Characterizing spatial and temporal deforestation and its effects on surface urban heat islands in a tropical city using Landsat time series

Abstract

With 1.76 million inhabitants, Mérida city has recently become the most populated city in southeast Mexico, being one of the most attractive cities for investment in real-estate projects. This has increased deforestation to the periphery, resulting in heat islands. We used time series of Landsat images and the BFAST algorithm to analyze annual deforestation around Mérida city over the 2000–2018 period. Land surface temperature was also estimated using Landsat images to compare temperatures before and after deforestation and examine heat islands in the city. The deforestation maps had a 96.82% overall accuracy on average, user’s and producer’s accuracy values of 91.62% and 95.46%, respectively, and an estimated total deforested area of 5413 ha over the study period. Land surface temperature increased in 2.36–3.94 °C after deforestation, and heat islands of varying intensity were detected in 80% of the urban territory, mainly where deforestation occurred. Our results demonstrate the effectiveness of Landsat images and the BFAST algorithm for detecting deforestation in peri-urban spaces, as well as the usability of Landsat images for estimating land surface temperature. These images are effective tools for urban land-use planning and for examining the formation of heat islands in tropical cities.