ArcMap: Difficulties performing Apparant Reflectance function upon Landsat Data

4220
4
02-03-2015 06:22 PM
JohnFullmer
New Contributor

Hi everyone, I'm trying to perform an apparent reflectance function upon some Landsat 7 ETM+ Level 1 Product rasters in ArcMap 10.2. I have been following the steps found at Apparent Reflectance Raster Function | ArcGIS Blog .  However the function is not working, and it appears to be because ArcMap isn't picking up the metadata from the file. Has anyone experienced this problem or have an idea what I may be doing wrong?

Tags (3)
0 Kudos
4 Replies
DanPatterson_Retired
MVP Emeritus

The link is buried unfortunately.  Can you elaborate on it not working?  Are you getting error messages or unexpected results?

0 Kudos
JohnFullmer
New Contributor

Oops, though the link is just the second result that appears when you google "Arcmap Apparent Reflectance" I get no results at all. I was guessing it was not reading the Metadata because the gain and bias fields in the formula window were not automatically populated. I've tried manually adding numbers to those categories, but then I press ok in the formula box, then ok in the template editor, and then nothing happens. If I reopen the template editor the apparent reflectance formula that I had added before is no longer there.

0 Kudos
DanPatterson_Retired
MVP Emeritus

What a circular description...but from the help file ... your conjecture may be correct, but there is no guidance on how to 'read' the metadata file.

All of the information required for the correction is extracted from the key metadata properties for each image when the function is initialized.  In order to be eligible for this function, an image must have the following properties:

  •                 Acquisition date and sun elevation for the dataset
  •                 Gain, bias, and sun irradiance for each band
0 Kudos
JayantaPoddar
MVP Esteemed Contributor

Hi John,

The values in each calculation might be taking integer values. Convert the values in each calculation to float.

e.g.: if a resultant of (x/y) = 0.3, it might be getting rounded upto its nearest integer, i.e. 0. Thus the issue is arising.

Thanks,

Jay



Think Location
0 Kudos