Spatio-temporal dynamics of urban heat island using Google Earth Engine: Assessment and prediction—A case study of Kathmandu Valley, Nepal

Abstract

This study examines UHI dynamics and impacts in the rapidly urbanizing Kathmandu Valley, Nepal, using remote sensing and predictive modeling. The primary goals are to evaluate UHI trends and explore how urbanization influences temperature and climate change. To achieve these objectives, the research investigates the relationship between spectral characteristics, Land Use Land Cover (LULC), and UHI, utilizing high-resolution data from MODIS and Landsat satellites to analyze land surface temperature (LST) and land use changes over recent decades. The study employs Cellular Automata-Markov (CA-Markov) modeling to predict future UHI dynamics, taking into account climatic variability, land use changes, and population growth. Findings reveal significant increases in LST and UHI intensity due to the expansion of impermeable surfaces and loss of vegetative cover. Predictions for 2030 indicate higher LSTs, with winter temperatures ranging from 9.34 °C to 30.12 °C and summer temperatures from 19.74 °C to 42.32 °C, showing an increase compared to 2020. Additionally, the UHI effect is predicted to intensify due to expanding built-up areas, with greater seasonal variation observed in summer. The results suggest that without effective mitigation, UHI will continue to worsen, exacerbating climate-related issues. Insights into the relationship between spectral parameters, LULC, and UHI can guide strategies to mitigate UHI effects, promote sustainable urban growth, and improve urban resilience. Integrating remote sensing technologies with predictive modeling is crucial for addressing urbanization and climate change challenges.