Negative Rrs_443 and Rrs_483 for OLI in a clear lake

[oneLaker]

Hi Quinten,

I am generating the Rrs of Landsat-8 OLI in a clear lake (38N, 119W) using the ACOLITE distributed in github. However, the Rrs at 443 nm and 483 nm were negative generated by both DSF and EXP method. At the same time, I used the SeaDAS to process the same image based on the NIR-SWIR (865nm, 1609 nm) method, the results are normal. I found the rho_rc data at 443 nm and 483 nm generated by the ACOLITE were negative as well.

The negative results distributed in many OLI imageries, attached are samples of Rrs_443 of OLI on June, 21 2013 derived by the DSF, EXP, and SeaDAS.

Given that the SeaDAS can generate positive values, the quality of these bands should be good. I do not know what actual factors caused this result, wrong Rayleigh correction, elevation correction (the lake locates on a mountain)?

Could you give me some suggestions?

Thank you!
Zhigang

seadas_nirswir_Rrs443_OLI20130621.png (434.86 KiB) Viewed 8593 times

acolite_dsf_Rrs443_OLI20130621.png (120.6 KiB) Viewed 8593 times

acolite_exp_Rrs443_OLI20130621.png (87.05 KiB) Viewed 8593 times

[quinten]

Hi Zhigang

Thanks for your question! Did you input the elevation of the lake or use a DEM for processing in ACOLITE?

Quinten

[oneLaker]

Hi Quiten,

Thanks for you mention! ACOLITE does have keywords to correct the atmosphere pressure.

I specified the dem_pressure=True, yet the results were still negative. And, the elevation=1946 (in m) and dem_pressure=True did not work as well. Finally, when I used elevation=1946 and dem_pressure=False, the results were positive!

I thought that the dem_pressure=true could automatically read the corresponding elevation of target lake from DEM file to correct pressure, but result of my examination was not. It should be caused by the missing DEM file, as I did not find the SRTM file under data/SRTMGL3.003. How can I get it?

By the way, it would be better to set the dem_pressure=True as the default value in Acolite settings.

Thank you very much!

Zhigang