Speaker: Mingyue Hu, Graduate Student, Department of Statistics, Iowa State University

Title: Estimating Control Total Acres for Desired Geographies Using Cropland Data Layer (CDL)

Abstract: The Conservation Effects Assessment Project (CEAP) evaluates the environmental impact of the conservation program of the United States Department of Agriculture (USDA). In this project, we determine the total acreage of pastureland in specified regions. Previously, we utilized a design-based estimator, which needed the target region’s control total. As the control total is unavailable for new target regions, we create a novel model-based estimator independent of the control totals by utilizing satellite-based cropland data layer (CDL) data as auxiliary information. In this talk, I will show the datasets and the procedure for capturing CDL pixel information near the National Resources Inventory (NRI) samples. For model training, we apply Lasso, LightGBM, SVM and Random Forest approaches. Based on the empirical results, the future plans for this project will also be discussed.