Using Climate Model Output in Hydrologic Impact Studies: Consequences of Correcting Biases at Different Scales

Abstract:

To estimate the impacts of the ongoing and accelerating climate disruption on societally important infrastructure and ecosystems, the most important primary tools are global climate models (GCMs), which simulate the earth system response to changing atmospheric concentrations of greenhouse gases. A considerable barrier to the use of GCM output for local and regional studies is that the biases in the GCM output must be corrected, and the regional impacts inferred from large-scale simulated climate features. A simple and generally effective method for accommodating systematic biases in GCM output is 'quantile mapping,' which has been applied to many variables and shown to reduce biases on average, even in the presence of non-stationarity. This type of bias correction has been applied at spatial scales ranging from areas of hundreds of kilometers to individual points, such as weather station locations. Since water resources and other models used to simulate climate impacts are sensitive to biases in input meteorology, the temptation is great to apply bias correction at a scale such that the corrected GCM output closely resembles historically observed data, though there can undesirable consequences to doing so, especially where spatial correlation of driving climate variables is important. In this presentation, the trade-offs involved in bias correcting GCM precipitation and temperature, especially at different spatial scales, are explored in the context of simulating extreme high and low stream flows in the Western United States.