Estimation of the Necessary Simulation Duration for Assessing Climatic Change Using the Pseudo-Global-Warming Downscaling Method in Snowy Area of Japan

Abstract

The Pseudo-Global-Warming Downscaling (PGWDS) method is a simple way to downscale future climate change using the reanalysis data added by the long-term mean difference between present and future climate data projected by a general circulation model (GCM). The PGWDS method has three substantial advantages as compared to a conventional dynamical downscaling method, (1) removing GCM biases, (2) reducing the amount of required GCM output, and (3) shortening the simulation duration. We have focused on the third advantage and applied the PGWDS method to a local area in Japan in January. The future changes in monthly mean precipitation, snowfall, and surface air temperature estimated by the 10-year averages can be regarded as the future climatic changes estimated by the 30-year averages because most of the 10-year averages are within the standard error of the 30-year averages. Meanwhile, the future frequencies in the 10 years are often larger than the error range of those in the 30 years in the extreme events. Short sampling duration seems to be a primary cause. It is necessary to check the possibility of shortening the simulation duration because the estimation of future change is not appropriate in some cases.