Monitoring and forecasting heat island intensity through multi-temporal image analysis and cellular automata-Markov chain modelling: A case of Babol city, Iran

An urban heat island is one of the most vital environmental risks in urban areas. The aim of this study was to assess the spatial-temporal patterns of land use changes and heat island intensity for the city of Babol, Iran, between 1985 and 2015 and to predict likely future heat island intensity variations. Multi-temporal Landsat images were acquired and analysed in this study. First, single channel algorithms were used to calculate the land surface temperature (LST), and a Maximum Likelihood Algorithm was utilized to classify images. Second, land use changes (LUCs) and LST were examined, and the relationship between the fractional vegetation cover (FVC) and land-use changes was analysed using the normalized land surface temperature. By using the mean and the standard deviation of the normalized thermal images, the area was divided into five thermal categories: very low, low, medium, high and very high. Then, by applying the heat island intensity index, the heat island changes in the studied period were investigated. Possible future land use changes were investigated using a cellular automata-Markov model, and the heat island intensity changes were anticipated. The results indicate that the area of built-up land increased by 92% between 1985 and 2015, and that the area of agricultural land noticeably decreased. The built-up land changes trend has an inverse relationship with the trend of FVC changes and follows the same trend as the normalized surface temperature changes. Most changes in the surface temperature of the area are located within 0-800 m of a built-up area. The main reason for these changes could be the conversion of agricultural and green space land areas into built-up land. The largest area of the temperature categories in all years is the medium temperature category, which covers the suburban land areas. The low- and very low temperature categories, which follow a decreasing trend, are related to land areas far from the city. In addition, the high- and very high-temperature categories, whose areas increased annually, are adjacent to the city core and the exits from the city. The average surface temperature in all land uses increased during the studied period. Nevertheless, the rate of temperature rise is higher in the built-up uses. The index ratio of the heat island during this period shows an increasing trend, and its value changed from 0/5 in 1985 to 0/67 in 2015. An increase in the heat island intensity has a direct relationship with the area population growth and, thus, the increase in the built-up land area. Anticipation of land use changes and the process of heat island intensity variations for the studied area show alarming results that call for decision-makers to address this important challenge.