Evaluating Evaporative Demand in CMIP5 Models and its Role in Characterizing Drought over the U.S. Northern Great Plains

Abstract:

Drought is a dominant climatic feature affecting socio-ecological systems across the U.S. Northern Great Plains. While drought conditions primarily occur in response to periods of low precipitation, they can be exacerbated by enhanced evaporative demand (E₀) during periods of elevated temperatures, net radiation, advection, and/or decreased humidity. Our understanding of future drought processes in this region is intrinsically linked to the uncertainty in climate models’ representation of E₀ in both historical and future simulations, and this in turn depends on how we choose to calculate E₀. While it is generally understood that physically based formulations of E₀ such as Penman-Monteith (PM) more appropriately account for the influence of radiative and advective forcings, temperature-based formulations such as the Hargreaves and Thornthwaite equations are often preferred for ecological and hydrological applications due to their simplicity and our higher confidence in input parameters. Here, we assess differences in E₀ representation across CMIP5 global climate models using physically and temperature-based formulations of reference evapotranspiration (ET_r; i.e., evaporative demand as modeled using a well-watered reference crop). Using daily outputs from the historical runs of selected CMIP5 models, we compare seasonal and annual ET_r trends from the PM and Hargreaves formulations over the CONUS and validate these against observational data (U.S. Class-A pan data). Next we use the PM-ET_r to compute the Evaporative Demand Drought Index (EDDI) and evaluate EDDI at weekly and monthly scales in historical simulations over the U.S. Northern Great Plains. We attribute variability, trends, and drought anomalies to each of its driving parameters, to tease out the influence of specific model biases and evaluate geographical nuances of ET_r drivers. Our findings will later serve as baseline for the evaluation of drought characteristics in simulations of future climate.