Long-term Landsat 5 and 7 Reflectance Inconsistencies Caused by Landsat Satellite Orbit Drifts

Abstract:

Keywords: Landsat, long term data record, orbit drift

The Landsat satellite series provide the longest temporal record of space-based earth observations and, with their free data availability, the systematic generation of consistent Landsat time series products has been advocated. The Landsat 5 and 7 satellites were launched into nominally the same orbits but temporally sparse station keeping maneuvers meant that their orbits drifted over the satellite mission lives with local crossing times varying differently between sensors and by up to 0.92 hours (Landsat 5 1982 to 2012) and 0.21 hours (Landsat 7 1999 to 2012). Consequently, their images were acquired with temporally variable solar zenith angles. Long-term Landsat 5 and 7 reflectance inconsistencies may be introduced by orbit drift induced solar zenith variations combined with surface reflectance anisotropy. The majority of terrestrial surfaces reflect optical wavelength radiation anisotropically with a directional dependence that varies as a function of the sun–target–sensor geometry, commonly described by the Bi-directional Reflectance Distribution Function (BRDF). This study quantifies the overpass time and observed solar zenith angles for all the Landsat 5 and 7 images available in the Landsat archive along an approximately north-south Landsat path over the Conterminous United States. The impact of observed solar zenith angle variations on red and near-infrared nadir view reflectance and on the derived normalized difference vegetation index (NDVI) with respect to different Moderate-Resolution Imaging Spectroradiometer (MODIS) BRDF land cover types is modelled. Results show that the 31 year Landsat 5 solar zenith variations and the 14 year relative Landsat 5 and 7 solar zenith differences vary latitudinally (up to 10º and 4º respectively) and impose small but significant reflectance and NDVI variations that should be minimized before long-term time series application.