资源环境科技发展态势分析平台

Precipitation over Southeast Asia is primarily controlled by the Southeast Asian monsoon system. This area features complex orography and morphology, and has limited surface precipitation observations. In this study, a statistical-dynamical downscaling approach that combines weather typing and dynamical downscaling is developed to obtain a high-resolution precipitation climatology for tropical Southeast Asia. A transient simulation with the regional climate model COSMO-CLM (COnsortium for Small Scale MOdelling-Climate Limited-area Modelling Community) driven by ERA-Interim (1979-2008) is performed for the study region. Focussing on Vietnam, six weather types (WTs) are selected for the Indochina Peninsula during the wet season (April to October) using a k-means cluster approach on daily 850 and 200hPa zonal wind components from ERA-Interim reanalysis. The six WTs can be physically interpreted as different stages of the seasonal progression of the planetary-scale monsoon circulation. For each WT, selected representatives from the COSMO-CLM run are dynamically downscaled to a resolution of 0.0625 degrees x0.0625 degrees (approximate to 7km). Using the present-day WT frequencies, the simulated COSMO-CLM representatives at 7km are recombined to a high-resolution rainfall climatology for the recent decades. The resulting high-resolution precipitation climatology is generally able to capture the present-day precipitation estimates derived from APHRODITE (Asian Precipitation - Highly Resolved Observational Data Integration Towards Evaluation of Water Resources) and station data. In spite of systematic biases our approach provides a valuable tool to obtain more robust regional climate change projections for the study area.