Hi Quinten,
I know there is now a modified ACOLITE version (the DSF option) which I've tested and found useful but I recently tested the EXP (SWIR) option and realized its accuracy is affected by wind speed and SZA (image is from Florida) in the 443 and 482 wavebands. I corrected my Rrs for the effect of windspeed and noticed an improvement in my derived product from images with relatively high wind speed (i.e., >= 3 m/s). Similarly, I corrected for the effect of SZA and noticed improvement in images with high SZA. I was wondering if you know/can speculate why.
Could the effects be due to:
SZA: ACOLITE uses a scene centre SZA, but perhaps there is pixel to pixel variations in my image.
Windspeed: imperfect Rayleigh correction or the use of a median Rayleigh-corrected reflectance as opposed to using per pixel. I'm assuming since surface roughness/wind speed is needed for Rayleigh correction, perhaps wind speed value obtained from NCEP for my image is not accurate.
Thanks.