Sensitivity of Projected Changes in Water Availability to Potential Evapotranspiration (PET) Estimates over Western Canada

The standardized precipitation and evapotranspiration index (SPEI) is a comprehensive measure of climatic water balance and availability in a region. However, the validity of computed SPEI values at the different scales depends on the reliability of the corresponding potential evapotranspiration (PET) estimates. This study investigates the sensitivity of 3 - and 12 - month SPEI (SPEI-3 and SPEI-12) over western Canada to the PET estimations using the Thornthwaite and Hargreaves methods. Both historical and  projected future changes in water availability under the RCP4.5 and RCP8.5 emissions scenarios are examined, the latter  using statistically downscaled, high resolution climate data generated by six Global Climate Models (GCMs) from the latest Coupled Model Intercomparison Project (CMIP5). Results indicate that, compared with the Hargreaves method, the Thornthwaite approach provides lower PET values over the southern and western part and higher PET values over the northern and north-eastern region of western Canada. This is also reflected in the climate change scenarios where the Thornthwaite method produced relatively uniform increases in PET over the region while the Hargreaves approach shows relatively higher PET increases in the southern half of the study region compared to the north.  As a result, the Hargreaves method produces relatively higher future water deficit (decreases in average SPEI) in the southern and western parts and relatively higher water surplus (increase in average SPEI) in the northern and north eastern portions of the study region. The reasons behind these reginal differences and their implications with respect to the water availability estimates are the subjects of an ongoing investigation.