A Multiresolution Approach to Estimating the Value Added by Regional Climate Models

Abstract

limate models have emerged as an essential tool for studying the earth?s climate. Global models are computationally expensive, and so a relatively coarse spatial resolution must be used within the model. This hinders direct application for many impacts studies that require regional and local climate information. A regional model with boundary conditions taken from the global model achieves a finer spatial scale for local analysis. In this paper the authors propose a new method for assessing the value added by these higher-resolution models, and they demonstrate the method within the context of regional climate models (RCMs) from the North American Regional Climate Change Assessment Program (NARCCAP) project. This spectral approach using the discrete cosine transformation (DCT) is based on characterizing the joint relationship between observations, coarser-scale models, and higher-resolution models to identify how the finer scales add value over the coarser output. The joint relationship is computed by estimating the covariance of the data sources at different spatial scales with a Bayesian hierarchical model. Using this model the authors can then estimate the value added by each data source over the others. For the NARCCAP data, they find that the higher-resolution models add value starting with low wavenumbers corresponding to features 550 km apart (or 11 total 50-km grid boxes per cycle) all the way down to higher wavenumbers at 150 km apart (3 grid boxes per cycle).