Quantifying the US Crop Yield in Response to Extreme Climatic Events from 1948 to 2013

Abstract:

The increasingly frequent and severe extreme climatic events (ECEs) under climate changes will negatively affect crop productivity and threat the global food security. Reliable forecast of crop yields response to those ECEs is a prerequisite for developing strategies on agricultural risk management. However, the progress of quantifying such responses with ecosystem models has been slow. In this study, we first review existing algorithms of yields response to ECEs among major crops (i.e., Corn, Wheat and Soybean) for the United States from a set of process-based crop models. These algorithms are aggregated into four categories of ECEs: drought, heavy precipitation, extreme heat, and frost. Species-specific ECEs thresholds as tipping point of crop yield response curve are examined. Four constraint scalar functions derived for each category of ECEs are then added to an agricultural ecosystem model, CLM-AG, respectively. The revised model is driven by NCEP/NCAR reanalysis data from 1948 to 2013 to estimate the US major crop yields, and then evaluated with county-level yield statistics from the National Agricultural Statistics Service (NASS). We also include MODIS NPP product as a reference for the period 2001-2013. Our study will help to identify gaps in capturing yield response to ECEs with contemporary crop models, and provide a guide on developing the new generation of crop models to account for the effects of more future extreme climate events.