DYNAMICAL MODELLING OF PRESENT AND FUTURE CLIMATE SYSTEM
Report No 1154/1/04

EXECUTIVE SUMMARY

The project on the dynamical modelling of the. climate system builds on earlier work to establish the base capacity to undertake climate simulations. This initiative is in recognition of the maturing of the climate model as a core tool for climate investigation, and for facilitating research on issues that have before proved largely intractable. However, the use of climate models requires great care in order to avoid blind application where the results may be fundamentally flawed. Appropriate implementation of climate models thus requires careful development in order to:

• Understand the inherent strengths and weaknesses of a particular model
• Assess the appropriate model configuration and refine the tool for the domain in question
• Evaluate the model output - assess bias and error and space-time dependent skill

This report outlines key developments addressing these issues. The project has significantly advanced the capacity to apply climate models in a robust manner, and at the same time addresses key questions of concern for the water and other sectors about the regional climate system of southern Africa. The diverse activities of the project are in themselves significant research projects, linked by the common use of the climate model. In the process a vibrant interactive research team has been developed, spanning institutions and involving many more students and scientists than only those who have authored this report.

Within the scope of developing model-based research on the southern Africa climate system. the original project objectives were:

• To continue the capacity building and development of dynamical modelling
• To identify and interpret dynamic process controls
• To develop scenarios of future climate variability
• To interpret the future dynamics of the processes controlling climate variability
• To investigate the dynamics of extreme events through case study modelling.

These objectives have been well met. The capacity developed in terms of infra-structure and scientific personnel has been substantially advanced, and as a direct consequence there are active research lines being followed by post-graduate students and established researchers that now utilize climate models as the principal tool.

The research questions addressed with models have focused on investigating those aspects of the climate system for \which there remains poor understanding. Central among these is the climate sensitivity to feedback mechanisms and the dynamic responses to external forcing - especially land-surface feedbacks to the atmosphere. The modelling studies have shown that the feedbacks are important, and the sensitivity of the climate system to these attributes is commensurate with the magnitude of inter-annual variability and projected climate change.

Furthermore; the results indicate the feedbacks may potentially exacerbate any climate change signal. The suggestion arising from these results is that the land-surface may play a notably more important role than perhaps has been previously thought in governing intra-seasonal variability. extreme events, and possibly the strength of the regional climate response to global climate change.

Complementing the study on climate system feedbacks has been development of the core modelling tools: an activity that tailors the characteristics of the model to optimize performance over the southern Africa domain. As a consequence of this -work the modelling tools are now well understood in terms of their skill and error, while the development of the models is an ongoing and necessary activity. Building on this, the models have been stretched to understand the limits of their performance, and in one application successfully simulated local scale weather systems down to 500m resolution, and in doing so provided valuable insight on extreme events in the Western Cape.

Further to these activities has been the development of a number of new methods of analysis tailored to the needs of the model-based research. These provide valuable and necessary understanding of the models capabilities, especially in the application to climate change questions. In this regard the work has supported a number of climate change activities, and provides core results underpinning research within other complementary projects.

This work has also greatly strengthened the interdisciplinary ties between the core modelling researchers and scientists in other disciplines; notably with scientists at the National Botanical Institute and international1v in the USA and the UK.

Perhaps the most important recommendation that can be made at this stage is that momentum must be maintained! An excellent research team has been developed through the support of this project; a team that has all the requisite skills to make significant advances in our understanding of the regional climate system. Such capacity is a pre-requisite to continue with the important research about the climate system, and which is of notable importance to a broad range of society. Specifically, the following issues are deserving of special attention:

• Investigation of the vegetation and soil moisture feedback mechanisms
• Continued development of the modelling tools
• Coupling of the climate models with other components of the physical system (e.g. hydrological models)
• Development of appropriate means for the dissemination of products to ends users.

This WRC project has contributed to a notable development in research capacity within South Africa. This capacity is well positioned to develop further with direct benefit to the pragmatic needs of many sectors, including that of water resources. As noted by the recent report "Poverty and Climate Change" (African Development Bank et al., 2000, prepared by 10 development and environmental agencies: "climate change poses a serious risk to poverty reduction and threatens to undo decades of development efforts." This is an issue that is possibly the single largest natural long-term threat to development in Africa, and an issue that can only be addressed through climate modelling.