2020. augusztus 13. csütörtök
IDŐJÁRÁS - OMSZ angol nyelvű folyóirat

Vol. 123, No. 4 * October - December 2019

Quarterly Journal of the Hungarian Meteorological Service

Special compilation: Environmental challenges – Smart solutions

letöltés [pdf: 3546 KB]
Validation of RegCM regional and HadGEM global climate models using mean and extreme climatic variables
Ildikó Pieczka, Judit Bartholy, Rita Pongrácz, and Karolina Szabóné André
 PDF (3932 KB)   |   Abstract

The horizontal resolution of global climate models (GCMs) is still too coarse to evaluate regional climatic differences, therefore, to analyze regional environmental changes, it is essential to downscale the GCM simulation results. One of the methods widely and most often used for this purpose is dynamical downscaling. In the present paper we examine the ability of a specific global (HadGEM2-ES) and a specific regional climate model (RegCM) to describe present climatic conditions in different geographical areas within the Med-CORDEX domain. Our main goal with this validation is to inform researchers, who are planning to complete climate change impact studies about the major characteristics of the simulation outputs, serving as important input in such studies. So we analyzed annual and seasonal mean fields, mean error fields relative to the reference measurements, and selected climate indices. On the basis of the results, dynamical downscaling generally cools the HadGEM results, which depends on the distance from the ocean, and orography. A clear improvement can be recognized in the root-mean-square error (RMSE) of temperature indices when using finer resolution. Moreover, dynamical downscaling with higher resolution often increases the precipitation in mountains. Furthermore, in order to quantify the potential added value of RegCM simulations, a complex measure was introduced to take into account both the magnitude and spatial extent of bias. The analysis shows a general improvement in the cold-related indices in Central Europe and all temperature-related climate indices in Western Europe. The influence of model resolution is usually so strong, that it results in the underestimation of precipitation indices changing into overestimation and vice versa.

IDŐJÁRÁS folyóirat