Climate data and scenarios: synthesis of recent observation and modelling results, chapter 3.4

3.4 Higher resolution

For many applications, climate changes projected by fairly coarse resolution global climate models may suffice. However, there are applications for which much more spatial detail is necessary. This is particularly true for applications in which a secondary model (such as an agricultural crop model or a basin-scale hydrological model) must be driven by climate model output. In such cases, higher-resolution regional downscaling may be required.

There are two general categories of downscaling: dynamical downscaling, using a regional climate model; and statistical downscaling, using empirical relationships between larger-scale meteorological variables and the local variables of interest. It is beyond the scope of the present document to provide a comprehensive review, and more detail regarding these methods can be found in the literature (see: (Hewitson & Crane, 1996; Murphy, 1999; Wilby & Wigley, 1997; Wilby, et al., 1998; Wilby, et al., 2004; and Schmidli, et al., 2006). However, by way of example, we provide here some results from two Environment Canada resources.