Lake Malawi’s threshold behaviour: A stakeholder-informed model to simulate sensitivity to climate change


Abstract

Over 90% of Malawi’s electricity generation and irrigation depend on Lake Malawi outflows into the Shire River. Recent lake level declines have raised concerns over future climate change impacts, including the risk of no outflows if the Lake Malawi Outflow Threshold (LMOT) is passed. Addressing calls for model co-production, we iteratively engage stakeholders in data collection, and eliciting local system insights and management priorities, to inform the development of a Water Evaluation And Planning (WEAP) model for the Lake Malawi Shire River Basin. We use a simple model setup and manual calibration to allow for data sparsity and limited documentation of historical management decisions. The model satisfactorily captures limited observed streamflow patterns of Lake Malawi tributaries and lake level variations for the period 1960–2009, however, small errors in lake level simulation significantly affect simulation of monthly outflows. The riparian countries, Malawi, Tanzania and Mozambique contribute approximately 55%, 41% and 4% respectively to lake inflows (1960–2009 average). Forced with 29 bias-corrected global climate model projections (2021–2050) and assuming no change in current operating rules of key infrastructure, the WEAP model simulates wide-ranging changes. These include much higher lake levels that would cause downstream floods, and much lower lake levels, including 11 projections that fall below the LMOT. Both outcomes would have major implications for downstream hydropower and irrigation. Future water management plans require identification and evaluation of strategies that can address multi-year shifts in lake levels and the uncertainty inherent in future climate and hydrological model outputs.

Ajay G. Bhave, Lauren Bulcock, Suraje Dessai, Declan Conway, Graham Jewitt, Andrew J. Dougill, Seshagiri Rao Kolusu, David Mkwambisi. In: Journal of Hydrology, Volume 584, 2020, 124671,
ISSN 0022-1694, https://doi.org/10.1016/j.jhydrol.2020.124671.