Prediction of seasonal urban thermal field variance index using machine learning algorithms in Cumilla, Bangladesh

https://doi.org/10.1016/j.scs.2020.102542Get rights and content

Highlights

  • Patterns of LULC change and seasonal LST shift in Cumilla city were analyzed.

  • Reduction of vegetation cover significantly increase the UHI effect in the city.

  • The cross- tabulation better explains the relationship between LULC vs UTFVI.

  • Seasonal UTFVI prediction demonstrate gradual decrease in overall thermal environment.

  • Predicted UTFVI vs LULC demonstrated the highest UTFVI concentration in urban area.

Abstract

The intensity and formation of urban heat island (UHI) phenomena are closely related to land use/land cover (LULC) and land surface temperature (LST) change. The effect of UHI can be described quantitatively by urban thermal field variance index (UTFVI). For measuring urban health and ensuring sustainable development, the analysis of LST and UTFVI are receiving boosted attention. This study predicted LULC, seasonal (summer & winter) LST, and UTFVI variations using machine learning algorithms (MLAs) in Cumilla City Corporation (CCC), Bangladesh. Landsat 4–5 TM and Landsat 8 OLI satellite images were used for 1999, 2009, and 2019 to predict future scenarios for 2029 and 2039. MLAs such as Cellular Automata (CA) and Artificial Neural Network (ANN) methods were used to predict the future change in LULC, LST, and UTFVI. The result suggests that, in the year 2029 and 2039, the urban area will likely to be increased by around 8 % and 11 %, where significant decrease will be taken place in green cover by 9 % and 14 %. If the rapid urban growth continues, more than 30 % of the CCC area will likely to be experienced more than 33 °C temperature and strongest UTFVI effect in the year 2029 and 2039. In addition, an average 4 °C higher LST was recorded in the urban area compared with vegetation cover. In urban construction practice, avoiding concentrated impermeable layers (built-up areas) and increasing green covers, are effective ways of mitigating the effect of UTFVI. This study will contribute in achieving sustainable development and provide useful insights to understand the complex relationship among different elements of urban environments and promotion of city competence.

Introduction

Climate is among the most critical environmental factors impacting not only our ecosystems but also the day-to-day activities (Hunt et al., 2017). Changes in urban land use/land cover (LULC) accelerate climate change because of its role in reducing biodiversity and creating the urban heat island (UHI) effect in cities. Urbanization and industrial development are causing widespread LULC changes with the construction of specialized infrastructures (Tang, Di, Rahman, & Yu, 2019; Rahman, Mohiuddin, Kafy, Sheel, & Di, 2019). Besides, this boost up the economic development accelerates construction of urban infrastructure and increase city competitiveness. It also has significant consequences, such as an increase in air pollution, reducing natural resources, hampering urban biodiversity and accelerates the UHI effect. Land surface temperature (LST) is receiving increasing attention as a measure of urban health and sustainable development. LST is associated with LULC changes, seasonal variations, climate change, and global warming (Huang, Huang, Yang, Fang, & Liang, 2019). Any significant conversion of vegetation area, water bodies to the impervious layer influences the LST variation. As the urban environment consists of a more impervious layer, higher LST is recorded in city areas, which dramatically contributes to UHI's formation and reduces environmental sustainability of the cities (Huang et al., 2019).

Traditional cities may fail to improve land utilization, which create urban sprawl compared to planned cities. There are numerous planned cities around the world, but most cities in Bangladesh are unplanned with small administrative areas and densely populated. Like other developing countries, Bangladesh has been experiencing rapid population increase (Van Schendel, 2020; Rahman et al., 2019). Rapid growth in population may have positive influences on economic development but create negative impacts on LULC change and sustainable development. The LULC composition in each city are different based on city characteristics, planning strategies and time periods. Cumilla, like other cities in Bangladesh, has been experiencing the consequences of rapid population change. Unplanned cities like Cumilla, which do not have organized towering buildings, complete infrastructure and orderly streets, have started to lose their city image and competitiveness. These disadvantages accelerate the UHI phenomenon due to unplanned development, the concentration of infrastructure expansion in the central area and insufficient green space and water bodies. According to the previous studies in Rajshahi, Bangladesh, unplanned infrastructural development and green cover reduction in the city area significantly increase the LST (Kafy, Rahman, Faisal, Hasan, & Islam, 2020). Effective LULC management plan by restricting unplanned urban development and by increasing green cover is currently one of the most significant challenging approaches to mitigate the UHI effects.

The urban thermal field variance index (UTFVI) is widely used to describe the UHI effect (Tomlinson, Chapman, Thornes, & Baker, 2011). The concentration of UTFVI is more in the urban areas due to human activities and substantially warmer than its surroundings neighbouring rural areas (Wang, Zhang, Tsou, & Li, 2017). UTFVI significantly causes negative impacts on the local wind patterns, humidity, air quality, indirect economic loss, reduce comfort and increase mortality rate (Sejati, Buchori, & Rudiarto, 2019). The excessive heat produced by the UTFVI results in a higher upward motion, leading to increased shower and thunderstorm activities (Singh et al., 2017). The UHI is responsible for producing pollutants like ozone which causes degradation in air and water quality (Lai & Cheng, 2010). Predicting the effects of future UTFVI can be an effective approach to identify the possible heat wave zones and help city officials to develop strategies for mitigating UHI effect and ensuring a sustainable environment.

Several factors influence LST to trigger the UHI as well as UTFVI phenomena, such as heatwaves, psychometry, modification of earth surfaces, and illumination intensity. LULC change performs as a significant predominant contributing factor to the UHI effect. The changes in the proportion of LULC types is the most influential factor affecting LST. Predicting future LULC changes and assessing the relationship with LST can effectively prevent the increasing trend of UHI and UTFVI phenomenon. LULC prediction studies can project future scenarios and help ensure environmental projection based on socio-economic development (Kafy et al., 2020). Various simulation studies for identifying the relationship between LULC and LST change have been performed by many researchers, which helped cities to develop future sustainable development strategies. A variety of spatially explicit models such as Cellular Automata (CA) (Balogun & Ishola, 2017; Losiri, Nagai, Ninsawat, & Shrestha, 2016) and Artificial Neural Network (ANN) (Shatnawi & Abu Qdais, 2019) have demonstrated successful prediction results by integrating Remote Sensing (RS) and Geographic Information System (GIS) techniques.

Although prediction of the future LULC and LST scenarios are conventional approaches, prediction studies for seasonal (summer and winter) UTFVI influenced by LULC change is an entirely new concept in RS perspective. The novelty of this study is to predict seasonal UTFVI variations and investigate,

  • spatial and temporal trends of LULC change, seasonal LST and UTFVI variations, and LST and UTFVI distribution over LULC categories using Landsat satellite images for Cumilla City Corporation (CCC) area over the past two decades (1999, 2009 & 2019) and

  • predict future changes in LULC, LST, and UTFVI using CA and ANN algorithms for the years 2029 and 2039.

Section snippets

Study area

CCC is considered for this study, which is one of the most influential city-centers in the Chittagong division of eastern Bangladesh. This area is located along the Dhaka-Chittagong highway which is one of the busiest highways of Bangladesh. The coordinate of this area falls between 22˚0'0" N to 26˚0'0” N latitude and longitude 88˚0’0” E to 92˚0’00” E, covering an area of approximately 46 km2 (Fig. 1). It serves as the administrative center of the Cumilla District. Formerly, CCC was under the

Materials and method

Three Multi-spectral Landsat 4–5 TM and Landsat 8 OLI satellite data were acquired for the year 1999, 2009 and 2019 from the United States Geological Survey (USGS) to explore the LULC and LST change and its effect on UTFVI in the study area. In every summer and winter seasons, satellite images were downloaded within the maximum one-month interval to eliminate the seasonal variability. During the downloading stage, the cloud cover was set to less than 10 percent for all images to have a

Result and discussion

As per methodology (in Section 3), the multi-year LULC pattern, LST, and UTFVI distribution with the prediction of future LULC, LST and UTFVI distribution were calculated for the study area. Results are presented in the following subsections.

Conclusion

This study examined the LULC types changes and investigated the relationship with seasonal LST and UTFVI in Cumilla from 1999 to 2019. The cross-linkage analysis was conducted to find out the association between LULC categories, LST, and UTFVI. By analyzing the data in 1999, 2009 and 2019, the authors clarified that an increase in the urban green cover among the city's built-up areas could significantly decrease the seasonal LST and UTFVI effect. The CA model was employed to predict the

Authors contributions

Abdulla - Al Kafy” developed the original idea for this research, study conception and design, monitor and guideline provider for acquisition of data and writing, analysis and interpretation of data, drafted and reviewed the manuscript and provided input in writing and finalizing the full manuscript. “Abdullah-Al-Faisal” reviewed and drafted the manuscript, predicted the seasonal LST, made language correction and finalized the manuscript. “Md. Shahinoor Rahman” updated the manuscript contents,

Declaration of Competing Interest

The authors report no declarations of interest.

Acknowledgments

The authors would like to thank the Cumilla City Corporation, Bangladesh Meteorological Department and the US Geological Survey for assisting this research with data-sets.

References (66)

  • A. Sekertekin et al.

    Modeling diurnal Land Surface Temperature on a local scale of an arid environment using artificial Neural Network (ANN) and time series of Landsat-8 derived spectral indexes

    Journal of Atmospheric and Solar-terrestrial Physics

    (2020)
  • J.S. Silva et al.

    Spatiotemporal impact of land use/land cover changes on urban heat islands: A case study of Paço do Lumiar, Brazil

    Building and Environment

    (2018)
  • D.R. Streutker

    Satellite-measured growth of the urban heat island of Houston, Texas

    Remote Sensing of Environment

    (2003)
  • J.-H. Su et al.

    Modeling habitat suitability of migratory birds from remote sensing images using convolutional neural networks

    Animals

    (2018)
  • S. Ullah et al.
    (2019)
  • Q. Weng et al.

    Estimation of land surface temperature–vegetation abundance relationship for urban heat island studies

    Remote Sensing of Environment

    (2004)
  • X. Yang et al.

    Impact of urban heat island on energy demand in buildings: Local climate zones in Nanjing

    Applied Energy

    (2020)
  • K. Abutaleb et al.

    Assessment of urban heat island using remotely sensed imagery over Greater Cairo, Egypt

    Advances in Remote Sensing

    (2015)
  • U. Avdan et al.

    Algorithm for automated mapping of land surface temperature using LANDSAT 8 satellite data

    Journal of Sensors

    (2016)
  • E. Bailly et al.

    Deliverable 5.3 Policy proposals and decision-making mechanisms for Climate change Adaptation and mitigation using NBS

    (2019)
  • I. Balogun et al.

    Projection of future changes in landuse/landcover using cellular automata/markov model over Akure city, Nigeria

    Journal of Remote Sensing Technology

    (2017)
  • BBS et al.

    District statistics 2011

    (2013)
  • M. Bilgili

    The use of artificial neural networks for forecasting the monthly mean soil temperatures in Adana, Turkey

    Turkish Journal of Agriculture and Forestry

    (2011)
  • M. Bilgili

    Regional assessment of monthly soil temperatures in the Aegean Region of Turkey

    Arabian Journal for Science and Engineering

    (2012)
  • A.R. Carrico et al.

    Extreme weather and migration: Evidence from Bangladesh

    Population and Environment

    (2019)
  • R.G. Congalton et al.

    Assessing the accuracy of remotely sensed data: Principles and practices

    (2008)
  • P. Fu et al.

    Responses of urban heat island in Atlanta to different land-use scenarios

    Theoretical and Applied Climatology

    (2018)
  • M.M. Hittawe et al.

    Abnormal events detection using deep neural networks: Application to extreme sea surface temperature detection in the Red Sea

    Journal of Electronic Imaging

    (2019)
  • M.S. Hossain et al.

    Climate change impacts on farmland value in Bangladesh

    Ecological Indicators

    (2020)
  • J.C. Hunt et al.

    Climate change and growing megacities: Hazards and vulnerability

    Proceedings of the Institution of Civil Engineers-Engineering Sustainability

    (2017)
  • IPCC

    Mitigation of climate change

    (2014)
  • F. Ishaque et al.

    Application of transform software for downscaling global climate model EdGCM results in North-Eastern Bangladesh

    Environmental Engineering Research

    (2020)
  • A. Kafy et al.

    Pond Filling Locations Identification Using Landsat-8 Images In Comilla District, Bangladesh

    1st National Conference On Water Resources Engineering (Ncwre 2018)

    (2018)
  • Cited by (122)

    View all citing articles on Scopus
    View full text