Statistical Downscaling Methods for Climate Change Impact Assessment on Urban Rainfall Extremes for Cities in Tropical Developing Countries – a Review

Abstract

Results of most global and regional climate model simulations cannot be directly applied in future change impacts and adaptation studies of urban drainage and flood risk management. A form of downscaling is required to increase the spatial and temporal resolution of the modelled rainfall data. This paper provides a critical review of the current state of the art statistical downscaling techniques that can be applied to quantify climate change impacts on urban rainfall extremes. Emphasis is placed on delta change methods and Poisson cluster stochastic rainfall models. The paper discusses the applicability and key limitations of statistical downscaling in climate impact and adaptation studies for cities in tropical developing countries. From the review, it can be concluded that simpler statistical downscaling techniques with modest resource requirements such as climate impact sensitivity analyses, use of simple Markov chain or semi-empirical models, construction of climate analogues and spatial interpolation of grid point data are appropriate for scoping of climate impacts and evaluation of mitigation and adaptation strategies at the city scale. Emerging resilience based approaches that combine both scenario based climate model projections and acceptability thresholds defined by key flood risk management stakeholders are promising for application in climate impact and adaptation studies for cities in tropical developing countries.