Evaluation of CMIP5 Models in the Context of Dynamical Downscaling over Europe

The quality of regional climate model (RCM) simulations is strongly dependent on the quality of data provided as lateral boundary conditions (LBCs). Frequently, the quality of near-surface variables of general circulation model (GCM) simulations like those from phase 5 of the Coupled Model Intercomp...

Full description

Saved in:
Bibliographic Details
Published in:Journal of climate 2015-07, Vol.28 (14), p.5575-5582
Main Authors: Jury, Martin W., Prein, Andreas F., Truhetz, Heimo, Gobiet, Andreas
Format: Article
Language:English
Subjects:
Atmospheric models
Boundary conditions
Climate
Climate change
Climate models
Climatic zones
Climatology
Correlation coefficients
Datasets
Domains
Free atmosphere
General circulation models
Global climate models
Intercomparison
Mathematical independent variables
Meteorology
Modeling
Modelling
Performance indices
Physics
Quality
Regional climate models
Regional climates
Simulation
Simulations
Studies
Variables
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The quality of regional climate model (RCM) simulations is strongly dependent on the quality of data provided as lateral boundary conditions (LBCs). Frequently, the quality of near-surface variables of general circulation model (GCM) simulations like those from phase 5 of the Coupled Model Intercomparison Project (CMIP5) is analyzed in the region of interest. However, such analysis does not necessarily lead to the selection of high-quality LBCs, as demonstrated in this study. The study region is the European domain of the Coordinated Regional Climate Downscaling Experiment (EURO-CORDEX), where a model performance index (MPI) is used to evaluate the skill of CMIP5 GCMs to reproduce near-surface variables within the EURO-CORDEX domain and free atmosphere variables along its lateral boundaries as a proxy for LBCs used in regional climate modeling. The results suggest that a GCM’s skill in simulating near-surface variables is correlated with 0.62 (Spearman’sr) to its skill in simulating LBCs for regional climate simulations. However, there is hardly any correlation between the performances of different variables, implying that a GCM evaluation solely based on surface parameters or a few variables is inadequate to select suitable driving data for regional climate models. The selection should include the evaluation of all variables passed to the RCM as LBCs in the lateral boundary zone (LBZ) on at least one midtropospheric level.
ISSN:0894-8755
1520-0442
DOI:10.1175/JCLI-D-14-00430.1