Climate science has generated unprecedented data and curated complex scientific tools for its analysis, yet this study addresses the persistent gap stemming from poor science communication and public disengagement. It validates a novel, readily accessible framework using Generative AI (GAI) as a supportive and facilitative tool to synthesize code for Google Earth Engine (GEE) data extraction and R statistical analysis, democratizing climate inquiry for non-experts. The methodology was applied to Kolkata, India, an exposed and vulnerable metropolitan area exhibiting notable policy deficits. Historical analysis (1990–2024) revealed a significant, long-term warming trend, particularly the Land Surface Temperature (LST) (0.0752 °C/year) strongly associated with rapid urban expansion (19.02 km2/year), which intensified the nocturnal Urban Heat Island (UHI) effect (Night-time UHII rose by +0.60 °C). This data quantifies the public’s lived experience of warmer winters, diminishing nocturnal cooling and accelerated flood risk from concentrated rainfall intensity. Future projections based on CMIP6/CanESM5 and derivative illustrations suggest that mean annual temperature may rise by up to 3 °C and project a critical 55.8% probability of late-monsoon months becoming an extreme rainfall month by 2030. The study re-affirms that translating scientific data into clear, validated, citizen-centric narratives is essential for transforming passive beneficiaries into proactive enablers. The proposed scalable framework offers a replicable model for developing community-informed, anticipatory climate action and driving structural improvements in urban disaster risk reduction (DRR).
Sil et al. (Mon,) studied this question.