Spatio-Temporal Flood Risk Modeling for Ugandan Road Infrastructure
The Challenge
An international research project investigating weather impacts on the Ugandan economy needed to quantify the vulnerability of the national road network to flooding, requiring analysis of complex spatio-temporal interactions.
Approach & Execution
Project Overview
As part of a consulting engagement (Nougat.ai) for a UK-based international research project, I focused on assessing the flood risk impacting Uganda’s critical road network. This analysis was vital for understanding potential economic disruptions caused by weather events.
Analytical Approach
- Data Integration: Acquired, cleaned, and spatially aligned diverse datasets including historical rainfall records, satellite-derived precipitation estimates, DEMs to calculate slope and flow accumulation, river networks, and road infrastructure vector data.
- Risk Factor Analysis: Developed models in R to assess flood susceptibility along road segments, considering factors like proximity to rivers, upstream catchment area, slope, historical rainfall intensity/duration, and soil type proxies (if available).
- Spatio-Temporal Modeling: Analyzed the temporal patterns of high-risk conditions to understand seasonal vulnerabilities.
- Visualization & Reporting: Produced maps highlighting road segments categorized by flood risk level and provided summary statistics and methodology documentation for the research team.
This analysis provided crucial, geographically specific insights into infrastructure vulnerability, supporting the project’s overall goals of understanding climate impacts on the regional economy.
Solution Overview
Integrated historical weather data (rainfall, satellite data), digital elevation models (DEM), hydrological features, and road network data. Applied spatio-temporal analysis techniques in R to model flood probability along road segments based on topographical and meteorological factors. Identified and mapped high-risk areas.
Key Results & Impact
- Delivered a detailed geospatial assessment identifying Ugandan road segments most vulnerable to flood-related disruptions.
- Quantified risk based on historical weather patterns and terrain characteristics.
- Provided critical data inputs for the research project's broader economic impact modeling.
- Developed a reproducible analytical workflow for assessing infrastructure risk from environmental factors.