Incorporating Orographic Effects on Precipitation into Downscaled GCM Output

Abstract:

General Circulation Models (GCMs) have promise for investigating the potential for natural hazards, such as landslides and floods, under a range of climate conditions. However, even downscaled output from GCMs can be too coarse to account for subgrid-scale variability in rainfall, especially in mountainous areas where orographic effects can enhance rainfall intensities to levels that trigger landslides. In order to investigate the potential for landslides, orographic enhancement of rainfall needs to be incorporated into downscaled rainfall estimates. Here, we describe one approach to account for orographic enhancement of rainfall estimates from dynamically downscaled GCM output. We use a stochastic simulation algorithm that generates site-specific rainfall time series conditioned on downscaled output. We demonstrate our approach by using this algorithm to generate rainfall time series at a 3-hour time step for a portion of Marin County, California over the time period 1990-2013. The conditioning data, which includes temperature, wind speed, and precipitation, were derived from output of the RegCM3 regional climate model at a spatial resolution of 15 km and a temporal resolution of 3 hours. We generated rainfall time series corresponding to several established rain gauge sites located from sea level to over 400 m in elevation to span a range of rainfall enhancement due to orographic forcing. This technique is applicable to areas with a sufficient rain gauge record, and will allow us to use downscaled output to investigate the potential for landslides under future climate conditions.