A new land cover classification based stratification method for area sampling frame construction

摘要

This paper proposes a new automated USDA National Agricultural Statistics Service (NASS) Cropland Data Layer (CDL) based method for stratifying U.S. land cover. The proposed method is used to stratify the NASS state level Area Sampling Frames (ASFs) by automatically calculating percent cultivation at the Primary Sampling Unit (PSU) level based on the CDL data. The CDL based stratification experiment was successfully conducted for Oklahoma, Ohio, Virginia, Georgia, and Arizona. The stratification accuracies of the traditional and new automated CDL stratification methods were compared based on 2010 June Area Survey (JAS) data. Experimental results indicated that the CDL based stratification method achieved higher accuracies in the intensively cropped areas while the traditional method achieved higher accuracies in low or non agricultural areas. The differences in the accuracies were statistically significant at a 95% confidence level. It is concluded that the CDL based stratification method will improve efficiency and reduce cost in NASS ASF construction, and improve the precision of NASS JAS estimates.