SheRules

Hotspot Detective is an autonomous geospatial diagnostic system that transforms passive satellite heat maps into causally-validated municipal cooling plans. In arid megacities like Jaipur, Rajasthan, summer surface temperatures routinely exceed 48 °C to 50 °C. Conventional satellite dashboards identify hot areas but cannot explain why they overheat or what specific interventions will produce cooling without unintended side effects. Traditional regressions suffer from geographic confounding, while full microclimate simulations require weeks of manual modeling per block. Hotspot Detective addresses this by partitioning Jaipur (1,018 km²) into 11,312 uniform 300 m cells. It integrates Landsat 8/9 daytime Land Surface Temperature (baseline median 48.19 °C), Sentinel-2 10 m land cover (built-up density, tree canopy, bare ground, turf, water, NDVI), Copernicus GLO-30 elevation, OpenStreetMap hydrological vectors, and WorldPop demographics. Hotspots are ranked by population-weighted excess heat ( Excess Heat ×Residents). An autonomous dual-agent system then diagnoses each area: the Lead Investigator (muse-spark-1.3) identifies anomalous factors and executes matched-area counterfactual tests across Jaipur’s database—holding elevation, distance to water, and baseline density constant—to isolate empirical cooling effects with Student's t-test significance (p<0.001). Hypotheses with fewer than 8 comparable areas are labeled 'Not testable'. The agent synthesizes a localized municipal Action Plan - such as native shade corridors (Neem, Khejri, Amaltas), high-albedo cool roofs (SRI > 78), or xeriscaping. Finally, the Scientific Critic (muse-wonder-1.0) audits findings against microclimate physics and flags collinearity traps before presentation. Delivered via a high-density, calm analyst interface (Linear/Observable aesthetic) with dual Diagnostics and Telemetry dashboards, Hotspot Detective enables municipal authorities to justify data-backed cooling investments in minutes.