A Bayesian approach towards facilitating climate change adaptation research on the South African agricultural sector

Abstract

Evaluating the range of proposed adaptation measures to combat the sensitivity of agriculture to climate change effects involves evaluating complex interactions between human and natural systems. Integrated strategy-making and implementation in the agricultural sector to reduce the risks posed by climate change requires the consideration of multiple, interdisciplinary factors and the sensitivities of their interrelationships. Lack of information on the sensitivity of agricultural activities to climate change in Africa hampers climate change adaptation research on the region. In water scarce South Africa, the growth of the agricultural sector is threatened by projected decreases in water availability due to climate change. This paper shows how Bayesian networks may be used to facilitate cross-disciplinary participation in elucidating these sensitivities. Bayesian networks provide a graphical framework for mixing quantitative and qualitative information and can be characterised using information associated with varying degrees of uncertainty. This enables a variety of domain experts to test key driver-response interactions through sensitivity analysis and enables visualisation of the complex inter-relationships between inter-disciplinary variables resulting from the impacts of climate change scenarios on South African agriculture. The ability to represent the sensitivities between key variables for which varying degrees of data-scarcity and uncertainty occur provides agricultural sector researchers with a facilitation tool that may helps visualise and formulate climate change mitigation strategies. The results presented here illustrates the extreme sensitivity of water-scarce South African agricultural sector to projected climate change impacts and provides a framework in which tradeoffs between activities can be preliminarily assessed in strategy-making for adaptation.

Keywords:

• Climate change adaptation
• water utilization
• Bayesian approach