P44 Sensitivities of parameter tuning and calibration for a convection parameterization scheme in the WRF model

Yan, Huiping, Pacific Northwest National Laboratory (PNNL) and Lanzou University, Yun Qian, Guang Lin, L. Ruby Leung, PNNL, Ben Yang, Nanjing University, and Qiang Fu, University of Washington

In this study, several issues related to parameter tuning and calibration processes in regional climate models are investigated based on the Weather Research and Forecasting (WRF) model. To quantify the response and sensitivity of model performance to model parameters, we identified five key input parameters in the Kain-Fritsch (KF) convection scheme in WRF and gave a bound for each parameter. We performed the parameter tuning and calibration focusing on KF scheme in WRF across different spatial scales, climatic regimes, and radiation schemes. Results show that the optimal values for the five input parameters in the KF scheme are close for all experiments conducted in this study. We found that the model overall performances in simulating precipitation are more sensitive to the coefficients of downdraft (Pd) and entrainment (Pe) mass flux and starting height of downdraft (Ph). Similar trends and variability of the simulated precipitation and skill score to the five parameters can be found across three spatial resolutions, two regions, and two radiation schemes that generate opposite bias in precipitation with their default settings. However, we found that the rainfall bias, which is probably related to the structural error, still exists over some regions in the simulation with optimal parameters, suggesting further studies are needed to target identifying and reducing the model bias or structural errors associated with the missed or misrepresented physical processes and problems with dynamical cores.