Projecting and hindcasting potential evaporation for the UK between 1950 and 2099

Abstract

Evaporation estimation is important for the assessment of a wide range of potential impacts of climate change, yet there are significant questions concerning the relevance of alternative methods for climate change studies, and the uncertainty associated with downscaled driving variables. Using principal components analysis, climate variables related to evaporation have been examined; results show significant differences in correlation structures between observed UK data and climate outputs from a Hadley Centre Global Climate Model (HadCM3). Although employing the GCM data directly in the Penman-Monteith combination equation appears to be practical for estimating current potential evaporation, this approach does not project realistic potential evaporation in the 2080s. A local calibration approach is taken to the derivation of an alternative empirical model for estimating potential evaporation based on GCM outputs, using the Generalised Linear Model (GLM) framework. This appears to provide a robust method for impacts assessment. From the GLM projections, the envisaged change in evaporation will be spatially variable across the UK. It is expected that the southern part of the UK will be more sensitive to the change in evaporation than the north. Moreover, in the 2080s, the range (variance) of the monthly potential evaporation appears to change more than the mean.