The operational role of remote sensing in assessing and predicting land use/land cover and seasonal land surface temperature using machine learning algorithms in Rajshahi, Bangladesh

• Published in: Applied geomatics Vol. 13; no. 4; pp. 793 - 816
• Main Authors: Al Kafy, Abdulla, Abdullah-Al-Faisal, Al Rakib, Abdullah, Akter, Kaniz Shaleha, Rahaman, Zullyadini A., Jahir, Dewan Md. Amir, Subramanyam, Gangaraju, Michel, Opelele Omeno, Bhatt, Abhishek
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.12.2021; Springer; Springer Nature B.V
• Subjects: Algorithms; Analysis; Carbon dioxide; Climate change; Climate effects; Data collection; Data mining; Developing countries; Earth and Environmental Science; Experts; Geographical Information Systems/Cartography; Geography; Geophysics/Geodesy; Global temperature changes; Heat tolerance; Heatwaves; Land surface temperature; Land use; Landsat; LDCs; Machine learning; Measurement Science and Instrumentation; Neural networks; Original Paper; Parameters; Remote sensing; Remote Sensing/Photogrammetry; Satellites; Spaceborne remote sensing; Summer; Surface temperature; Surveying; Temperature; Urban planning; Urbanization; Winter; Bangladesh; Urbanization; Land surface temperature; Cellular automata; Artificial neural network; Participatory planning
• ISSN: 1866-9298; 1866-928X
• DOI: 10.1007/s12518-021-00390-3

Human activity has boosted carbon dioxide emissions, causing temperatures to rise. The average temperature on Earth is roughly 15 °C, but it has been much higher and lower in the past. There are natural climatic changes, but experts say temperatures are already rising faster than at any other period in history. Unplanned urbanization can sometimes backfire, causing negative consequences that harm the economy and contribute to environmental damages, especially in developing countries like Bangladesh. Because of the strong association between land use/land cover and land surface temperature (LST), the study attempted to analyze and estimate LULC and seasonal (both summer and winter) LSTs using Landsat satellite images at 5-year intervals from 1995 to 2020. Later, the study forecasted both LULC and seasonal LSTs for 2030 and 2040 using cellular automata (CA) and artificial neural network (ANN) algorithms for Rajshahi district. As supporting parameters for determining the magnitude of climate change effects owing to urbanization and temperature rise, primary data collection procedures such as focus group discussions (FGDs) and key informant interviews (KIIs) with experts from diverse sectors were used. Results reveal that the built-up area was increased from 158.22 km 2 (6.64%) to 386.74 km 2 (16.23%) in this 25 years’ timeframe, and it contributed the highest average temperature (41.68 °C in 2020 in summer) comparing with other LULCs. The LSTs were increasing at an alarming rate with 1–2 °C standard deviations per 5 years and maximum temperature was increased from 1995 to 2020 by 37.22 to 42.7 °C) in summer and 22.18 to 28.94 °C in winter. Prediction states that net increase of built-up area will be 2.51 and 5.29, respectively, in 2030 and 2050 from 2020. Maximum LST will likely to be increased to 43.23 °C (2030) and 45.92 °C (2040) in summer, and 30.94 °C (2030) and 31.77 °C (2040) in winter. FGDs and KIIs assessments indicate that frequent LULC change was the main reason for increasing LSTs (71%) and 76% experts agreed that heat waves are the most influencing factors for adverse climate change, among other parameters. The work introduces new methods for integrating remote sensing data with primary datasets, which will provide substantial insights to urban planners and policymakers in terms of participatory and sustainable planning.