Climate variability modes and their impact on the hydrological cycle - A model evaluation study

Kurzfassung

Climate models are essential tools for studying and projecting climate change and climate variability. A thorough assessment of a model’s ability toreproduce the observed present-day climate is therefore a crucial prerequisite for building up confidence in the model’s ability to project the future climate. This includes modes of natural climate variability such asthe El Niño-Southern Oscillation (ENSO)and the Pacific Decadal Oscillation (PDO)that are spontaneously generated and that need not exhibit the same chronological sequence in models as in nature.The main focus of this thesis is an assessment of representation of ENSOand PDO as well as the impact of these climate modes on cloud parameters and precipitation. For this,different simulations from theMax Planck Institute Earth System Model (MPI-ESM) and the ECHAM/MESSy Atmospheric Chemistry Model (EMAC) are evaluated. These include a coupled fully coupled ocean atmosphere simulation from the MPI-ESM, a simulation with prescribed SSTs from both MPI-ESM and EMAC, and an EMAC simulation which isnudgedtowards realistic meteorology. To evaluate the model simulations, a variety of different observational data setsare used.To evaluate the climate modes,the National Center for Atmospheric Research(NCAR)Climate Variability Diagnostics Package (CVDP) has been implemented into the Earth System Model EvaluationTool (ESMValTool)as part of this thesis. The CVDP computes statistical key metrics of internal climate variability for model simulations and observational data sets. The ESMValTool isa software tool developed by multiple institutions that aimsat improving routine Earth System Models (ESM) evaluation. Additionally, new diagnostics have been integrated into the ESMValToolto evaluate the response of clouds and precipitation to the selected modes of climate variability ENSO and PDO. The results show that the spatial structure of the PDO simulated by the MPI-ESMcoupledis quite realisticand in terms of temporal variability, the MPI-ESM coupled produces a periodic response of approximately the same frequency as observed –that is, periods of 3 –7 years for ENSO, and ~20 years for the PDO. However, the MPI-ESMcoupledmodel only weakly resembles the observed asymmetry between the El Niñoand La Niñaevents, but simulates a quite realistic SST evolution to ENSO. Teleconnections of the atmospheric circulation with ENSO including corresponding changes in temperature, precipitation, and cloud properties are reasonably well captured by the MPI-ESM and EMAC modelsimulations.Systematicdifferences in the performance of the coupled and theatmosphere-only model runs with MPI-ESMand between the free-running and the nudged EMAC model simulationsin reproducing the observed relationships between PDO, ENSO, and local SST changes and cloud parameters and precipitation are rather small.This points to the necessity toimprove model parameterizations of the relevant physical processes such as convection or boundary layer processes.