1.04. Impacts of weather events on ontario crop yields

revista MBR

Autores: J. Cabas, A. Weersink.

INTRODUCTION
In order to determine how Ontario farmers can best adapt to any changes in climate, the first stage is to estimate the impacts of climate variability on crop yields. Such predictions can be based on crop biophysical simulation models, such as CERES or EPIC (see Rosenzweig et al.).  An alternative is to use regression models with actual crop yield or profit as the dependent variable and climatic measures as explanatory variables (Newman 1978; Waggoner 1979; Granger 1980; Dixon et al. 1994; Segerson and Dixon 1999).

Regression models have the potential flexibility to integrate both physiological determinants of yield, such as climate, but also socio-economic factors.  For example, Kaufmann and Snell (1997) estimated a hybrid regression model integrating physical and social determinants of corn yield in a way that is consistent with crop physiology and economic behaviour.  They found that climatic variables account for 19% of the variation in corn yield for counties in the US Midwest while social variables accounted for 74% of the variation. The discussion and analysis on climate change effects on crop yield has tended to focus on the effects of predicted increases in average values of climate variables on average crop yields (Adams et al. 1999).  Rather than focus on the mean, others have suggested that the greatest challenge facing the agricultural industry will arise from an increase in the frequency and intensity of extreme events resulting from climate change (CCAF  2002).  There are several important questions that the agricultural industry will therefore have to answer:  What is the relative influence of climatic and non-climatic factors on crop yield?  How sensitive is the inter-annual average crop yield to climate variability?  How sensitive is inter-annual crop yield variability to climate variability? The purpose of this study is to estimate the effects of weather on average yield and the variance of yield for corn, soybeans and winter wheat in Ontario.  The first section of this study reviews the applications of regression models to estimate the impact of climate variability on crop yields in order to develop an appropriate model for Ontario.  The empirical model is then presented along with a description of the necessary data.  The fourth section presents the regression results.