Agrohydrological sensitivity analyses with regard to projected climate change in Southern Africa.

Abstract

Climate change resulting from the augmented "greenhouse effect" is likely to have significant effects on the terrestrial hydrological system and the social and ecological systems linked to it. Climate change could potentially affect inputs to the agrohydrological system such as rainfall, temperature and potential evaporation; processes within the system such as vegetation dynamics and crop production; and hydrological responses such as runoff, recharge of soil water into the vadose zone and net irrigation demand. This study outlines the use of a daily water budget model, ACRU, and SCENGEN, a climate change scenario generator, to assess potential impacts of global climate change on agricultural production and hydrological responses in southern Africa. This study also considers potential impacts of climate change on plant response which may determine the extent of potential impacts of climate change on agricultural production and hydrological response. Two approaches to climate change impact studies are adopted for use in this study. The first, and more conventional approach considers the impact of a specified climate change scenario, in this case developed with the use of SCENGEN, on the terrestrial hydrological system. The second approach considers the degree of climate change, in this case precipitation change, required to perturb the hydrological system significantly in the various climate regimes found in southern Africa. A comparative analysis of the sensitivity of selected hydrological responses to climate change produced the following results, in ascending order of sensitivity: net irrigation demand < stormflow response < runoff < recharge into the vadose zone. The impacts of a specific climate scenario change on hydrological responses produced unexpected results. A general decrease in mean annual precipitation over southern Africa is predicted for the future with SCENGEN. However, widespread simulated increases in runoff, soil moisture content in the A- and B-horizon and recharge into the vadose zone are obtained. These increases are a product of the CO2 "fertilisation" feedback, which is incorporated as a maximum transpiration suppression routine, in the ACRU model. Net irrigation demand, which is not linked to this routine, is simulated to increase in the future.