Modelling the impacts of climate change and multimodal transport on healthcare access in seasonally flooded regions

Climate change poses an escalating threat to global health, with the most severe impacts threatening low-income, rural communities in the Global South who have contributed least to global emissions yet face greatest vulnerability. Climate change is altering the frequency and magnitude of hydrological hazards, yet little research has examined how floods affect the ability to reach essential services. This thesis developed a coupled climate-flood-health framework to quantify the effects of floods on healthcare accessibility under present and future climates. The framework was demonstrated in the Barotse Floodplain, Zambia, where Indigenous Lozi livelihoods are intertwined with the seasonal floodwave. Climate forcing data were input into hydrological and hydrodynamic models, generating flood extents for a baseline (2000s) and three future decades (2030s, 2050s, 2070s) under three socio-economic pathways (SSP1-2.6, SSP3-7.0, SSP5-8.5). Future floodwaves exhibited significant reductions in flood duration and magnitude, with later flood rises and earlier flood recessions fundamentally altering the shape of future seasonal floodwaves. This increased aridity heightens vulnerability for the Lozi who are reliant on floods for agriculture, fishing, and boat transport. The modelled flood extents were then integrated into a novel geographic health access model, capable of representing the impacts of floods on walking, ox-cart, and boat transportation. Using this approach, the first spatio-temporal, hydrologically dynamic assessment of maternal health access under climate change was produced. Without adaptation, future population growth could quadruple the number of women unable to access maternal care during floods by the 2070s. However, resilient access to all three transport modes mitigated accessibility constraints; the use of ox-carts and canoes enabled ~90% of women to reach delivery sites within a timely two hours, regardless of seasonal or climate-driven flood changes, compared with walking alone permitting only 50% of women to achieve timely access in the dry season and declining further to ~22-26% under present and future peak floods extents. Finally, a sensitivity analysis of floods datasets assessed how different data sources (optical sensing, SAR, and hydrodynamic modelling) affect outcomes, revealing a trade-off between methodological complexity and accuracy in flood representation. Collectively, this thesis established a reproducible, globally transferable framework for evaluating how climate change reshapes health access amid changing flood regimes, a core element of creating climate-resilient health systems.