Measuring the risk of climate variability to cereal production at five sites in Spain

This study provides a method with which to measure the risk of climate variability to agriculture among geographic regions, and analyses the potential risk to crop yields in Spain. Our methodology comprised 3 steps: (1) models were developed for each region to estimate the risk of climate variability, and functional forms were derived from 60 yr of empirical data on wheat; (2) Monte Carlo models were used to analyse in more detail the probabilistic properties of the agricultural yields; (3) a risk factor index was applied to compare among 5 sites. An advantage of this methodological approach is that it links agricultural areas with representative meteorological stations and uses a Monte Carlo approach to define large samples of crop yields that more accurately reflect the statistical properties needed for risk analysis. The methods were robust enough to develop climate and management scenario analysis, which was applied to 5 case studies that exemplify other Mediterranean areas in which climate is a main source of agricultural risk and exerts pressure on limited water supplies, for which there are competing demands. The results show that risk characterization is complex, owing to the multiple attributes of risk beyond climate variability, and that our method of risk analysis facilitates comparisons among locations.