A Weather Generator-Based Statistical Downscaling Tool for Site-Specific Assessment of Climate Change Impacts

Abstract

Abstract. Statistical downscaling approaches are usually used to bridge the gap between climate model outputs and data requirements of impact models such as crop and soil erosion models. This study synthesizes, integrates, and standardizes a statistical downscaling method that was initially developed in 2005 and subsequently evaluated and improved during the last decade. A new downscaling software program, Generator for Point Climate Change (GPCC), has been developed to automate and visualize the method to assist end users with detailed technical and user documentation. GPCC readily generates daily time series of climate change scenarios for local and site-specific climate change impact studies using monthly projections from global climate models or regional climate models. The downscaled variables include precipitation and maximum and minimum temperatures. This software provides a simple but effective climate downscaling tool for assessing the impacts of climate change on crop production, soil hydrology, and soil erosion at a field scale. The tool can also provide an alternative downscaling method to facilitate the international collaborative efforts of the Agricultural Model Intercomparison and Improvement Project (AgMIP) for simulation of world food production and food security assessment. The detailed downscaling methods, their scientific bases, and the advantages of GPCC over other commonly used downscaling methods are presented. GPCC is written in the Matlab language, and a standalone version can be run on Windows XP or above without Matlab software. The tool has a graphical user interface that is simple and easy to generate downscaled climates as well as to visualize downscaled outputs. Each interface tab and key button and their functions are described to facilitate its widespread application. Keywords: Agricultural system models, Climate change impacts, Generator for point climate change, Interface, Statistical downscaling, Stochastic weather generator.