Investigating the relationship between land use/land cover change and land surface temperature using Google Earth Engine; case study: Melbourne, Australia

[thumbnail of sustainability-14-14868.pdf]
sustainability-14-14868.pdf - Published Version (28MB) | Preview
Available under license: Creative Commons Attribution

Jamei, Yadhar ORCID: 0000-0003-2568-3760, Seyedmahmoudian, Mehdi, Jamei, Elmira ORCID: 0000-0002-4270-0326, Horan, Ben ORCID: 0000-0002-6723-259X, Mekhilef, Saad ORCID: 0000-0001-8544-8995 and Stojcevski, Alex (2022) Investigating the relationship between land use/land cover change and land surface temperature using Google Earth Engine; case study: Melbourne, Australia. Sustainability (Switzerland), 14 (22). ISSN 2071-1050


The rapid alteration to land cover, combined with climate change, results in the variation of the land surface temperature (LST). This LST variation is mainly affected by the spatiotemporal changes of land cover classes, their geospatial characteristics, and spectral indices. Melbourne has been the subject of previous studies of land cover change but often over short time periods without considering the trade-offs between land use/land cover (LULC) and mean daytimes summer season LST over a more extended period. To fill this gap, this research aims to investigate the role of LULC change on mean annual daytime LST in the hot summers of 2001 and 2018 in Melbourne. To achieve the study’s aim, LULC and LST maps were generated based on the cost-effective cloud-based geospatial analysis platform Google Earth Engine (GEE). Furthermore, the geospatial and geo-statistical relationship between LULC, LST, and spectral indices of LULC, including the Normalised Difference Built-up Index (NDBI) and the Normalised Difference Vegetation Index (NDVI), were identified. The findings showed that the mean daytime LST increased by 5.1 °C from 2001 to 2018. The minimum and maximum LST values were recorded for the vegetation and the built-up area classes for 2001 and 2018. Additionally, the mean daytime LST for vegetation and the built-up area classes increased by 5.5 °C and 5.9 °C from 2001 to 2018, respectively. Furthermore, both elevation and NDVI were revealed as the most influencing factors in the LULC classification process. Considering the R2 values between LULC and LST and their NDVI values in 2018, grass (0.48), forest (0.27), and shrubs (0.21) had the highest values. In addition, urban areas (0.64), bare land (0.62), and cropland (0.61) LULC types showed the highest R2 values between LST regarding their NDBI values. This study highlights why urban planners and policymakers must understand the impacts of LULC change on LST. Appropriate policy measures can be proposed based on the findings to control Melbourne’s future development.

Dimensions Badge

Altmetric Badge

Item type Article
DOI 10.3390/su142214868
Official URL
Subjects Current > FOR (2020) Classification > 4005 Civil engineering
Current > Division/Research > College of Science and Engineering
Keywords sustainability, Google Earth, land surface temperature, LST, land use and land cover, LULC, Melbourne, Australia
Download/View statistics View download statistics for this item

Search Google Scholar

Repository staff login