Spectral line curvature detection (for pushbroom sensor data)

GENCLUT generates a spectral line curvature correction lookup table (CLUT) for a set of image data. The CLUT is a two-dimensional lookup table of center wavelength and FWHM shift values as a function of image band and along-scanline position. These shift values represent the effect of spectral line curvature on the image data and are used to apply the correction.

GENCLUT undertakes an iterative optimization process to find the shifts involving at each step:

an at-sensor radiance to surface reflectance transformation
smoothing of the surface reflectance data to reduce distortion
a surface reflectance to at-surface radiance reflectance transformation

The goal is to find the shifts where the original and transformed-smoothed-transformed spectra most closely match. See GENCLUT for further details on the algorithm.

The CLUT may be used in ATRLUT during the actual atmospheric transformation of the at-sensor radiance image data set to a surface reflectance data set. ATRLUTwill resample the resample the radiance spectra from the RLUT according to the shifts recorded in the CLUT such that they more closely match the actual (distorted) sensor relative response profiles for different band and across-scanline positions. This will reduce the effect of the spectral line curvature distortion on the transformation results.

SLCCOR is then used to convolve the surface reflectance data set with the nominal relative response profiles of the sensor so that those profiles will be valid over the entire data set.

See GENCLUT or SLCCOR for a detailed application example.