search

olivierhagolle / Start_maja

To process a Sentinel-2 time series with MAJA cloud detection and atmospheric correction processor
Other
51 stars 15 forks source link

running maja in backward mode #13

Closed Giovanni-Sabatino closed 6 years ago

Giovanni-Sabatino commented 6 years ago

Dears

I am trying to run MAJA on the following two products:

S2A_MSIL1C_20170108T095401_N0204_R079_T33UWQ_20170108T095355 S2A_MSIL1C_20170111T100351_N0204_R122_T33UWQ_20170111T100351

obtaining the following error:

2018-07-18T04:01:58.623175 wnode0029.gpod.local maja-processing 01.00 [000000007683] [E] vns::Data::ERROR: ComputeEnvironmentCorrectionImageFilter(0x2461430): The images 'l_IPTOCPtr' (size [10980, 10980]) and 'l_L2EDGPtr' (size [1098, 1098]) have not the same size ! [vnsComputeEnvironmentCorrectionImageFilter.txx:CheckInputs:128]

see here for the complete log file: http://gpod.eo.esa.int/proxy4.aspx?url=http://grid-eo-serv01.esrin.esa.int/gridify-logs/S_153187496758091685373661161833/MAJA/30464/majaproc_L2BACKWARD.log

Indeed I notice that the python script producing AUX data, tuilage_mnt_eau_S2.py, is producing 1098X1098 AUX data raster size, while the input tile is 10980X10980. What can cause this reduction of a factor 10X10?

Thank you. Giovanni Sabatino

Giovanni-Sabatino commented 6 years ago

maybe this information adds something: when I run the python code to create S2 AUX data I notice the message:

e011n44 missing SWBD watr file : e011n44 it is a fully land tile /data/MAJA/aux_data/SRTM_WaterBody_Data//e011n44.gml #############Fichier eau : /data/MAJA/aux_data/SRTM_WaterBody_Data//e011n44.gml gdal_rasterize -burn 0 -l e011n44 /data/MAJA/aux_data/SRTM_WaterBody_Data//e011n44.gml /data/MAJA/aux_data/SRTM_WaterBody_Data//srtm_39_04_tmp.tif Warning : the output raster dataset has a SRS, but the input vector layer SRS is unknown. Ensure input vector has the same SRS, otherwise results might be incorrect. 0...10...20...30...40...50...60...70...80...90...100 - done.

olivierhagolle commented 6 years ago

Hi Giovanni, did we solve your issues (we discussed via email), or do you still need help on that point ? Best regards, Olivier

Giovanni-Sabatino commented 6 years ago

Hi Olivier, forcing the aux data size to be 10980X10980 fixed the issue: we need to add an extra 0

proj=UTM29N EPSG_out=32629 chaine_proj=EPSG:32629 tx_min=0 ty_min=0 tx_max=0 ty_max=0 pas_x=109800 pas_y=109800 orig_x=499980 orig_y=6100020 marge=0

olivierhagolle commented 6 years ago

OK thanks, so I can close the issue. By the way, I am interested to know about your use of MAJA and the results you find (via email, not on github) Best regards Olivier