Validate 2009 calibration

[mark1]

Specim calibrated the Eagle and Hawk sensors at the start of the 2009 season. We need to validate the calibration as best we can - this will primarily require an independent verification by FSF.

Mismatches in the overlap region between Eagle and Hawk still exist, as in previous years:

[mark1]

Eagle/Hawk spectra using 2009 calibration over water

[mggr]

Comparison of the 2007 vs 2009 calibrations shows significant changes at some wavelengths. As we currently have no independent verification, we can only assume these changes are correct..

Mark: could you add the gnuplot images you made to this ticket?

[mark1]

Percentage difference of data processed with 2009 calibration to 2007 calibration - Eagle

[mark1]

Percentage difference of data processed with 2009 calibration to 2007 calibration - Hawk

[mark1]

An Eagle and Hawk image have been processed with both the 2007 and 2009 calibrations. The difference between the 2007 and 2009 radiance values, for a sample of pixels, has been converted to percentage of the 2007 values, and is graphed below.

[mggr]

We now have internal calibrations and the means to test them. The Specim wavelengths appear correct to within subpixel error. Radiance multipliers are within 5%, which is pretty fair given our measurement error and uncertainty about Specim's setup.

The overlap regions will still have poor calibration as they are very low on signal to noise ratio. We can now quantify this.