VIIRS FAQ
1. What does
Going forward, we encourage everyone to use the following version:
VIIRS = Visible Infrared Imaging Radiometer Suite
For historical reasons, VIIRS may have been defined differently in various documents, for example:
Visible Infrared Imaging Radiometer Suite
Visible/Infrared Imager/Radiometer Suite
Visible/Infrared Imager Radiometer Suite
Visible Infrared Imager/Radiometer Suite
Visible-Infrared Imager/Radiometer Suite
Visible Infrared Imager Radiometer Suite
While there is nothing wrong with these versions and the variations are small, it's better to use one version consistently going forward.
2. What's the solar irradiance model used in the operations for calibrating the reflective solar bands of VIIRS?
The actual model used was from Modtran 4.3 (Newkur). The original plan was to use the Thuillier 2003 model as documented in the VIIRS radiometric calibration ATBD. However, this did not happen during the transition after launch and it is difficult to switch models at this point because of the potential impacts on EDR products. This issue will be revisited in the future. This issue is discussed in the recent JGR paper by Cao et al. 2013 (Suomi NPP special issue).
3. Is the VIIRS solar band calibration reflectance or radiance based?
The answer is a little more complicated. As you are probably aware that MODIS calibration is reflectance based. Similar to MODIS, the VIIRS reflective solar band calibration is based on the onboard solar diffuser but the calibration equations are very different. For practical purposes, the VIIRS calibration can be thought of as reflectance based. By that I mean that the effect of using different solar models can be canceled out if used properly so the reflectance is derived from the ratio between solar diffuser and solar signals (not affected by the solar model). The radiance on the other hand may have more biases due to the use of solar models. This issue has been discussed in the recent JGR paper by Cao et al. 2013 (Suomi NPP special issue).
4. What's the radiance unit for the
The DNB radiance at night has very small values with a number in the order of magnitude of 10^-9. The unit is W/(cm^2 -sr). Note that unlike other bands, it does not have the per wavelength (um) in the unit. Also, I found that you can make your analysis easier if you simply multiply the DNB night radiance by 10^9, then the resultant numbers are more manageable by most software package. If you do that, the unit simply becomes nano watts per square centimeter per steradian nW/(cm^2 -sr).
5. Why some of the VIIRS bands don't have "RadianceFactors" in the SDR files?
Some VIIRS bands such as M3, 4, 5, 7 and others may not have the "RadianceFactors" in the SDR file because the radiance values are within a reasonable range and do not need to be scaled. In those cases, the values in the file are the actual radiance without a need for any calculations. However, be aware that this is in normal circumstances and should be differentiated from cases where there are anomalies, such as corrupted files. In any case, the radiance values should be reasonable compared to the Ltyp, Lmin, and Lmax for the VIIRS band. If you find values grossly out of range, please report to us.