Assessing scale effects for statistically downscaling precipitation with GPCC model

摘要: The resolution of general circulation models (GCMs) is too coarse to assess the site-specific impacts climate change. Downscaling approaches have been developed meet this requirement. As model increases, it imperative know whether finer regional (RCMs) would result in any improvement statistical downscaling quality at station scale for particular methods. objective study effects resolutions on precipitation using generator point change (GPCC) model. conducted across three scales, from GCM, and mid- high-resolution RCMs a two Canadian stations Quebec province. Observed gridded corresponding scales also studied parallel, totalling six experiments. results show that statistics downscaled are somewhat overestimated Sept-Iles all experiments with relative error mean daily, monthly, annual ranging between 1.1 5.0%, 2.7 5.9% 0.5 6.2%, respectively, but underestimated Bonnard −4.1 −10.2%, −3.2 −13.3%, −2.6 −12.0%, respectively. number wet days per year well preserved difference observed data −2.9 6 d stations. similar observations model-simulated both Furthermore, there no noteworthy effect when output GPCC, indicating regions without RCM projections, high-quality daily series can be derived directly GCM projections © 2013 Royal Meteorological Society