Conterminous United States demonstration and characterization of MODIS-based Landsat ETM plus atmospheric correction

The potential of Landsat data processing to provide continental scale 30 m products has been demonstrated by the NASA Web-enabled Landsat Data (WELD) project. The integration of a recent MODIS based Landsat atmospheric correction algorithm into the WELD processing is described and demonstrated by application to 12 months of conterminous United States (CONUS) Landsat 7 ETM + data. A large volume assessment of the atmospheric correction is presented considering approximately 53 million 30 m pixel locations sampled systematically across the CONUS for December 2009 to November 2010. Monthly 30 m reflectance and derived normalized difference vegetation index (NDVI) data are assessed comparing the top of atmosphere (TOA) and the MODIS-based atmospherically corrected surface reflectance values with respect to spectral, temporal, land cover, and a per-pixel atmospheric correction quality storage scheme. The mean CONUS absolute difference between surface and TOA NDVI expressed as a percentage of the surface NDVI was 28% and the surface NDVI was on average 0.1 greater than the TOA NDVI for vegetated surfaces. The mean difference between surface and TOA reflectance (surface minus TOA) increased monotonically with increasing surface reflectance. On average the change from a negative to a positive mean difference occurred when the surface reflectance was 036, 0.22, 0.17, 0.14, 0.07, and 0.02 for Landsat ETM + reflective bands 1, 2, 3, 4, 5, and 7 respectively. These values are of interest as they depict the average CONUS Landsat ETM + surface reflectance values where the atmosphere has on average no impact and provide the average boundary values for positive and negative atmospheric contributions to ETM + TOA reflectance. The CONUS mean absolute differences between surface and TOA reflectance expressed as percentages of the surface reflectance were 45%, 22%, 12%, 6%, 5%, and 13% for Landsat ETM + bands 1, 2, 3, 4, 5 and 7 respectively. (C) 2013 The Authors. Published by Elsevier Inc. All rights reserved.