Array dimension error

[FCarotenuto]

Dear all,

I'm running ATCOR4 and testing different correction routines for some airborne images (e.g.: running ATCOR4 flat terrain, vs. ATCOR4 with DEM and other options). 

Sometimes, apparently randomly, while ATCOR4 is computing the Atmospheric Correction (after having clicked on "RUN Atmospheric correction", after having updated visibility for excessive negative reflectance pixels), it suddenly stops with this IDL Error:

">>> constant0_atm_4r-Error <<<
---------------------------------
Array dimensions must be greater than 0."


I can't understand what's causing it. Sometimes it simply solved by restarting the computer, sometimes it is enough to re-prepare all the DEM files from scratch and repeating the procedure, sometimes, as in this moment, I can't seem to solve it and keeps crashing all my attempts at running the atmospheric correction. 

In attachment there's the log for the atmospheric correction procedure. It mentions the error happening while running APDA1_LUT_CONSTVIS_4R, so something related to visibility I'd say. But it doesn't really help me at all. 

Any suggestion?

SMALL UPDATE: It seems to be tied to the "variable water vapour option", if that's turned on the error does not appear. What could that be?

[reseadmin]

This error may occur during the water vapor retrieval, if one tries to retrieve the water vapor from a spectral region where no image bands are available. So, one needs to check if there are bands in the 940nm or/and the 1130nm absorption region and only use the region which is available for the given sensor - if none is available one can not use the 'variable' water vapor option. In order to catch the error in a more comprehensive way, we'd have to check a sample data set at ReSe Support.

[reseadmin]

This error may occur during the water vapor retrieval, if one tries to retrieve the water vapor from a spectral region where no image bands are available. So, one needs to check if there are bands in the 940nm or/and the 1130nm absorption region and only use the region which is available for the given sensor - if none is available one can not use the 'variable' water vapor option. In order to catch the error in a more comprehensive way, we'd have to check a sample data set at ReSe Support.

Thank you for your quick answer, The point is that the error happens if I do -not- use the variable water vapor option. If I, instead, activate it, the error disappears and processing continue as normal. Also, when the option is activated, the bands are chosen automatically by ATCOR (window 1 is 792.1-792.2, window 2 is 836.7-836.8 and wv absorption region is 817.1-834.0) and they do match a spectral region that's available in my image (it is sampled by a HySpex VNIR sensor with a spectral range between 403.08 to 992.9 nm. The automatically chosen channels on the image (as displayed by the ATCOR4 variable water vapor menu) are 123 for window 1 (792.06 nm) , 137 for window 2 (836.69 nm) & 131-136 (817.57-833.51 nm) for the wv absorption.

[reseadmin]

This error may occur during the water vapor retrieval, if one tries to retrieve the water vapor from a spectral region where no image bands are available. So, one needs to check if there are bands in the 940nm or/and the 1130nm absorption region and only use the region which is available for the given sensor - if none is available one can not use the 'variable' water vapor option. In order to catch the error in a more comprehensive way, we'd have to check a sample data set at ReSe Support.

Thank you for your quick answer, The point is that the error happens if I do -not- use the variable water vapor option. If I, instead, activate it, the error disappears and processing continue as normal. Also, when the option is activated, the bands are chosen automatically by ATCOR (window 1 is 792.1-792.2, window 2 is 836.7-836.8 and wv absorption region is 817.1-834.0) and they do match a spectral region that's available in my image (it is sampled by a HySpex VNIR sensor with a spectral range between 403.08 to 992.9 nm. The automatically chosen channels on the image (as displayed by the ATCOR4 variable water vapor menu) are 123 for window 1 (792.06 nm) , 137 for window 2 (836.69 nm) & 131-136 (817.57-833.51 nm) for the wv absorption.

ok, I'll have to check this with some sample data then. Most probably, there's a problem with how the spectral bands for water vapor retrieval are initialized in the code. I'll contact you off-forum for the further support.