Premium
Circulation‐Conditioned Wintertime Temperature Bias in EURO‐CORDEX Regional Climate Models Over Central Europe
Author(s) -
Lhotka Ondřej,
Kyselý Jan
Publication year - 2018
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
eISSN - 2169-8996
pISSN - 2169-897X
DOI - 10.1029/2018jd028503
Subject(s) - climatology , advection , downscaling , environmental science , climate model , cloud cover , flow (mathematics) , atmospheric sciences , climate change , meteorology , geology , precipitation , mechanics , geography , cloud computing , physics , thermodynamics , oceanography , computer science , operating system
We analyze circulation‐conditioned wintertime temperature biases in regional climate models (RCMs) over Central Europe and link them to simulation of cold days. Daily minimum and maximum temperatures from the European Coordinated Regional Climate Downscaling Experiment RCMs are evaluated against reference data from the E‐OBS gridded data set. Flow direction is analyzed through u‐ and v‐wind components at the 850‐hPa pressure level taken from the ERA‐Interim reanalysis. Cloud cover and snow‐area fraction from the same reanalysis are used to associate the temperature biases with physical mechanisms. Distribution of flow directions during cold days is evaluated using efficiency coefficients, calculated as ratios of the relative abundance of flow directions during cold days and their wintertime climatology. RCMs driven by the reanalysis have both positive and negative minimum temperature biases, ranging from +0.7 to −2.7 °C. By contrast, all reanalysis‐driven RCMs have negative maximum temperature biases, especially during northerly and westerly advection, which is probably related to overestimated cloud cover. The too‐cold westerly and northerly flows are associated with an overestimated number of cold days and relatively high values of efficiency coefficients under these directions. Efficiency coefficients are simulated better under southeasterly advection, during which the temperature bias tends to be the lowest. The results suggest that simulated cold events should be analyzed and interpreted with caution, since they may develop also under zonal flow in some models, which contradicts observations.