Predicting spatial and temporal changes in surface urban heat islands using multi-temporal satellite imagery: A case study of Tehran metropolis

Nowadays, predicting spatial and temporal changes of surface urban heat islands (SUHIs) is both a demanding challenge and a subject of interest for researchers working on urban thermal en-vironments. This study aimed to predict the spatial-temporal changes of SUHI in Tehran city. The research data consisted of Landsat multi-temporal images, MODIS-derived water vapor and land surface temperature (LST) products, as well as climatic data. The cellular automata (CA)-Markov model was used for predicting the spatial and temporal land cover changes. Afterward, based on the statistical and spatial information of the impact of land cover changes on LST changes, maps of the LST and SUHII were procured for future years. The results showed that the area of built-up lands increased by 88% from 1985 to 2019. The built-up lands area is expected to increase to 446.3, 463.9, and 480.1 km(2) by 2026, 2032, and 2038, respectively. The mean LST of Tehran city in 1985, 1992, 2000, 2009, and 2019 was 36.45, 41.95, 43.15, 44.95, and 47.25 degrees C, respectively. By 2026, 2032, and 2038, the mean LST may reach 48.65, 49.95, and 51.15 degrees C, respectively. Analysis of the SUHI ratio index in Tehran showed an increase from 0.01 in 1985 to 0.34 in 2019, and is expected to reach 0.38, 0.45, and 0.51 by 2026, 2032, and 2038, respectively.

Automated summary

This study aims to predict the spatial-temporal changes of surface urban heat islands (SUHIs) in Tehran city. The research utilizes Landsat images, MODIS data, and climatic data to predict land cover changes and analyze their impact on temperature changes. The results show an increase in built-up lands area and a predicted rise in average temperature and heat island index in the coming years.