Assessing and predicting land use/land cover, land surface temperature and urban thermal field variance index using Landsat imagery for Dhaka Metropolitan area

• Main Authors: Abdullah-Al- Faisal, Abdulla - Al Kafy, Abdullah Al Rakib, Kaniz Shaleha Akter, Dewan Md. Amir Jahir, Md. Soumik Sikdar, Tahera Jahan Ashrafi, Saumik Mallik, Md. Mijanur Rahman
• Format: Article
• Language: English
• Published: Elsevier 2021-08-01
• Series: Environmental Challenges
• Subjects: Land surface temperature; Urban thermal field variance index; Support vector machine; Cellular automata; Artificial neural network
• Online Access: http://www.sciencedirect.com/science/article/pii/S2667010021001712

Climate change is occurring because of an increase in greenhouse gases such as carbon dioxide, methane, and others, which act as a partial blanket for the planet and store solar energy radiation, resulting an increase in land surface temperature (LST). Cities that are already suffering from the urban heat island (UHI) effect , which will withstand the worst of these more extreme heat events. The extent of the impermeable layer and changes in LST are inextricably linked to the severity and commencement of UHI events, which can be measured using the urban thermal field variance index (UTFVI). Land use/Land cover (LULC) change was assessed using support vector machine (SVM) supervised classification, seasonal (summer and winter) LST, and UTFVI variations from Landsat 4–5 TM and Landsat 8 OLI satellite images for the years 2000, 2010, and 2020. Furthermore, in Dhaka, Bangladesh, the cellular automata-based artificial neural network (CA-ANN) algorithm was utilized to forecast LULC, seasonal LST and UTFVI for 2030. From 2000 to 2020, the results demonstrated a large net change in urban areas (+20.52%), whereas vegetation, bare land, and water bodies were all decreased with net changes of -5.72%, -11.19%, and -3.6%, respectively. According to projected LSTs, the net increase in summer and winter temperatures from 2020 to 2030 will be 13% and 20%, respectively, in the highest temperature group (greater than 35 °C). Furthermore, the projected UTFVI showed that in 2030, roughly 72% (up from 58% in 2020) and 69% (up from 47 percent% in 2020) of total area will be covered by stronger and strongest UTFVI zones. Correlation analysis was statistically significant (p value < 0.05), and the relationship between LST and NDBI was strong and positive, but strongly negative with NDVI. The accuracy examination of all the maps revealed a high degree of estimating, with kappa values of more than 80%. By giving significant insights on urban settings and promoting city competency, the study will broaden the perspectives of city planners and policymakers.