission event rate is NOT COMPUTABLE without an absolute received-fluence distribution, completeness and exposure definition.
No empirical wide-FOV event rate is inferred from the conditional geometric/population proxy; source geography and coverage remain unknown.
Sources for this question (7)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
The central missing step is connecting an observed waveform and red upper limit to a unique physical source class, depth or current. Keep emitted photons, atmospheric delivery, aperture photons and registered counts separate. The next decisive work is selected-instrument calibration, realistic labeled injections and unbiased population/exposure measurement, rather than adding another unqualified photon-budget estimate.
Physics 101
We were asking several different questions as though they had one answer: does light escape the cloud, does it hit the aperture, does the detector register it, and does the recorded pulse identify the desired physical event? Each step can fail for a different reason.
A useful way to stay honest is to label what we know. A conservation law can be established.A mathematical model can be solved under its assumptions.A different mission can supply a real measurement.But the proposed instrument still needs its own response and uncertainty measurements.
The new calculations help identify dangerous assumptions and useful tests. They show why cloud reflection can add delayed light, why the same rise can have more than one cause, why a red null needs a confidence limit, and why field size changes both coverage and noise. They do not create the missing faint-event population or certify hardware.
The project can now move from a broad feasibility question to specific measurable requirements. A credible design will say how many photons it needs to detect a flash, how many to measure its rise, how strong a red upper limit it can set, and how those answers change across field, background, temperature, radiation and recovery state. That is the evidence needed before predicting a useful mission yield.
Full technical answer
The largest correction is to treat this as a coupled inverse and classification problem rather than a single photon-count calculation. Optical brightness, cloud depth, source current and physical discharge class are not uniquely interchangeable. A fast blue rise with weak red emission is informative evidence, but alternative source histories, ordinary-lightning components, cloud paths and instrumental artifacts must be tested.
Keep four photon reference planes separate: emitted source photons, cloud-exiting angular light, aperture photons and detected photoelectrons/registered counts. Keep source and received spectra, time definitions, bandwidth units and air/vacuum wavelength coordinates explicit. The earlier universal 3 ms source duration, universal amplifier floor, constant aurora assumption and unqualified 4 nm filter response are not retained as facts.
The full source pipeline is part of the measurement. Selected survey counts depend on trigger/live time, HV state, particle vetoes, threshold/background recursion, processing version and association windows. Literal ISUAL 337 positive/777 zero flags are reproducible catalogue labels, not independently verified thermal false positives or physical zero-red events. A published corrected waveform cannot be accused of raw-dead-time distortion merely because a separate raw mean-response diagnostic exhibits it.
The most effective remaining program is a bounded sequence: choose a physically feasible optical/detector candidate; measure its photon-scale, spectral/angular/timing/noise/dynamic-range response; validate source/cloud hypotheses with realistic injections and independent labels; then integrate useful exposure over the actual orbit and source population. Obtain independent full-path calibration and end-of-life tests before using peak brightness as a current estimator. Additional literature reduces uncertainty but cannot replace the missing selected-instrument and faint-population measurements.
Corrections to earlier assumptions
Do not collapse detection, timing measurement, class identification and current inversion into one threshold.
Preserve measurement/model/example/access states and independent cohorts.
More decimal places or more papers do not replace missing mission inputs.
Equations and their conditions
G22 · expression 1
evidence states: measured / conditional model / method / missing
A source link is not a mission validation statement.
Conditions: Retain original units, geometry, selection, access and uncertainty.
What supports this answer
crossstream qualification
Source selection, red-null controls and current inversion retain empirical/access limits.
Selected receiver calibration and end-of-life response.
Realistic source/cloud/noise classes and independent physical labels.
Unbiased joint population/exposure and validated operational false-event control.
Related executed studies (0) and limits
These retained R4 calculations are conditional on their stated inputs. This answer page does not constitute a new execution.
No question-specific numerical study is assigned in the reviewed register.
Sources for this question (8)
Each link opens the detailed record below, including review/access scope and limitations. A linked source is not necessarily a measurement or a validation of this mission.
The reviewed registry contains 116 records; 110 distinct records are cited by these question/evidence entries. These include measurements, model papers, official methods and manufacturer specifications. The access and limits fields define what was actually inspected. They are not 116 independent experiments.
Open publisher methods, comparison table, inversion tests, uncertainty table and availability inspected
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_round_01.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Publication date as recorded
2026
Where to inspect
Published model description/optical-output claims; API availability is separately audited in BG 021. No exact 337 numerical result taken from this source.
Limits
Model scenario output requires particle inputs and is not a Kp-binned observational climatology.
Ground atmospheric corrections cannot be applied blindly to upward/nadir geometry.
Current API docs must be checked: September 2026 release changed defaults and fixed mean-versus-characteristic energy handling.
Open primary methods and instrument bands inspected
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_round_02.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
LIME instrument/model band coverage and validation; actual v 2 observed-wavelength archive inspected separately in campaign.
Limits
An interpolated or extrapolated 337 nm prediction needs an independent UV uncertainty budget; the actual photometer has no 337 nm band.
NIST primary abstract and instrument/uncertainty summary inspected
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_round_02.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_round_02.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Official page mission-considerations and uncertainty discussion; lunar viewing/opposition constraints.
Limits
These are general mission/calibration characteristics, not a calibrated 337 nm lunar irradiance or a free operational resource.
Primary USGS publication record and abstract inspected; original paper discovered
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_round_02.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
ROLO observed wavelength bands and empirical phase/libration model; lowest fitted band about 350 nm.
Limits
337 nm lies outside the measured wavelength range; absolute accuracy and fit residual are different quantities.
NASA 27 pageaccepted manuscript: instrument, noise, observingwindows, moon, aurora, lightning and availability inspected; publisher metadata compared
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_round_02.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
PDF§2.1, 2.4, 3, 4.2, 4.3, 4.6, Open Research
NASA metadata publication date 2025-10-01, acquired 2026-02-10
Limits
Do not treat limb examples or visible moon examples as a polar nadir 4 nm 337 climatology. Need raw twilightradiance, postprocessing, errorcovariance and filterfolding.
Primary full instrument and calibration sections inspected
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_round_02.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
§2.2 and Table 1
§3.1, Figure 5
Limits
The described bands are 427.8, 844.6, 777.4 and 650-700 nm, not a measured 337 nm profile.
Illustrates calibration/energy-ratio degeneracy, not 30µspolar 337 climatology.
Actual Open API document fetched 2026-10-21 T 22: 28: 53 Z, 91408 bytes, SHA 256 d 38533 aa 0 dc 168 a 5458 c 2617 df 4 bc 3 b 1 c 906 dff 46 a 9504 ec 0 b 6 d 23 f 52 c 4 7e3 dd
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_round_02.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Current public Open API response schema and optical output fields, archived during campaign; arbitrary 337/777 outputs not declared.
Limits
A paper-level capability is not proof of a callable 337/777 API. Do not send invented flags or relabel 427.8 as 337.
This audit corrects the earlier broad BG 006 API availability implication; source code/access or another model must be checked before running 337/777 scenarios.
Pinned source, license, example inputs and output inspected; rounded example integrated locally. No compiler/model execution.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds_03_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
model_code/NCAR_GLOW_Glowlicense.txt
model_code/NCAR_GLOW_in.basic.aur
reference_numbers_audit.json
Limits
An imposed precipitation scenario is not a measured moderate-Kp radiance or orbital distribution.
Academic research license retained; no fresh atmospheric model run claimed.
Primary abstract and Table 1 inspected. Full text and figures are subscription restricted.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds_03_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Abstract
Table 1
Limits
Not justified to say UV lunar measurements do not exist. Actual 337 nm albedo, geometry and error budget have not been extracted from this restricted full text.
Primary publisher cached Section 8.4/Table 19 recovered. Direct PDF access 403/timeout, not a fully read local PDF.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds_03_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Table 19
Section 8.4
reference_numbers_audit.json
Limits
Interpolation choice is not an uncertainty interval. Incident celestial radiance must be integrated over cloud illumination and BRDF; it is not an Earth-filled nadir background.
README, Newsletter 6, directory listings and one small public Net CDF inspected; header and geometry audit saved.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds_03_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
round 04/raids_readme.pdf, pp 1--2
round 04/raids_nuv_qualification.json
round 04/raids_nuv_sample_header.json
Limits
Single commissioning-day limb sample is neither nadir cloud nor polar climatology. No radiance numbers exported.
README requests PI authorization and advance draft review before publication/presentation. Private qualification only; no external contact made.
A scan is not automatically simultaneous across wavelength; pixel timing matters.
Primary instrument, calibration, continuum-separation, conclusion and data-availability sections inspected; full XML retained.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds_03_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Section 2
Section 3
Data availability
round 05/Noll et al.(2024)_primary.xml
Limits
Low-latitude ground data need extinction/sky-component separation and satellite illumination geometry.
A continuum-only spectrum omits structured O2 bands and must not replace total 337 background or double-count a complete sky model.
No 337 nm value has been inferred from a graph or uninspected archive.
Official variants/metadata inspected; 2.44 MB v 2 HDF 5 file retrieved/hashed and solar-only filter integrations executed; 2021 primary uncertainty text inspected.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds_03_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
round 05/HSRS_p 1 nm_v 2.nc
round 05/spectral_geometry_audit.json
Coddington 2021 Section 4.2
Limits
Sunlight is not nightside or lunar irradiance. No absolute Moon flux was derived.
1.3% UV broadband uncertainty is not an independent error for every fine spectral pixel; individual line discrepancies can be larger.
Air/vacuum wavelengths and actual measured filter response must be resolved before mission folding.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds_03_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Sections 5.4, 6.3--6.6, 7.2
Limits
No machine-readable coefficient set retrieved yet.
UV calibration/polarization features and GOME 2 comparison differences limit an unqualified high-accuracy 337 claim.
Do not confuse 5 nmreflectance grid with complete solar-spectrum resolution.
Official full overview, model-range descriptions and data-resources section inspected.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds_03_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Overview
Table 1
Data resources
Limits
LCS service scope starts 350 nm and the discussed base-function fit grid starts 350 nm, despite broader model range elsewhere on the same page. Actual 337 nmv 2 coefficients/support need clarification.
No claim that coefficients are publicly downloadable or that we ran this model. No external contact made.
Verifiedranged ZIP centraldirectory and 6 smallmembers CRC 32/SHA checked; whole 151 MBZIP notdownloaded. Actualmean spectrum and README inspected. CC BY 4.0.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds08_11.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Zenodo 8335837 mean continuum table, 0.5 nm grid; processed selection/line-removal conventions in Noll et al.(2024).
Limits
The extracted pseudo-continuum includes many unresolved O2 ultraviolet bands; strong resolved lines were removed. It is not the total polar 337 or 777 nm sky background.
Broad extraction windows and a 0.5 nm grid do not establish a sub-nanometre filter optimum.
ESO documents pipeline air wavelengths. An explicit final author wavelength conversion was not verified. Registration of the spectrum and filter remains conditional; the small photon-energy change does not establish line-position accuracy.
Fullauthorarxivv 1 HTML inspected, final CDSR ead Me/99 row Jtable/15.63 MB gzip 49152 x 99 sky cube downloaded andhashed. Finalpublisher directfetch 403; indexedfinaldataavailabilityqualifies CDS route.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
backgrounds\source_register_rounds08_11.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
CDS J/A+A/691/A 330; 99-bin mean intensity table and celestial NESTED Nside64 direct+diffuse sky cube.
Limits
The ultraviolet extension uses stellar-atmosphere/extinction models and an empirical diffuse-light fit, rather than a direct 4 nm measurement at 337 nm.
The roughly 13 nm wavelength bracket cannot resolve 4 nm spectral structure. Interpolating it does not provide a new pixel covariance or confidence interval.
Six hypothetical surface normals are neither global bounds nor an orbit distribution. Actual orbit date, cloud BRDF and atmospheric transfer are required.
Author preprint dated 2023-11-01; all 7 pages text inspected, Figure 6 p 5 visually inspected. Published JQSRT 312(2024)108809.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Spectroscopic line-position/strength model validated by author against experimental FTS and SPECAIR under stated laboratory conditions. Not an NBE radiometric calibration.
Limits
Official current ExoMol metadata identifies the 20240705 C-B extension. Public .states/ERJ.trans requests timed out here: this is a local access failure, not global unavailability.
Zenodo 5717009 contains the older 20190916 archive, created and modified in 2021; it does not include the Jans et al. 2024 C-B extension.
195-page scanned official NIST PDF retrieved with SHA 256; selected printedpp 113, 115, 117, 123, 139, 226, 306, 307 visually inspected. Not all 195 pages read.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Actual SQL ite 11225088 B and selected sourcefiles retrieved andhashverified. No software installed or outsidepackagecode executed. Own SQL ite/arithmetic audits executed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Pinned Moose/original Massive OES data/query audit; lower-state link, energy/J and coordinate checks. These are conditional reconstructions, not an accepted NBE spectrum.
Conditions
Numerical spectroscopy implementation and diagnostic, not a measurement.
Limits
The attempted initial v0-v0 filter fold was invalidated; no physical capture fractions from it are retained. Later candidate mappings are conditional diagnostics.
Population/spin weights, original line positions and Einstein A values still require independent primary comparison.
Moose code uses MIT licensing, while original MassiveOES terms include local-use, sharing and attribution conditions. Their application to bundled spectral data remains unresolved; restricted raw data are kept private.
Official NIST documentation sections IV, VI, VII and Jones 1981Eq 1/introduction printedp 27 inspected throughbrowser. Wholevolume notdownloaded/read.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
NIST refractive-index-of-air/wavelength documentation; standard-air coordinate conversion, not a physical filter-vacuum shift.
Conditions
Ownstandardaircoordinatecalculation using publisheddispersion formula; not measured anvil refractivity.
Limits
No nominal wavelength was silently changed in the frozen calculations.
NIST discusses air-index models over 300-1700 nm with condition-dependent uncertainty. A nominal dry-air conversion does not certify old table accuracy or an event band centroid.
Cross-stream source package: exact ten-source list/access levels in sourcebranchprimaryreview, not a single independent dataset.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Primary full HTML methods, results, appendices anddataavailability inspected; XML/metadata access separate inreceipt. No rawimage/video download.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Original 77 pagepublic NASA PDF locallyretained; summary/calibration/filtertable andselectednight/daytables visuallyverified.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
NTRS metadata: Workof USG ov, Public Use Permitted; privateworkingcopy.
Publication date as recorded
1971
Kind as recorded
Measured filter-channel equivalent brightness, dark-current corrected; not spectrally pure N2(0,0).
Where to inspect
NASA-CR-124719, PDF pp 4-5 filter/calibration; pp 8, 11 selected night observations; page 67 bright sample is midday.
Conditions
Measured filter-channel equivalent brightness, dark-current corrected; not spectrally pure N2(0, 0).
Limits
Airglow, starlight, moonlight and other features were not removed; the weak-end accuracy at a few tens of Rayleigh is described qualitatively.
The inspected 1971 summary does not expose a full filter curve, equivalent bandwidth, formal uncertainty, per-wavelength timestamps or complete zenith extinction correction.
The visually verified 1.350 kR example on printed page 67 is Flight 13 midday aurora; it must not be labeled nightside.
Original 102 pagepublic NASA PDF locallyretained; printedpp 38-40, 46-50, 54-55, 60 inspected; keycalibration/tablepagesvisuallyverified.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_01.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Authors as recorded
P. Yang; L. Bi; B. A. Baum; K.-N. Liou; G. W. Kattawar; M. I. Mishchenko; B. Cole
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_02.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
publisher_full_HTML; methods 2.1-2.3, conclusion and code availability examined
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_04.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Publication date as recorded
2021-01-12
Where to inspect
sections 2.1-2.3 and code-availability statement
methods and conclusions; our applicability assessment
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_04.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
complete developer README and selected changelog entries read
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_04.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
author/institution_hosted_full_PDF; all 8 pages extracted, sections 1-4 read closely
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_05.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Authors as recorded
Barry A. Bodhaine; Norman B. Wood; Ellsworth G. Dutton; James R. Slusser
Official docs and source transient_prbvolpath.py read; no renderer installed or executed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_06.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Original 243 pagegovernment PDF privatelyretained; selectedprinted 2-3, 36-39 visuallyread. Scanned OCR ispoor. NASA 2005 Appendix A independentlycrosschecksformulas.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_07.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Primary 4-pageauthorpreprintmaintextread, publisherabstractverified. Nojournalpaywallbypass. Preprint PDF dateline 2018-09-03 differsfromv 1 submissionhistory; version URL andhashretained.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_08.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Authors as recorded
T. Svensson; R. Savo; E. Alerstam; K. Vynck; M. Burresi; D. S. Wiersma
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_08.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Authors as recorded
Stephen G. Warren; Richard E. Brandt
Publication date as recorded
2008-07-31
Where to inspect
section 2 paragraphs 6-7, p 2; section 7 paragraph 34, p 7
Selected primary methods/results and table 1 read; primary XML retained privately. Actual OASIS table namespace parsed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_09.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Current official documentation and basic-usage pages read; no software installed/run. Manual link advertises 2.0.6, not yet privately retrieved.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
cloud_transport\source_register_round_09.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Authors as recorded
libRadtran developers
Publication date as recorded
Live documentation, last modified2024-12-22
Where to inspect
documentation lines 30 and 49-58; basic_usage MYSTIC section
Full author manuscript v 2 and primary author-institution published-version metadata; Astroparticle Physics 31 (2009) 305-311.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
sections 4.1-5.2; Table 2; Eq.(2)
Limits
A bright calibrated photodiode transfer is useful even if a PIN detector is too noisy for the faint science signal.
This is not a 1% absolute 337 nm standard or a 30 µs pulse simulator. Whole-path and piecewise component calibration can differ; narrow mission filters need finer wavelength sampling than this study.
Official STS cI page and primary FITS file retrieved; header verified, 288000 bytes, SHA 256 7201cd8566d53d384430c3d5fd2ae8d8122e9 a7f765711748b49336f63237672.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
CALSPEC composition/cautions/history; FITS primary HISTORY and SCI BINTABLE header
Limits
A stellar standard exists at 337 nm, but this does not give a complete 1% receiver calibration. Actual response, backgrounds and pointing remain required.
The private Vega fold is an ideal-band photon-statistics illustration; spectral resolution is about 0.55 nm near 337 and 0.98 nm near 777, not a line-resolved nitrogen/oxygen reference.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Primary author preprint; journal 2010; automated preprint footer has an inconsistent 2024 date
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Full primary team-hosted published article, 2012; 7024538-byte PDF read in memory, extracted text/provenance saved. Shared source with alternative_check; complementary calibration-path qualification.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
sections 2-2.1; printed pp.73, 75-76; Tables 1-2
Limits
These are internal mainly-red LED response trends after preflight voltage-gain correction, not direct 337 nm PDE aging or isolated radiation damage. LED stability is assumed from duty, dose and temperature arguments. The reference pathway does not establish science-wavelength throughput stability.
The source-population stream independently retrieved the same paper; cross-link its record rather than count this as independent evidence.
Primary publisher full HTML verified through af=R; direct full URL initially failed. Supporting DOCX returned 403. Official data recovered independently.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
No instantaneous 337 peak-current law is determined; source conditions differ from 15-18 km NBEs.
The corona PMT's 340 nm/12 nm filter and a 10 nm spectral sum are not a 4 nm 337.1 nm receiver.
The released workbook lacks the qualitative Figure S16 optical waveform; Table 2 photon-rate normalization was not independently reconciled with flux-distance data.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
LEDD1B modulation specifications table and notes
Limits
This driver is not automatically a 30 µs optical-rise simulator. Current step, LED load, bias and optical tail must be measured.
The driver waveform is not a calibrated optical waveform; use a fast reference detector and verify the complete attenuator/sphere path.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
M340L5 table row; operating conditions
Limits
It can supply light through a 337 nm passband, but this is not monochromatic 337.1 nm emission. Spectral overlap, attenuated photon number, heating and optical pulse shape require characterization.
An exact M337L4 part was not verified; do not recommend that unconfirmed identifier.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
sections 3-4, pp.3-6
Limits
Those uncertainty figures are for the demonstrated setups, not a guaranteed 337 nm wide-field two-channel service.
A CW efficiency measurement must be checked against actual transient rate, background and detector recovery behavior.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Calibrating the spectral responsivity of detectors
Limits
The better-than-0.01% cryogenic-radiometer statement applies at specific visible laser wavelengths; it is not a quoted 337 nm receiver uncertainty.
A separate routine single-photon-detector service starts at 380 nm. Analog responsivity coverage does not automatically establish 337 nm photon-counting calibration, timing capability, available booking, or an end-to-end uncertainty. No facility affiliation is claimed.
Primary open publisher HTML, 14 June 2022; methods and figures inspected.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
sections 4.4-4.6; section 5 and Figures 9-11
Limits
Its large CCD instrument and orbit are not this receiver. Do not import its degradation rate or long-exposure correction performance into microsecond data.
The row anomaly is probably external obscuration by loose insulation, not ordinary detector radiation aging. Internal LED constancy alone cannot exclude a blocked entrance path.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Figure 1 and associated introductory calibration paragraphs
Limits
This demonstrates an architecture for separating reference aging from instrument aging, not a universal UV stability guarantee for a Cube Sat.
The result is relative temporal stability; initial absolute calibration and retrieval soft-calibration have distinct uncertainties. A single unmonitored LED cannot uniquely separate its own drift from detector gain drift.
Primary author manuscript 26 Oct 2023; journal NIMA 1064, 2024
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Full primary author manuscript v 2, 18 June 2017; NIM A 863, 35-46.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
The equations assume linear independent response and Poisson photoelectrons; background and under-amplified pulses bias occupancy if misclassified.
This calibrates charge per collected photoelectron and multiplication noise, not absolute incident-photon PDE. Keep optics/photocathode calibration separate; do not blindly apply to correlated SiPM avalanches or saturated data.
Official landing metadata/file inventory verified; API 403. NoPSDATA downloaded.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Description; Files; v 1 released 2026-03-24
Limits
Compatible Pico Scope software/export and complete relative calibration are required for waveform reanalysis.
ENTLN data require an external request; no contact made. No unrestricted data license established from retrieved landing text.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Table 1, p 1001
sections 2.1-2.2, pp 1001-1002
Figures 3-4
Conditions
R 7400-06 UV PMT; R 7400-20 red PMT; S1227-1010BQ broadband PIN
Design/preflight paper; 337±5 nm; 42.7 deg UV full FOV; 20 kHz 12-bit; 8 mm PMT photocathode.
Limits
The design sensitivity of 0.1 uW/m2 differs from the later flight table.
Telecentric chief-ray control does not establish zero angular spread.
All six channels must not be assumed to fit within 1U; box and power definitions require checking.
Official current manufacturer product specification.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
specifications and operation modes
Limits
A short minimum exposure is not a 50 kHz full-area frame rate. Reducing the ROI changes accepted area/field unless the optics are redesigned.
Summed independent read noise scales approximately with square root of pixel count; 1.6 electrons per pixel is not 1.6 electrons for the full receiver. Exact 337 nm QE and flight qualification remain unverified.
Official datasheet TPMO 1 122E01, October 2022; current H16722-50 product page also verified.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
PDF p.1 specifications and footnotes; p.2 cooling and dimensions
Limits
Red PMTs need not all have only a few percent QE. But a 5 mm cathode imposes an étendue constraint, and thermal rejection in vacuum needs design.
No flight qualification or radiation result is given. Dark anode current is not an RMS noise specification.
Official manufacturer datasheet; p.3 indexed table and footnotes verified.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
p.3 electrical/optical table, footnotes 4-5
Limits
Scaling to a larger mosaic scales dark counts and capacitance too. Bias, package/window, temperature and readout must be retained.
A spectrum or separately calibrated PDE is needed at 337.1 and 777 nm; nominal peak PDE cannot be reused.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
p.1 table, conditions Ta=25 °C
Limits
The 420 nm figures are not calibrated 337 or 777 nm values. Large area trades increased capacitance against collection area; the receiver amplifier must be evaluated with that capacitance.
Catalog availability does not imply flight qualification.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
sections 3-4 and 6; printed pp.3-5, 8
Conditions
S13361-3050-NE-08 8 x 8 array, 3 x 3 mm per tile, 50 micrometre cells; RADIOROC/Artix7 prototype
Quoted reproducibility is not a full absolute uncertainty budget.
The text labels S150 C as Hamamatsu although S150 C is a Thorlabs product; reference-detector identification needs checking before reproducing the calibration.
Official Texas Instruments technical article SSZT 554, Bryan Lizon, January 2019; full text read.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
questions 4-5, 9-10: ENBW, total chain filtering, output-data-rate and clocks
Limits
The article's 6 nV RMS example uses 2.5 samples/s, not 10-50 kS/s; no such low noise is assumed for this photometer.
Nonwhite detector/TIA noise and aliased out-of-band noise require PSD integration through the actual chain. A single bandwidth or nominal bit depth does not determine mission photon threshold.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Noise in TIAs
Achieving Higher Bandwidth with Higher Gain TIAs; Figures 6-9
Conditions
LTC6268 FET-input amplifier; 499 kohm feedback
Limits
Specific low-capacitance setup; not prooflargearea UV receiver reaches equalnoise.
Counterexample disproves a universal 1 e 10 V/A·Hz limit, not fundamental photonbudget limitations.
Official manufacturer technical note, full 10-page PDF.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Multiplying an already multiplied total dark current by M² would overestimate the noise incorrectly. A catalog total dark current alone cannot uniquely partition bulk and surface noise.
For fixed measured total dark current, the inferred dark PSD lies between 2 qI_D and 2 qMFI_D under this model, not a single universal value.
Official NASA report, 21 January 2026; Appendix 3 updated 5 October 2025.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Appendix 3, printed pp.62-64; Figures A 3-1 to A 3-3
Limits
A lidar range tail converts to time using 2 d/c. These observations do not establish a 0.2 ms tail for our detector.
The useful requirement is amplitude- and duration-dependent impulse/step characterization: cloud timing and receiver timing cannot be separated by assuming a single universal linear kernel.
Official ET Enterprises technical reprint R/P 065 by A.G. Wright
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
section 3, printed page 2; Figure 4; section 5.3
Limits
Legacy example amplifiers and nanosecond scintillation pulses are not our chosen receiver.
Actual transfer function, saturation and recovery must be measured over the millisecond pulse and background range.
Official manufacturer 5-page PDF, revision R 19, 2 January 2019; current product listing verified.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
p.1 block diagram; p.2 specifications and measurement footnote
Limits
Noise at one frequency must not be multiplied by square root of 50 kHz and substituted for the measured integrated figure.
Detector capacitance adds voltage-noise current. Bench hardware is not flight qualified; this is a counterexample to the claimed universal gain-bandwidth bound.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
section 3.3, printed pp.9-11 and Table 5; section 4.1 p.12
Limits
The 250 nm internal check does not establish absolute drift of the whole 337 or 777 nm optical path.
Heritage proves an architecture, not the performance of a 25 mm entrance aperture within the same volume and power. Exact PMT part numbers are not published here.
Official Analog Devices tutorial by Walt Kester; full 12-page PDF, Rev.A 10/08.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
pp.1, 4-8, Figures 1, 4-7; baseband sampling and antialias filtering
Limits
A Nyquist frequency does not specify a practical filter cutoff or its stopband margin. The tutorial's separate bandpass-undersampling examples do not apply directly to a DC-to-transient photometer.
Exact sub-sample parameter inference requires pulse-model, phase, noise and filter assumptions; raw cadence alone does not validate rise-time discrimination.
Full primary publisher HTML now verified using?af=R after initial timeout. Table 1 and relative calibration methods inspected.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Primary publisher search-index full text plus author institution metadata; direct publisher open returned 403. Published online 24 June 2026, volume 46 August 2026.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
sections 2.2, 2.4.1, 3.1.3-3.1.4; Tables 1, 3, 4
Limits
This refutes a universal 1 nA amplifier floor, but gives no 337 nm detector-coupled or radiation-qualified result. Actual sensor/cable capacitance matters.
The reported average density is RMS divided by square root of nominal bandwidth, not a measured flat white-noise PSD. Its 250 kHz variant has substantially higher 66.18 pA RMS noise.
Peer-reviewed consensus article; accepted manuscript available at Cambridge, repository DOI 10.17863/CAM.122981
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Noise, NEP/detectivity, temporal response and benchmarking sections
Limits
Method guidance is not experimental performance of a candidate flight detector.
Responsivity and timing measured at bright illumination cannot automatically establish behavior near the faint-event threshold.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
p.2 f/5 modeled cone; p.3 thermal results
Limits
Neither thermal coefficient is a measured 337 nm property. The result does not solve a 60-90° full field in a compact receiver.
Environmental cycling and hard coatings do not by themselves establish radiation, contamination, or vacuum qualification.
Official current catalog, 12.5 mm product; page specifications verified.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
product specifications table
Limits
The OD 6 product name does not guarantee visible/NIR blocking. Leakage must be integrated over the detector response, especially for silicon.
The 25 mm page did not expose a complete primary specification during this review; secondary reseller values were not transferred. No space qualification follows from the catalog.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
printedpp 5-7; equation 1; section 3
Limits
The geometrical-optics relation is transferable; the far-infrared materials, bolometer performance and diffraction numbers are not.
Finite filter cones and detector acceptance constrain practical ultraviolet throughput.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Official NIST tables and handbook wavelength convention; full table inspected.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
O I 7771.94, 7774.17, 7775.39 Å rows; Handbook section A Wavelengths
Limits
Source spectra, solar standards, monochromator calibration and filter specification must use a declared common convention.
An air-vacuum coordinate conversion is separate from an actual coating passband change when changing ambient pressure. Both require metadata.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Primary author manuscript v 2, September 2011; section 3.2 and figures inspected. PDF rendering shows a later automatic 2018 header.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
section 3.2, PDF pp.5-6; Figures 2-3
Limits
This is a cold underwater/ice assembly, not a satellite rate prediction or proof that our fused-silica filter will behave identically.
It demonstrates that millisecond internal light bursts are physically possible; a >3 ms duration criterion alone is not a universal particle/noise veto. Copying the dead-time cure would impair this mission’s rise-time measurement.
Primary author manuscript v 2, 1 July 2021, full PDF.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
section 2 pp.3-4; section 3.1 pp.5-8
Limits
Trigger fractions are selected acquisition statistics, not a transferable probability for this photometer. A UV-chain-only artifact need not appear in a red null channel.
The separate first-day high-voltage/daylight fault damaged 51 channels; it must not be presented as cumulative radiation degradation.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Primary open-access full text verified; raw datasets available only on author request, which was not made.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
sections 2-4; Figures 3-7; Appendix annealing time
Limits
The stated orbital years assume 600 km and 10 mm Al; this is accelerated bench evidence, not flight validation.
Conditional versus periodic differences may include small-sample variation. A 2 kcps maintenance threshold chosen for QKD is not a photometer threshold.
Very cold operation/thermal cycling requires an actual receiver thermal-power and qualification design.
open author manuscript; complete methods inspected
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
open author manuscript; methods and measured slopes inspected
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Methods: Radiation effects on GM-APDs; Results Figs.6-7; Data availability
Conditions
Two SAP500 GM-APDs; free-running passive quench with feedback maintaining excess bias
Spoo Qy-1, 51.6 degree inclination, altitude 418 to 378 km during 600 days; DCR comparison normalized to 10 C
Limits
Part-to-part variation and anisotropic geometry prevent transferring a universal growth coefficient to a fixed 550 km polar orbit.
Underlying data and code are available on request, not public.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
sections 2-5; Figures 2-4
Conditions
Tower 180 nm 4 Tteststructures similar tofinal CIS
Co 60 HZB/CERN; maximumapplicablebias; roomtemperature; measurements up 2 monthsafterirradiation.
Limits
These are test structures, not a final flight device; nominal 300 s exposures differ greatly from a 20 us transient sample.
Dark-current random telegraph signals and prompt charge hits require separate treatment. Reported dwell times range from microseconds to weeks.
No generic CMOS NEP or polar-orbit transient rate follows from this study.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
detectors_optics/evidence_ledger.jsonl
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
specification table; spectral-response curve
Conditions
sglux SG01D-18S, 0.50 square millimetre SiC photodiode
Dark current at 25 C, 1 V reverse bias; peak responsivity at 280 nm
Limits
Peak responsivity must not be applied at 337 nm.
No internal multiplication; capacitance and front-end voltage noise remain relevant.
Optical UV-aging evidence is not automatically proton displacement-damage qualification.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Public supplier/reference document, scope qualified by the linked audit; no unrecorded full-text access claimed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Open primary article; source/model methods inspected in project and official article reopened 22 October 2026
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
Official article and frozen project source/model audit
Rights
Article CC BY 4.0. No third-party restricted raw observations reproduced.
Where to inspect
Sections 2-4: Monte Carlo cloud source/observer implementation and applications
2015 publisher abstract and snippets; full 2014 Uppsala author dissertation downloaded
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Three openauthor PDFs downloaded, text extracted, SHA 256 saved in emission_kinetics/retrieval.json
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
AIRFLY 2013 absolute result
AIRFLY 2007 pressurepaper
AIRFLY 2008 Table 1 and Eq 1-3
Limits
The 216.65 K case is an extrapolation, not a measured flight cloud profile or source electron-energy distribution.
A nanosecond excited-state lifetime cannot constrain a continually re-excited source to a nanosecond optical duration.
Public 95.1 MB ZIP inspected by HTTPR ange; selected MATLAB figures retrieved with ZIPCRC checks. Full paper inspected. Numeric figure arrays decoded without executing MATLAB.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Selected MATLAB figure entry retrieved by HTTPR ange, ZIPCRC/SHA 256 verified; boundedparser readsnumericdata without executing MATLAB objects
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Published Figure 4 numeric axes andlineobjects
Limits
A useful replay fixture, with no classifier confusion matrix or population completeness.
Publisher/full redirects toabstract; abstract anddata statement verified, full methods not obtained. Public source tagv 0.1.1 inspected statically, not executed.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Full primary HTML; 18, 338, 400 byte CCBY catalogue retrieved with exact supplied MD 5 verification
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Complete primary main paper; supplemental access failed
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Laboratory streamer electric-field method and 70 mbar profiles; optical curves normalized, not an absolute 337 nm-current calibration.
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Instrument/resultssections
Spectralresultfigure/text
Limits
The stated raw MAJIS data access remains proprietary until about six months after the 2029 EGA3 encounter; no archive spectra were downloaded.
No 337/777 ratio, intrinsic line width, source-current calibration or mission population inference follows.
Full publisher HTML inspected; numerical Table 1 retained as published
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Table 1
Limits
Overlaps Liu Nature 2021 twelveeventstudy and 2025 elves subset; do notadd counts as independent samples.
Shows source population not universally>3 ms. No completeness.
Full text accessible via publisher web reader; PMC frontpage presentedbrowserchallenge
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Primary ESA technicalequation plus NASA sourceand GOLD twilightpaper
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Full publisher HTML, including March 2025 erratum, inspected. Public code/data DOI 10.26208/xtw 7-zk 59 metadata retrieved; actual archive files not retrieved.
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Sections 2.2 and 3, equations 7-10
Section 3 Figure 3
Erratum, corrected 29 March 2025
Limits
Forward reproduction demonstrates consistency within the assumed source model, not uniqueness across geometries or current profiles. The paper explicitly says deconvolution non-uniqueness is not addressed.
No optical/current calibration or new 337 nm measurements.
Full text accessible through web reader; publisher shell GET previously 403; author-repository PDF also located but not downloaded
Access date
2026-10-21/22 campaign; exact requests retained in source receipts
Role
Independent primary measurement, published model, manufacturer specification or official reference as explicitly indicated in access/conditions; not all sources are measurements.
Registry provenance
sources_population/evidence_ledger.json
Rights
License/redistribution not established by this record. Paraphrased findings and original links only; no restricted raw archive data reproduced.
Where to inspect
Figure 5 caption
Figure 6 caption
Table 1
Section 4 paragraph after Figure 6; concluding paragraph
Open the reproducibility package for numerical verification. An implementation check and Monte Carlo standard error do not establish the physical assumptions or a mission event rate.