Integration of the Cropland Data Layer Based Automatic Stratification Method Into the Traditional Area Frame Construction Process

A new automatic stratification method utilizing United States Department of Agriculture (USDA) National Agricultural Statistics Service (NASS) geospatial Cropland Data Layers was recently implemented in NASS operations. Recent research findings indicated that using the automated stratification method significantly improved area sampling frame stratification accuracies in intensively cropped areas (>15% cultivation) and overall stratification accuracies when compared to traditional stratification based on visual interpretation of aerial photography or satellite data, while reducing the cost of area frame construction (Boryan et al., 2014). Though the new automated stratification method has improved stratification efficiency, objectivity, and accuracy in the intensively cropped areas, it inherits the Cropland Data Layer classification errors and has lower accuracies in low or non-agricultural areas. This implies that the automated stratification process is not a perfect solution to directly replace the NASS traditional stratification method for area frame construction operationally. This paper describes a hybrid approach: an operational area frame construction process that integrates the automated stratification results with manual editing/review methods. New 2014–2015 NASS area frames for South Dakota, Oklahoma, Arizona, New Mexico, Georgia, Alabama and North Carolina were successfully built using the new integrated operational process. The improvement measures used to assess the traditional, automated and hybrid methods for area frame construction include: 1) area frame stratification accuracy; 2) mean stratum primary sampling unit size, mean stratum percent cultivation and stratum standard deviations; 3) the variances of key estimators; and 4) labor cost. The seven updated area frames delivered significant improvements in objectivity, operational efficiency, and frame accuracy, based on 2013–2015 June Area Survey reported data.