X
PrevPrev Go to previous topic
NextNext Go to next topic
Last Post 04 Mar 2011 08:41 AM by  anon
scale factor file for FLAASH correction of Hyperion image
 1 Replies
Sort:
You are not authorized to post a reply.
Author Messages

anon



New Member


Posts:
New Member


--
04 Mar 2011 08:41 AM
    Hello, I have a L1R Hyperion image that I am trying to atmospherically correct using the FLAASH module in ENVI. I understand that the L1R Hyperion data are in integer data type where the VNIR and SWIR range bands have been rescaled by multiplying by 40 and 80, respectively. The units are W/m2*sr*micrometers I read in the ENVI QUAC and FLAASH module that FLAASH requires the input data in floating point data type with microW/cm2 * nm* sr. My questions are: 1) When I prepare the scale factors that I import to be used in FLAASH, I take into account the unit conversion and 40 and 80 scaling factors used to rescale the L1R data. Since they were multiplied with these numbers, I divide the integer pixel values by 40 and 80 and also apply the unit conversion factor, which is 0.1. As a result, I end up with 0.0025 and 0.00125 for the VNIR and SWIR bands. Then I create an Excel file where I have two columns, one with band number, and one showing the associated scale factor for that band. I save it as a tab delimited text file. Is my scale factor calculation correct and is this the correct way to prepare an Ascii scale file for Hyperion FLAASH analysis? 2) Since the input data are integer type, do I need to do a separate operation to convert the integer data type to floating point or does the use scale factor convert the data type to floating point? I would appreciate the help of anyone out there. Thank you. Regards, Kemal Gökkaya

    Deleted User



    New Member


    Posts:
    New Member


    --
    14 Mar 2011 12:08 PM
    I'll answer your second question first: FLAASH can accept integer data as input, as long as you provide a scale factor that will convert it to floating point values in units of of µW/(cm2 * nm* sr).. You do not need to convert it to floating point first. To answer your first question, I think you have the idea right, but I believe you've missed the fact that FLAASH scale factors are *divided* into the input data to get the floating point radiance values in the proper units. So, I believe you will need to provide the inverse of the values you are currently providing. Unless I have misunderstood what you wrote. The format of the ASCII file you describe sounds like it will work for the input scale factor file. - Peg
    You are not authorized to post a reply.