Spatiotemporal Optimization of Mobile Health-Clinic Deployment Under Seasonal Flooding

Seasonal flooding can fragment rural road networks precisely when dispersed communities require dependable primary care. This study develops a two-stage stochastic robust location-routing model for stationing and routing mobile health clinics under uncertain road closures, travel times and service demand. Candidate bases are selected before flood conditions are known; community assignments and vehicle routes adapt after a scenario is observed. Scenario-specific shortest paths are computed on a flood-modified road graph, while a population-weighted spatial accessibility index, unmet-demand penalties, an equity term and conditional value-at-risk control the efficiency–resilience trade-off. A reproducible synthetic case representing a linear rural settlement system contains 24 communities, six candidate bases, 76 road links, three clinics and 12 seasonal flood scenarios. The robust policy selected bases B2, B4 and B5. Relative to stationing optimized for dry conditions, it increased the probability-weighted objective by 3.3% but reduced the worst-scenario objective by 9.0% and the mean of the three most adverse outcomes by 2.8%. All 900 expected weekly consultations remained allocable in the tested scenarios, although the population within 45 minutes of a selected base fell to 89.1% during central, prolonged and bridge-failure scenarios. The framework converts flood forecasts and road-status observations into stationing, routing and accessibility outputs that health managers can update before and during extreme weather. Results are illustrative rather than estimates of any named district and require field calibration before operational use.