Water Budget over the Tibetan Plateau and Its Dependency on Horizontal Resolution Simulated By an MRI High-Resolution Global Atmospheric Model
Abstract:We use an MRI atmospheric global model (MRI-AGCM3.2) with three different horizontal resolutions (20km:SPA/60km:HPA/180km:LPA) to examine water budget and its dependency on horizontal resolution of AGCM over the Tibetan Plateau (TP) and surrounding mountainous regions at which headwater regions of large rivers are located.
Seasonal cycle of atmospheric water budget shows that MRI-AGCM3.2 has a bias to overestimate precipitation over TP all year round, but its bias becomes smaller as the horizontal resolution becomes finer. Surface evaporation has little dependence on the horizontal resolution. Over western TP, precipitation shows a bi-modal seasonal cycle. In SPA and HPA, the primary (secondary) peak is in Jul.-Aug. (Mar.-Apr.), while two peaks have the same magnitude in LPA. A magnitude of the peak in spring becomes smaller as the horizontal resolution becomes finer. In eastern TP, precipitation has its annual maximum in July. In LPA, a larger positive precipitation bias than that in SPA and HPA is explained by a larger positive bias of moisture flux convergence, which implies that there is more moisture inflow into TP in LPA than that in SPA and HPA. Many of global climate models in CMIP5, whose horizontal resolution is 200km or coarser, has potential not to simulate water budget over TP properly, which may impact on a reproducibility of seasonal cycle of streamflow.