TM ETM+ Rrc results of ACOLITE are negative

[xiaolan1989]

Dear Quinten,

I processed several scenes of Landsat TM and ETM+ data using ACOLITE. When comparing the ACOLITE Rrs with the insitu Rrs, i found the ACOLITE over estimate or under estimate the Rrs. I found a very serious problem, the Rrc out put results for some data were negative (for example: image LT05_L1TP_014033_20110714_20160831_01_T1).
I wonder if this is a normal phenomenal?
I also processed the data ( LT05_L1TP_014033_20110714_20160831_01_T1) using Rayleigh correction algorithm described by Hu et al. (2009), the results is postive.
So if you use the same Rayleigh correction method with Hu et al. (2009) in the ACOLITE?
XING

[quinten]

Hi Xing

Thanks for your comment, the Rayleigh correction for this scene is indeed bad. Negative Rayleigh corrected reflectances are not normal, especially to the degree observed here. As you say this is a serious problem, and I will investigate!

Quinten

[xiaolan1989]

Hi Quinten,
I am trying to calcuate the Rrs using ACOLITE.
I am not sure if this is a serious problem. But if the Rrc are negative, can we trust the Rrs product for this scence?

Xing

[quinten]

Hi Xing

I was agreeing with you that this is a serious problem! Negative Rayleigh corrected reflectances means the Rrs will be bad (negative) as well. I'm assuming something goes wrong with the view and sun geometry for this scene, or the computation of the TOA reflectances, but I need to find a moment to investigate.

Quinten