Evaluation of uncertainties in regional climate change simulations

Abstract

Abstract. We have run two regional climate models (RCMs) forced by three sets of initial and boundary conditions to form a 2x3 suite of 10-year climate simulations for the continental United States at approximately 50 krn horizontal resolution. The three sets of driving boundary conditions are a reanalysis, an atmosphere-ocean coupled general circulation model (GCM) current climate, and a future scenario of transient climate change. Common precipitation climatology features simulated by both models included realistic orographic precipitation, east-west transcontinental gradients, and reasonable annual cycles over different geographic locations. However, both models missed heavy cool-season precipitation in the lower Mississippi River basin, a seemingly common model defect. Various simulation biases (differences) produced by the RCMs are evaluated based on the 2x3 experiment set in addition to comparisons with the GCM simulation. The RCM performance bias is smallest, whereas the GCM-RCM downscaling bias (difference between GCM and RCM) is largest. The boundary forcing bias (difference between GCM current climate driven run and reanalysis-driven run) and intermodel bias are both largest in summer, possibly due to different subgrid scale processes in individual models. The ratio of climate change to biases, which we use as one measure of confidence in projected climate changes, is substantially larger than 1 in several seasons and regions while the ratios are always less than 1 in summer. The largest ratios among all regions are in California. Spatial correlation coefficients of precipitation were computed between simulation pairs in the 2x3 set. The climate change correlation is highest and the RCM performance correlation is lowest while boundary forcing and intermodel correlations are intermediate. The high spatial correlation for climate change suggests that even though future precipitation is projected to increase, its overall continental-scale spatial pattern is expected to remain relatively constant. The low RCM performance correlation shows a modeling challenge to reproduce observed spatial precipitation patterns.

Evaluation of a WRF dynamical downscaling simulation over California

“…For example, simulations forced by reanalysis (which should have more realistic moisture fluxes) also display positive precipitation biases (Pan et al 2001;Han and Roads 2004;Leung et al 2003;Salathe et al 2008). In particular, Han and Roads (2004) found that relative humidity profiles in their RCM simulation were higher than in the reanalysis forcing it, suggesting that regional model behavior may be playing a role.…”

“…Evaluation of the full RCM+GCM bias has generally been limited to a cursory assessment of the RCM's spatial distribution and seasonal cycle of precipitation and temperature en route to more lengthy discussions of climate change projections (e.g. Giorgi et al 1994;Pan et al 2001;Snyder et al 2002;Leung et al 2004). Exceptions include Leung and Ghan (1999) and Han and Roads (2004), which additionally compare modeled geopotential height and moisture fluxes against reanalysis.…”

Changes in extreme, cold‐season synoptic precipitation events under global warming

“…Analysis of observations covered the same 8-year period as the NCEP-driven run, 1981NCEP-driven run, -1988. Further details of this simulation suite are given by Pan et al [2001].…”

“…The HadCM2 scenario-climate simulation assumed a 1% per year increase of effective greenhouse-gas concentrations after 1990. The period for the scenario-climate was the decade 2040 -2049 [Pan et al, 2001]. Here, we refer to the contemporary and future RCM climates driven by HadCM2 as the control and scenario simulations, respectively, and the climate change is the scenario minus control difference.…”

“…The model computed precipitation using the Grell [1993] convection parameterization and a simplified version [Giorgi and Shields, 1999] of the Hsie et al [1984] explicit moisture scheme. Pan et al [2001] and Gutowski et al [2007] provide further details of the model configuration.…”