Accuracy, Bias, and Improvements in Mapping Crops and Cropland across the United States Using the USDA Cropland Data Layer

The U.S. Department of Agriculture’s (USDA) Cropland Data Layer (CDL) is a 30 m resolution crop-specific land cover map produced annually to assess crops and cropland area across the conterminous United States. Despite its prominent use value for monitoring agricultural use/land (LULC), there remains substantial uncertainty surrounding CDLs’ performance, particularly in applications measuring LULC at national scales, within aggregated classes, or changes years. To fill this gap, we used state- class-specific accuracy statistics from USDA 2008 2016 comprehensively characterize performance CDL space time. We estimated nationwide area-weighted accuracies specific as well classes non-cropland. also derived reported new metrics superclass within-domain error rates, which help quantify differentiate efficacy mapping (e.g., cropland) among constituent subclasses (i.e., crops). show that aggregate embody drastically higher accuracies, such correctly identifies user’s perspective 97% time greater all years since coverage began 2008. quantified biases throughout these data generate independent bias-adjusted crop estimates, may complement other survey- census-based statistics. Our overall findings demonstrate CDLs provide highly accurate annual measures areas, when appropriately, are an indispensable tool landscapes.