Statistical models for simulating extreme yield anomalies

WP1 has developed a set of eight core indices to gauge impacts of critical temperatures and water shortage on crop performance. The set includes two indices addressing drought, two indices quantifying crop damages due to extreme heat and giving probability of hot days over the growing season, and indices for valuating the probability of frost damage at different crop stages. The indices were used to assess the occurrence of critical conditions for growing wheat and maize in two countries, France and Spain, on the basis gridded weather data for 1975-2013 as compiled by the European Commission Joint Research Centre.
Indices and climate variables averaged over the monthly or longer time scale were used to develop statistical models for simulating extreme yield anomalies. Logistic models were found particularly suitable for describing extreme yield anomalies in terms of combinations of indices and climate variables. Such models can be used either as prediction tools or to post-process of operational yield forecasts.

Posted in Science briefs.