Normal 0 MicrosoftInternetExplorer4  I have used two different types of software to atmospherically correct my thermal channels (ATCOR and winvicar). I have then applied emissivity normalization in ENVI. I believe the radiance units for both are mW/m^2/Um/sr so need a data scale factor of 0.001 to get to W/m^2/um/sr? However the emissivity and temperature outputs are very different when I use the scaling factor of 0.001, yet when I leave the data scale factor as 1 (the ENVI default) the results are almost identical. Why would applying the data scale factor make the results different? Should I be applying different data scale factor for each method? I am not sure if this is the appropriate forum for this type of question so suggestions for another forum would also be appreciated.