DNB Cross Talk Review - PowerPoint Presentation

Slide1

DNB Cross Talk Review

1 July 2014Steve Miller (CSU-CIRA)

Contributions from

William

Straka (UW-CIMSS), Curtis Seaman (CSU-CIRA), Lushalan Liao (NGST), Stephanie Weiss (NGST), and Steve Mills (Renaissance Man Eng.)Slide2

Overview

Raytheon FP13 lab tests suggest cross talk may exist between certain detectors & aggregation modes in the High Gain Stage.Out of specification by up to factor of 3, depending on certain pixel/aggregation mode combinations.

Would manifest as possible spurious signal in detectors 1 and 16 (i.e., first and last scan lines of a 16 scan line bundle).

In the product paradigm, the cross talk is from detector 9 into 16, and from 8 into 1 (this is flipped from Raytheon charts which are shown in ‘sensor order’.)

Options to correct non-compliance:

Accept waiver for use-as-is.

No cost impact, no delivery schedule impact

Change require to encompass existing performance as measured for F1 VIIRS NPP and JPSS J1 (requires a corresponding verification

documentaion

change to TPs, RVP, RVSS, etc.)

“Very low” cost impact for req. and

verif

. document changes, no delivery schedule impact

Redesign DNB Timing Card (FPGA timing)

“Low” cost impact, no impact on instrument delivery schedule

3

rd

option (timing card) ostensibly benign, but “low cost” is relative and there may be other risks and unanticipated impacts to a redesign

Need to establish that the issue is significant enough to warrant Option 3

This document does not constitute an official guidance/recommendation on the waiver request—simply supporting informationSlide3

FP-13 Testing

Cross talk into HGB side in response to HGA side illumination

Recipient detectors 1 and 16 (off-diagonal) are the main offenders, although some adjacent detector cross talk is prevalent in HGB (broadening of the diagonal)

No cross talk of note in the mid and low gain stages—this is a HG issueSlide4

EFR3461 Investigation

DNB Module Level using J1 FPA at operating temperature (simulating operation conditions)

Note that the cross talk into detectors 1 and 16 are most prominent at

Agg

Mode 1 (nadir) and nearly negligible at

Agg

Mode 32 (edge of scan).

Timing must be more precise at lower aggregation modes, when more sub-detectors are being combined.

Note the significant performance differences between HGA and HGB…nominally the two should be identical

HGA and HGB are typically averaged to produce HGS in operationsSlide5

Courtesy, L. Liao (NGST)

Possible sign of spurious cross talk in detectors 1, 16

(Note: instantaneous values may be much higher)

Statistics suggest that the effects of the cross talk do exist in the data

Difficult to find (uncommon) but may be significant when they do appear.

November 2012

HGSdn

= 0.5(HGA+HGB)Slide6

Possible Cross Talk

Occasional spurious signals in the presence of strong gas flares.

Adjacent darkening in adjacent lines of same scan angle—sign of possible cross talk.

Not seen in cities. Sometimes seen in fires…

7 May 2014 2129 UTC Persian Gulf Gas FlaresSlide7

Provisional Summary

Operational data show indications of the cross talk problem.Difficult to identify in DNB imagery. Extreme cases of isolated gas flares may provide examples of worst-case scenarios.Effects not discerned over most anthropogenic light sources (cities, ships, etc.)

We continue to search for examples, including possible HGA and HGB exclusive imagery collected during VROP 702 (pending).

An Option 3 fix to the DNB timing card may introduce unanticipated risks.

There exist other issues (stray light leakage) which may present more pressing issues in terms of DNB performance.