Evaluation of the South Asian Monsoon in the ECHAM/MESSy

Kurzfassung

The South Asian Monsoon has been evaluated in the ECHAM/MESSy Atmo-spheric Chemistry (EMAC) model in comparison to observations and models par-ticipating in the 5th Phase of the Coupled Model Intercomparison Project (CMIP5).Existing diagnostics of the Earth System Model Evaluation Tool (ESMValTool)have been applied to historical CMIP5 simulations and to an EMAC timeslice ex-periment representing the year 2000 to investigate precipitation in the South AsiaMonsoon period. The EMAC simulation generally overestimates precipitation inSouth Asia during the local summer and winter seasons compared to observations,and also the global precipitation intensity, which is calculated as summer minuswinter difference. This bias in precipitation intensity results in an overestimationof the global monsoon domains, which are areas that exceed a precipitation inten-sity of 2.5 mm per day. Compared to the CMIP5 historical simulations, not only theCMIP5 multi-model mean but also each individual model is more skillful in termsof pattern correlation index. Possible reasons for biases in the EMAC simulationcould for example arise from biases in circulation or the representation of clouds.The simulation of the 850 hPa low-level wind was better represented than the pre-cipitation intensity, although some biases compared to meteorological reanalysisexist. Diagnostics have been newly developed and implemented into the ESMVal-Tool to examine the representation of Cloud Radiative Effects (CRE) as externalsolar radiation is the main driver for monsoon. Deviations from observations arefound in particular for the simulated shortwave radiation, which could be relatedto problems in the representation of low-level clouds. The biases found in theparticular EMAC simulation evaluated here could partly be due to the boundaryconditions used, in particular the prescribed sea surface temperatures (SST) thatare taken from a CMIP5 model simulation with the CMCC model. In addition, acoupled ocean could possibly partially compensate for the biases in precipitation.Additional work is required to investigate this further. For example, an EMACsimulation with observed SSTs or a coupled ocean could be evaluated to betterunderstand the reasons for the biases found here