Scaling Methods in Regional Integrated Assessments: From Points Upward and from Global Models Downwards

Abstract

Scaling methods in regional integrated assessments are often adopted as givens, when in fact there are a range of methods that each have their strengths and weaknesses. Methods such as a site within a polygon, spatially uniform grids, grids with relational data on polygons, interpolation and stochastic spatial models are reviewed for crop-climate modeling of climate change impacts developed in the European Union's Clivara project. A similar suite of methods for downscaling from global climate models to local conditions exists, and is reviewed. Up- and down-scaling issues relate to availability of data, the level of technical expertise in the project team, validation, uncertainty and risk, stakeholder participation, and modeling of actor-agents. Given the many aims of integrated assessments, no one approach is best.