Condensed water path in GCMs and RCMs simulations over Europe

Condensed water path (CWP) is the sum of liquid water path (LWP) and ice water path (IWP). CWP is an essential parameter affecting a wide range of atmospheric processes including cloud properties, precipitation formation and both short and long wave radiation properties of the atmosphere. Therefore, it is desirable to validate this quantity in both global and regional model outputs and to quantify its impact on other climate characteristics. Finally, this knowledge could help to understand differences between CMIP5 and Euro-CORDEX climate simulations as they are discussed among modelers and climate change impact communities and lead to better representation of physical process in climate models. 

We focused on the area of Central Europe and recent climate and its projection according to scenarios RCP 4.5 and SSP-2.45. We have selected all model simulations containing CWP available from Euro-CORDEX initiative and CMIP5 and CMIP6 over Europe. CWP data were then aggregated for each of the 4 distinct ensembles (Euro-CORDEX high resolution, Euro-CORDEX low resolution, CMIP5 and CMIP6) in monthly time step and then compared to observations represented here by ERA5 re-analysis and CLARA 2.1 satellite measurements data. 

We have found that ERA5 and CLARA data agree well with each other. CWP available from CORDEX models and represented as ensemble mean is then significantly underestimated compared to the reanalysis, observation dataset and both CMIP ensemble means. CMIP5 models processed in the same way as CORDEX models then overestimate CWP compared to reanalysis and observations, whilst CMIP6 ensemble offers the best agreement with observed data. The future climate projections based on RCP 4.5 scenarios (CMIP5 and CORDEX models) and SSP-2.45 (CMIP6) models were also assessed. CMIP6 models then project statistically significant loss of CWP over the 21st century, CMIP5 data and CORDEX data project strong increase in CWP. However, values from both CORDEX ensemble means by the end of the century are still not reaching the levels from observed data.