Integrating Survey Data with Auxiliary Sources of Information to Estimate Crop Yields

• Activity Number: 624
• Type: Invited
• Date/Time: Thursday, August 13, 2015 : 8:30 AM to 10:20 AM
• Sponsor: Survey Research Methods Section
• Author(s): Nathan Cruze*
• Companies: USDA/NASS
• Keywords: Bayesian hierarchical models ; composite estimation ; model-based estimation ; survey sampling

Abstract

The USDA's National Agricultural Statistics Service conducts multiple surveys for major crops, including winter wheat, during a growing season. These surveys are designed to capture the current status of crops at state, regional and national levels. Each of the surveys also provides an indication of end-or-season yields. A Bayesian hierarchical model gives improved yield forecasts by combining these indications from three different types of surveys together with auxiliary data. Modeled state forecasts are benchmarked against a regional forecast and rigorous measures of uncertainty are provided. Advantages of this model are the flexibility for inclusion of new sources of auxiliary information and the incorporation of expert judgment while retaining reproducibility and estimability of standard errors. The model for winter wheat is shown to perform well over a wide variety of conditions, as illustrated by data from the 2012 crop year.