CLIMATE-CHANGE SCENARIOS FOR THE ASSESSMENTS OF THE CLIMATE-CHANGE ON REGIONAL ECOSYSTEMS

This paper outlines the different methods which may be used for the construction of regional climate change scenarios. The main focus of the paper is the construction of regional climate change scenarios from climate change experiments carried out using General Circulation Models (GCMs). An introduction to some GCM climate change experiments highlights the difference between model types and experiments (e.g. equilibrium or transient). The latest generation of climate change experiments have been performed using fully coupled ocean-atmosphere GCMs. These allow transient simulations of climate change to be performed with respect to a given greenhouse gas forcing scenario. There are, however, a number of problems with these simulations which pose difficulties for the construction of climate change scenarios for use in climate change impacts assessment. The characteristics of the transient climate change experiments which pose difficulties for the construction of climate change scenarios are discussed. Three examples of these problems are: different climate change experiments use different greenhouse gas concentration scenarios; the ''cold-start'' problem which makes it hard to link future projections of climate change a given calendar year; and the drift of the climate is noticeable in the control simulations. In order to construct climate change scenarios for impacts assessment a method has therefore to be employed which addresses these problems. At present the climate modelling and climate change impacts communities are somewhat polarised in their approach to spatial scales. Current GCMs model the climate at resolutions larger than 2.5 degrees x 3.75 degrees, while the majority of impacts assessment studies are undertaken at scales below 50 km (or 0.5 degrees). This paper concludes by addressing the problems in bringing together these two different modelling perspectives by presenting a number of regional climate change scenarios.