Title: Modelling precipitations fields for agroclimate risk management

Abstract: Managing agroclimate risk often requires reliable estimates of the precipitation indices, and possibly the temperature, over the spatial and time domains. In this talk, we will describe how the precipitation and temperature fields can be spatially interpolated for new locations of interest, as well as temporally backcasted to historical timepoints, using measurements collected over times at several monitoring stations. The relationship between the precipitation and the temperature fields will be examined, aiming to use one factor as a predictor for the other. Finally we will describe how the probability of exceedance (POE), a key element of risk management can be computed using our approach. Alberta climate data maintained by Environment Canada will be used and preliminary results will be presented.