I have been trying to use VISAT to reproject MERIS SMILE-corrected radiances, and cloud probabilties to GeoTiff. When I export the 15 MERIS bands, and then import into Idrisi, the radiance values are rescaled to a very large range. However, if I delete all but one band, and export just that one, the radiance values in the GeoTiff file (read in Idrisi) are not altered. I realize that VISAT may select a new data type to accomodate differences among the bands in the product, but in this case, I assume all numbers are the same data type.

Can anyone tell me what I am doing wrong?

Thanks for the response.

The original values are radiances; e.g. band 7 values range from 6.7-249.2. The output as Idris reads it is 0-1,132,015,616.

In this case, I deleted all bands from the SMILE product except 7,8,9 and 10, and exported the product containing only these 4 bands to GeoTiff.

I placed zipped sample files on the web in a freeware filesharing sited called 4shared. You should be able to access them via the links below.

The original smile-corrected (& reprojected to utm) product is in
http://www.4shared.com/file/246284007/71375b68/BERENS090912_smile-correctedre.html

Two exported tifs are in
http://www.4shared.com/file/246282716/3f0f13e6/BERENS090912_exported_tifs.html

One tif contains all 15 radiance bands (but not the flags and detector index bands, because I thought they miight be part of the problem.
The other tif contains only band 15.

When I import the latter tif into Idrisi, I get correct numeric values, e.g. for band 15, values range from 0 - 119.2 mW/sr/m2/nm.

When I import the former file containing all 15 radiance bands, for band 15, values range from 0 - 1122978360.

I also included in the zip file the two Idrisi raster files for band 15.

I hope you can make sense of it!

Thanks, Greg

Thanks for looking into it, Marco.

I will try Idrisi and let you know if they can help me.

Greg