JPSS-1 VIIRS Prelaunch Reflective Solar Band Testing and Performance
, Hassan Oudrari 2, Evan Haas 1, Qiang Ji 2, Frank De Luccia 1 and Xiaoxiong Xiong 3Round 1
Reviewer 1 Report
This paper delivers successful prelaunch calibration results of JPSS-4 VIIRS instrument in the Reflective Solar bands using TVAC tests at different temperatures. This paper includes results from the dynamic range, calibration coefficients, detector noise and radiometric uncertainties for JPSS-1 VIIRS. This paper is well planned, organized and presented. I recommend this paper to be published with minor corrections. One last suggestion is that the uncertainty estimation section (last one) was too brief. Please elaborate a little bit more. Please refer to my comments below.
Thank you for your hard works.
At Page1 line 32, I guess the altitude is 829km. Please check and add a reference.
Line 39: ‘that covers’
Page3 & line 87, J2 VIIRS uses third order polynomials in SWIR bands. Please add an explanation. Also in equation 4 in page 6 line 177.
Line 187; Change ‘dnout’ to ‘dnout’ (italic and subscript for out).
Line 210; Please define ‘LSIS’.
Line 218; Why do you need TMCSIS since you have SIS100 with different lamp power levels and attenuation screens? à Answered around 251.
Line 276; What is the source of SV target. It was not the deep space. In this case, how dark and stable was the source. Please add some explanations. Is there any figure of test bench layout for SIS, TMCSIS, and SV sources?
Line 343; There was an in-depth study on this saturation issues. Please mention follow work.
W. Wang, C. Cao, S. Blonski, Y. Gu, B. Zhang and S. Uprety, "An Improved Method for VIIRS Radiance Limit Verification and Saturation Rollover Flagging," in IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-11, 2022, Art no. 5403011, doi: 10.1109/TGRS.2021.3097896.
Line 387; Would you add radiance scale image for Figure 9? How’s the dual gain transition? Was radiance values smoothly changed around the point?
Line 422; What do you mean by quantization? I assume that the radiance or dn response stuck at a certain level. Please revise the following sentence to explain this situation instead of saying “This may cause some on-orbit noise (?) and radiance biases at very low illumination scenes”. In this case, the SV responses are not reliable for these detectors which will lead inaccurate radiance values. Please explain the potential problems.
Line 432; In Figure 10, why the SNR is slightly lower than trend near the Lmax? Please explain and add in the caption.
Line 449; Have a question about the validity of Equation 6 and this mythology. Usually, the C0 value is expected to be zero in an ideal detector case. When C0 is zero, the C0/C1 fit becomes always zero.
Line 519: Is there any reference for the improper bias voltage issue?
Quality of the Figures from 12~14 can be improved.
Quality of the Figure 18 needs to be improved.
Line 601: It would be nice to have brief descriptions of M11 BRF characterization issues here with some plots.
Line 611: Table 6 needs to be improved. Please change orientation of the table that we can see numbers clearly.
Author Response
Thank you for your thorough review of the paper. I've updated the contents based on your comments and the red text is a response.
At Page1 line 32, I guess the altitude is 829km. Please check and add a reference.
Added the correct altitude and included a reference.
Line 39: ‘that covers’
Added the s to cover
Page3 & line 87, J2 VIIRS uses third order polynomials in SWIR bands. Please add an explanation. Also in equation 4 in page 6 line 177.
Added text to describe a 3rd order term to equation 3 and some words as to how the VNIR and SWIR use different sets. Equation 4 is for the VNIR and equation 10 is the third order fit for the SWIR.
Line 187; Change ‘dnout’ to ‘dnout’ (italic and subscript for out).
Done
Line 210; Please define ‘LSIS’.
Done
Line 218; Why do you need TMCSIS since you have SIS100 with different lamp power levels and attenuation screens? à Answered around 251.
Added that the TMCSIS is a higher radiance source than the SIS100
Line 276; What is the source of SV target. It was not the deep space. In this case, how dark and stable was the source. Please add some explanations. Is there any figure of test bench layout for SIS, TMCSIS, and SV sources?
Added text to describe the SV source on line 191
Added a figure to show the TVAC test source layout.
Line 343; There was an in-depth study on this saturation issues. Please mention follow work.
- Wang, C. Cao, S. Blonski, Y. Gu, B. Zhang and S. Uprety, "An Improved Method for VIIRS Radiance Limit Verification and Saturation Rollover Flagging," in IEEE Transactions on Geoscience and Remote Sensing, vol. 60, pp. 1-11, 2022, Art no. 5403011, doi: 10.1109/TGRS.2021.3097896.
Done
Line 387; Would you add radiance scale image for Figure 9? How’s the dual gain transition? Was radiance values smoothly changed around the point?
Done
Line 422; What do you mean by quantization? I assume that the radiance or dn response stuck at a certain level. Please revise the following sentence to explain this situation instead of saying “This may cause some on-orbit noise (?) and radiance biases at very low illumination scenes”. In this case, the SV responses are not reliable for these detectors which will lead inaccurate radiance values. Please explain the potential problems.
Added text to further explain this issue.
Line 432; In Figure 10, why the SNR is slightly lower than trend near the Lmax? Please explain and add in the caption.
Added test to the figure to identify that as partially saturated data.
Line 449; Have a question about the validity of Equation 6 and this mythology. Usually, the C0 value is expected to be zero in an ideal detector case. When C0 is zero, the C0/C1 fit becomes always zero.
That’s correct. We fit to equation 6 to see how the tau, c0/c1 and c2/c1 behave. Ideally tau would match the vendor specified transmission, c0/c1 would be zero and c2/c1 would capture any sensor response non-linearity. However, the test setup, sources and sensor performance are not perfect and cause those terms to have uncertainties in them. The least squares fit algorithm tries to optimize the fit that includes those test uncertainties. The c0/c1 will absorb some of those uncertainties (non-linearity, quantization, source drift…) and not be zero to get the minimum residuals in the least squares fit. Hopefully the c0/c1 values have uncertainty bars that straddle zero (the term is insignificant) and lets us set it equal to zero.
Line 519: Is there any reference for the improper bias voltage issue?
There’s no official reference. This was a result of the contractor’s internal investigation into the root cause.
Quality of the Figures from 12~14 can be improved.
Added titles to identify the rows and columns in the figures.
Quality of the Figure 18 needs to be improved.
Added titles to identify the rows and columns in the figure.
Line 601: It would be nice to have brief descriptions of M11 BRF characterization issues here with some plots.
Added some text and a reference to a paper that covers the M11 BRF extrapolation.
Line 611: Table 6 needs to be improved. Please change orientation of the table that we can see numbers clearly.
Rotated it and made it bigger on the page.
Reviewer 2 Report
This article is clearly structured and carefully researched. There are two issues that I would like the authors to further clarify as follows.
1. Many published articles mentioned that the reflectance difference between the SNPP/VIIRS and NOAA-20/VIIRS was 2-3% in the calibration verification of post-launch, do authors have the same finding in pre-launch? This kind of study is very meaningful for the traceability of the previous differences between the two sensors.
2. I suggest the authors to review some calibration validation results of post-launch to compare with the results of this paper to see if they are consistent.
Author Response
Thank you for the thorough review of the paper. I've added a few sentences about the bias between the S-NPP and JPSS-1 VIIRS instruments in the uncertainty section. I've also added a reference to the VIIRS bias study for further details. I'm not sure of you wanted a section on the on-orbit performance, but this paper was intended to focus on pre-launch only. That could be added if you think it is an important piece of information to include. The on-orbit performance of JPSS-1/N20 VIIRS has been very good. The RSB radiometric performance has been stable and the only issue I've been aware of is the bias between the two VIIIRS instruments. The pre-launch cause of that has been discussed in an SPIE paper and a journal paper is currently being written for that as well.
Round 2
Reviewer 1 Report
Authors provided reasonable answers and responses to my comments and questions. Thank you.
Reviewer 2 Report
I asked the question to know if the difference between NOAA-20 and NPP comes from pre-flight, which is important for the traceability of the difference between them, and the authors just need to mention it in the article. Looking forward to the authors' subsequent research.
