Bherney/Ibadan-Surface-Urban-Heat-Island

This project characterises the surface urban heat island of Ibadan, one of the largest cities in West Africa, using Landsat thermal remote sensing for 2025. Working in Google Earth Engine and ArcGIS Pro. # Surface Urban Heat Island of Ibadan, Nigeria (2025) Manuscript, figures, data, and analysis code for a single-year study characterising the surface urban heat island (SUHI) of Ibadan, Nigeria, using Landsat thermal remote sensing. The study retrieves land surface temperature (LST), maps the heat island and its intensity, quantifies the relationship between temperature and vegetation, identifies statistically significant heat clusters, and summarises temperature by land cover class. ## Key findings - Mean LST across the study area was 35.03 degrees C (range 27.84 to 47.60). - The urban core averaged 39.77 degrees C against 34.73 degrees C for the rural ring, a SUHI intensity of 5.04 degrees C. - LST and NDVI were strongly negatively correlated (r = -0.93, R2 = 0.86), with LST = 43.8 - 18.7 x NDVI. - Statistically significant heat clustering (Getis-Ord Gi*) covered 25.1 percent of the study area, dominated by a high-confidence hot core of about 345 square kilometres. - Mean LST by land cover ranked built-up (39.35) and bare land (38.12) warmest, and vegetation (33.60) and water (31.76) coolest. ## Repository structure ``` figures/ All figures used in the manuscript, plus the methodology flowchart and dataset panel data/ Result tables as CSV (SUHI, hot-spot distribution and area share) code/ Analysis code: Earth Engine export and ArcPy/ArcGIS Pro notebook docs/ Supporting documentation and guides ``` ## Methods in brief Data were prepared in Google Earth Engine and analysed in ArcGIS Pro using ArcPy. LST and NDVI were derived from Landsat 8 and 9 Collection 2 Level 2 data for the 2025 dry season, assembled as a cloud-masked median composite. A four-class land cover map (water, built-up, vegetation, bare land) was produced by supervised Random Forest classification of Landsat imagery in Earth Engine. The heat island was measured as an urban-minus-rural difference and along urban-to-rural transects; hot spots were identified with th …