Towards Development Of A Decision Support System For Water Resource Development In Semi-Arid Micro-Catchments

Abstract

The objective of the paper is to bring together, within a single holistic framework, the principle findings of the research into the physical and human aspects a study in Romwe micro-catchment in southern Zimbabwe. The framework is developed using Bayesian Belief Networks (BBNs) to identify and model the central aspects of both physical and human environment (at both the micro and macro scale), and their impacts on each other. In particular I look at the likely impact of effective catchment management (here taken to be physical and technical interventions) on the probability of livelihoods improving. BBNs offer an exciting new way of bringing together very disparate data sources within useful frameworks, which could then be used not only for decision making about development needs, but also to target better the next generation of research. For groundwater, a sizeable improvement in supply can be achieved by moving from poor to good catchment management under medium rainfall conditions, but not under poor or good rainfall. For improved surface water and improved soil moisture, it is particularly under low rainfall conditions that it is worth moving from poor to good catchment management. It is clear from the results that physical catchment management on its own is incapable of having a major impact on peoples’ livelihoods other than within a very narrow range of parameters. The suggestion is that outside this range, the decision as to whether or not to lake a physical catchment management approach has a negligible impact on water resources and even less on livelihoods. In relative terms a number of factors both extraneous (wider economy, underlying geology), and internal (community cohesion, skills levels) to the community rank equally or more highly in terms of general impact on well being.