Urban Heat VulnMapping

Urban heat vulnerability mapping for Kochi, India — built for the GIS Mapathon hosted by NEST Digital.

Overview

This project identifies Priority Intervention Zones in Kochi: areas where residents face the highest combined risk from urban heat, based on four overlapping factors:

• High land surface temperature
• Low vegetation cover (NDVI)
• Poor access to hospitals (distance-based)
• Low access to cooling/community infrastructure

By layering these datasets, the analysis flags neighborhoods where heat exposure and lack of supporting infrastructure compound each other — the places that should be prioritized for interventions like tree cover, shaded public spaces, and improved emergency access.

Methodology

1. Boundary data — Kochi's administrative boundary extracted from GADM.
2. Points of interest — Hospital, school, and park locations pulled via QuickOSM (QGIS plugin) within the Kochi boundary.
   • Where OSM returned polygons (e.g. school/park footprints) instead of points, centroids were computed using the QGIS Processing Toolbox and merged with existing point layers.
3. Vegetation index — NDVI computed to represent vegetative cover across the city.
4. Hospital accessibility — Distance-to-hospital raster generated to represent proximity to emergency medical care.
5. Composite scoring — Heat, vegetation, and hospital-distance layers combined into a single raster to derive Priority Intervention Zones.

All spatial processing was done in QGIS.

Data Sources

Layer	Source
Administrative boundary	GADM
Hospitals, schools, parks	OpenStreetMap (via QuickOSM)
Vegetation index	NDVI (satellite-derived)
Building footprints	Not included — see Limitations

Repository Structure

prfvl-urban-heat-vulnmapping/
├── README.md
└── data/
    └── processed/
        ├── hospital_distance.tif       # Distance-to-hospital raster
        └── hospitals_raster.tif        # Hospital point raster

Limitations and Future Work

• Building footprints were planned as an additional layer (finer-grained proxy for built-up density and heat retention) using Geofabrik's Kerala OSM extract, but the file size was too large to process within the competition's time and hardware constraints. This is earmarked as a future improvement.
• Parks and vegetation layers were finalized under time pressure and could benefit from further refinement and validation.
• Priority zone scoring is currently a simple layer overlay; future iterations could explore weighted scoring or machine learning-based classification.

Team

Built as part of a team effort for the GIS Mapathon (NEST Digital). GIS/spatial analysis pipeline built by [prfvl]; StoryMap, presentation and frontend built by [merin-07], [sarj29] and [lakshmisiju1111git].

Tools Used

• QGIS (QuickOSM, Processing Toolbox)
• GADM (administrative boundaries)
• OpenStreetMap data