Page Banner

United States Department of Agriculture

Agricultural Research Service

Research Project: MANAGEMENT OF AGRICULTURAL AND NATURAL RESOURCE SYSTEMS TO REDUCE ATMOSPHERIC EMISSIONS AND INCREASE RESILIENCE TO CLIMATE CHANGE Title: Predicting agricultural impacts of large-scale drought: 2012 and the case for better modeling

Authors
item Elliott, Joshua -
item Glotter, Michael -
item Best, Neil -
item Boote, Ken -
item Jones, Jim -
item Hatfield, Jerry
item Rosenzweig, Cynthia -
item Smith, Leonard -
item Foster, Ian -

Submitted to: Soil Science Research Network (SSRN)
Publication Type: Research Notes
Publication Acceptance Date: March 25, 2013
Publication Date: March 26, 2013
Citation: Elliott, J., Glotter, M., Best, N., Boote, K., Jones, J., Hatfield, J.L., Rosenzweig, C., Smith, L.A., Foster, I. 2013. Predicting agricultural impacts of large-scale drought: 2012 and the case for better modeling. Soil Science Research Network (SSRN). Available: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2222269.

Technical Abstract: We present an example of a simulation-based forecast for the 2012 U.S. maize growing season produced as part of a high-resolution, multi-scale, predictive mechanistic modeling study designed for decision support, risk management, and counterfactual analysis. The simulations undertaken for this analysis were performed in December 2012 using weather data up to and including November 30 2012, making it less a forecast of the harvest itself (which was largely completed before this date) than a forecast of the official county-level statistics of the 2012 harvest, scheduled for release on February 21, 2013. The presence of useful predictive information in a zero lead time forecast such as this is a necessary, although obviously not sufficient, condition of a framework’s ability to provide useful predictive information at longer lead times. Droughts and other climate extremes call for a comprehensive approach to monitoring, modeling, and predicting growing seasons globally by using a combination of statistical models, real-time satellite observations, and high-resolution process-based models. Learning from the successes of weather forecasting, researchers need to bring together data and models in a probabilistic framework that leverages real-time data and high-performance computing to improve risk assessment for a range of scales, such as demonstrated here.

Last Modified: 12/22/2014
Footer Content Back to Top of Page