VIIRS L2 STRAYLIGHT flag

[josselin.aval]

Hi,

We are currently using the VIIRS L2 data provided by Earthdata for water quality monitoring.

What is the algorithm used to compute the STRAYLIGHT flag? What is the principle?

Thanks in advance for your support.

Best regards,

Josselin

[OB.DAACx - SeanBailey]

For most sensors, the straylight flag is a simple dilation of the cloud and high Lt flags. For VIIRS, a 3x3 kernel is used following the recommendation of Hu et al. (2019)

Hu, C., Feng, L., Lee, Z., Franz, B. A., Bailey, S. W., Werdell, P. J., & Proctor, C. W. (2019). Improving satellite global chlorophyll a data products through algorithm refinement and data recovery. Journal of Geophysical Research: Oceans, 124.

FYI, for SeaWiFS, there was sufficient prelaunch characterization information (and a need based on the sensor response to bright targets) to derive a correction for straylight. Pixels for which the correction is deemed inadequate (those closest to the bright target) are flagged.

Sean

[josselin.aval]

Thank you very much for your quick answer and for the references.

> cloud and high Lt flags

Following https://oceancolor.gsfc.nasa.gov/resources/atbd/ocl2flags/, these correspond to HIGLINT, HILT, CLDICE and SEAICE?

Josselin

[OB.DAACx - SeanBailey]

Just HILT (high top-of-atmosphere radiance, Lt) and CLDICE (clouds...and ice, because the simple radiometric test for clouds cannot distinguish it from bright ice). SEAICE is based on an ice fraction ancillary dataset (which is rather coarse resolution) and is not used in the straylight masking. HIGLINT is also based on an ancillary dataset (well, windspeed from said dataset, and a model). Glint rarely is bright enough at TOA, and when it is it trips the HILT flag.


Sean

[josselin.aval]

Thanks again for your quick answer.

Therefore, the STRAYLIGHT flag does not include any straylight or adjacent light from the coasts (adjacency effects)? Except if such adjacent light from the coast would impact the HILT and CLDICE flags (which seems to be not possible due to the very high intensity required)?

[OB.DAACx - SeanBailey]

Correct, though coasts can be bright enough to trigger the HILT flag. There is no specific adjacency effect flagging or correction applied. Although "stray light" in this context does refer light from outside the direct viewing path, and so should capture any impact from adjacent bright sources.

Sean

[josselin.aval]

Perfect, thanks, this answers my question.