Cross-calibration of two small footprint sensors

This paper presents a method for determining the calibration of a small footprint satellite sensor using another small footprint system. The work uses the System Pour l´Observation de la Terre (SPOT) 3-High Resolution Visible (HRV) Sensor and Landsat 5-Thematic Mapper (TM) to determine calibration coefficients for TM using HRV given the different look angles, bandpasses, crossing times and footprints. The cross-calibration results are referenced to the reflectance-based vicarious calibration results from October 8 and 9, 1994, from White Sands Missile Range, NM. Although the images are from two different days the atmospheric and reflectance data are very similar and both satellites had near-nadir views. Bands 1, 2, and 3 of HRV, are used to calibrate bands 2, 3, and 4 of TM. The initial attempt of cross-calibration does not account for spectral differences. The calculated calibration coefficient for Chuck Site for TM is within 4% of the reflectance-based values. Several other sites are also used with dark areas providing the worst results. Other bright areas have errors ranging from 1-12%. Areas within the gypsum dunes have results within the 1-4% range. To determine misalignment effects, the TM image is moved relative to the HRV image. This produced changes of less than 1% for up to a 40 meter misalignment