Conterminous United States (CONUS) field extraction

An automated computational methodology to extract cropland fields from WELD processed Landsat 5 TM and Landsat 7 ETM+ time series was developed. In 2010 for all the CONUS a total of 4,182,777 fields were extracted with mean and median field sizes of 0.193 km2 and 0.278 km2 respectively. The results were validated by comparison with field boundaries manually digitized from Landsat 5 and Google-Earth high resolution imagery. The overall per-pixel crop field classification accuracy was 92.7% and the overall crop field producer's and user's accuracies were 93.7% and 94.9%. Comparing all the reference and extracted field objects, 81.4% were correctly matched and the extracted field sizes were on average underestimated by 1.2% relative to the reference field objects.