Review: Assessment of climate change impacts on groundwater-dependent ecosystems in transboundary aquifer settings with reference to the Tuli-Karoo transboundary aquifer

ABSTRACT Climate change effects on groundwater-dependent ecosystems (GDEs) are a function of interaction mechanisms between aquifers (with unique hydraulic properties) and ecosystems. It is currently difficult to determine the magnitude and direction of impacts, and to achieve integrated groundwater resources management; partly because most impact assessments do not incorporate GDEs. This review aims to present existing knowledge regarding the assessment of GDEs resilience to climate change in terms of methods, applicability, and outputs reliability; in reference to transboundary aquifer (TBA) settings. Climate change impacts estimation on groundwater includes future scenarios utilization through Global Climate Models (GCM) to produce atmospheric forcing for hydrogeological models to provide independent recharge estimates. Local level studies are recommended to rectify GCM projections scale coarseness through downscaling techniques available for application towards bridging the spatial resolution gaps, catering for decision-making. Many studies highlight output uncertainties stemming from e.g., the selection of downscaling methods and compatibility with coupling models. Application of a combination of models and different scenarios is recommended to avoid a single non-representative outcome. Ecosystems’ protective measures in most TBAs are country-specific in southern Africa and do not consider neighbouring countries activities while based on low-flow requirements through baseflow, excluding terrestrial GDEs to a large extent. Therefore, a collaborative approach to resource management and research is required among riparian states but the understanding of groundwater–ecosystems–climate change interactions is poor; hence a study was initiated in semi-arid Tuli-Karoo TBA, shared between Botswana, South Africa and Zimbabwe to address this scientific knowledge gap.