The atmospheric transformation is performed by the ATRLUT function. This function reads the data set and the RLUT (and, optionally a CLUT (see Spectral line curvature detection (for pushbroom sensor data)) and uses the RLUT to apply either a radiance-to-reflectance or a reflectance-to-radiance transformation to the data set (the former is typically of the greatest interest to users because it produces an atmospheric correction).
The image metadata on nominal location, time, and heading; sensor field-of-view; sensor tilt; band response profile; and DN-to-physical units radiometric transformations are also used by ATRLUT. The nominal location and time are used to compute sun position. Sensor field-of-view and sensor tilt are used to estimate the view zenith angle and relative view-solar azimuth for each data set pixel. Because of this, the transformation is suitable for use with data acquired by wide field-of-view sensors. The field of view of most airborne hyperspectral sensors is wide enough to not allow the view zenith angle and relative view-solar azimuth variation to be neglected in an atmospheric transformation.
ATRLUT also requires data on the elevation, aerosol optical depth, or surface meteorological range, and atmospheric water vapor content for the imaged scene. Nominal values for these parameters that apply to the entire scene may be entered, or, maps for these parameters that give a value for each individual pixel in the image may be used instead. The latter option will result in better transformation results with increasing scene variability.
See the ATRLUT help for a detailed application example.
© PCI Geomatics Enterprises, Inc.®, 2026. All rights reserved.