统计降尺度方法研究进展综述

统计降尺度方法是将大气环流模式GCMs输出的低分辨率的气象资料转换为流域尺度的主要方法之一，现已发展成为气候学中较为完善的领域。本文简要介绍了统计降尺度方法的基本原理，包括基本假设条件、主要步骤和关键点；重点介绍统计降尺度方法，大致分为转换函数法、天气分析技术和天气发生器这三类，并对几种方法的国内外应用进展做了阐述；对统计降尺度方法的不确定性研究做了简要介绍。指出未来研究应重点研究统计降尺度模型的适用条件及范围、提高降水模拟的精度；统计降尺度与动力降尺度两种降尺度结合的方法将是降尺度主要发展方向之一。 Statistical downscaling method is one of the methods that transform the meteorological data with low resolution GCMs output to the basin scale data, and this method has been studied in depth and widely used in the area of climatology. In this paper, the basic principles of the statistical downscaling method were introduced, including the basic assumptions, main steps, and key points. The statistical downscaling methods can be classified into the transfer function method, weather typing method, and weather generator model, and the applications of each method were also introduced. The uncertainty analysis of the statistical downscaling method was briefly introduced. Moreover, the future study of the statistical downscaling method should focus on its applicable conditions and range and the improvement of precision of precipitation simulation. Finally, it was pointed out that the coupling approach of statistical and dynamic downscaling methods is one of the main development directions of the downscaling study. 国家重点基础研究发展计划（973计划）课题（2010CB428402）；和国家科技支撑计划课题（2012BAC21B02）