MODIS L1B RED/NIR band saturation
Posted: Fri Jan 25, 2019 9:51 am America/New_York
Hello all,I am using MODIS aqua images for water quality modelling, and notice it saturates on 667,678, 748nm L1B bands at highly productive or turbid water. Is it caused by sensor failure or processing algorithms? If it is the latter, any solution to restore valid values in these saturated water pixels?thank you,------ Details -----------I cannot see a full image from L1A product, and I used Seadas OCSSW to process L1A toward L1B, where I see the saturation. I posted previously about this issue but mainly discussed at the L2 product generation. https://oceancolor.gsfc.nasa.gov/forum/oceancolor/topic_show.pl?pid=42964;hl=modis%20aqua%20saturat, and you can see the saturation of a MODIS image (MYD01.A2017266.1830.061.2018163171904.hdf) saturation example at here: [url= //drive.google.com/open?id=1HNCeLSIrLDIOgMEltdE40YdzWnSdCe2j%5b/url%5d ] This is a cloud-free image, so the saturation is caused by cloud is very unlikely. I plot its L1B image band histogram, and notice the saturated value is much higher than all the rest pixel values, almost double the highest value in the rest pixels. It suggests half of the value range is not used, so if that is the case, obviously the saturation is not because it reaches the sensor detector maximum value. (in that case, the highest value in the non-saturated pixels should be very very close the saturated value)