Impact of land use and land cover change on land surface temperature in Iskandar Malaysia using remote sensing technique

Abstract

Iskandar Malaysia is one of the impressive development projects ever undertaken in Malaysia that has been experiencing rapid rate of land use change since 2006. Land use change is due to the urban expansion and reduction in natural green areas resulted from enhanced economic growth. The three objectives of this study are (i) to estimate the land use and land cover changes (LULC) in Iskandar Malaysia from 1989 to 2014, (ii) to investigate the effect of LULC changes on land surface temperature (LST) change in the study area and (iii) to predict the LST by 2025. Remote sensing data namely Landsat (Landsat 5, 7 and 8) and Moderate Resolution Imaging Spectroradiometer (MODIS) of Terra product (MOD11A1) were used to classify various LULC and to calculate the LST in Iskandar Malaysia. There are two digital classification techniques used to classify and test the different LULC in this study area. Maximum Likelihood Classification (MLC) technique provided higher accuracies compared to the Support Vector Machine (SVM) technique. Consequently, the classified satellite images using the MLC technique were used to monitor changes in LULC in Iskandar Malaysia. LST was extracted using mono window. The mean LST using Geographic Information System (GIS) analysis according to LULC shows that water areas recorded the highest night time LST value, while forest recorded the lowest day time LST value. Urban areas are the warmest land use during the day and the second warmest land use during the night time. Moreover, the weighted average used to predict the mean LST of entire Iskandar Malaysia, it was found that if green space increases LST value would decrease by 0.5○C. To predict the effect of LULC changes on mean LST of each LULC types linear curve fitting model was used. According to the results, the mean night LST from 2000 to 2025 will increase in Iskandar Malaysia as urban (20.89°C to 22.39°C±0.45), mangrove (20.88°C to 22.59°C±0.50), forest (20.39°C to 21.04°C±0.18), oil palm (20.39°C to 21.25±0.25), rubber (20.34°C to 22.36°C ± 0.57), and water (21.61 °C to 23.31°C ± 0.51). The results show increment in day time at urban (29.26°C to 32.78°C±1.07), mangrove (26.23°C to 28.82 °C±0.89), forest (25.76°C to 27.54°C±0.49), oil palm (27.02°C to 29.54±0.70), rubber (26.49°C to 27.24°C ±0.29), and water (26.10 °C to 28.77 °C ± 0.8) respectively. Moreover, the relationship between LST and several impervious and vegetation indexes show that there is a strong relationship between impervious indexes and LST, and an inverse relationship between vegetation indexes and LST. Finally, this study concluded that replacing green natural area with improvise surface can increase the land surface temperature and have negative effect on urban thermal comfort.