UPM Institutional Repository

Relationship between deforestation and land surface temperature across an elevation gradient using satellite imagery in Cameron Highlands, Malaysia


How, Darren Jin Aik (2021) Relationship between deforestation and land surface temperature across an elevation gradient using satellite imagery in Cameron Highlands, Malaysia. Doctoral thesis, Universiti Putra Malaysia.


The Cameron Highlands has experienced multiple land encroachment activities and repeated deforestation, leading to extensive land-use and land-cover change (LULCC) during the past six decades. The recent deforestation has possibly contributed to the warming and increased LST. On the other hand, the rise in LST could be directly linked to deforestation due to the expansion of urban areas, including agriculture. However, deforestation and land cover dynamics and their effect on land surface temperature (LST) in the highland areas are not well known. This study aims to explore the drivers and impacts of deforestation as a direct cause of urbanization and land expansion and its effects on the land surface temperature of the Cameron Highlands between 2009 and 2019 using satellite imagery. The specific objectives were three folds; (i) to detect Land Use and Land Cover Change (LULCC) between 2009 and 2019 in Cameron Highlands, (ii) to evaluate the relationship between Land Use and Land Cover Change (LULCC) and Land Surface Temperature (LST) using Landsat and MODIS imageries, and (iii) to assess Land Use and Land Cover Changes (LULCC) of different forest types across an altitude gradient. Geospatial techniques and remotely sensed data were employed to analyse Landsat 7 Enhanced Thematic Mapper Plus (ETM+) and 8 Operational Land Imager (OLI/TIRS), Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), Global Digital Elevation Model (GDEM), and Moderate Resolution Imaging Spectroradiometer (MODIS) 11A sensors. First, land cover classes and detection were identified using an Object-based Image Analysis (OBIA) classification technique on both Landsat 7 and 8 sensors, using a combination of nearest neighbour and multiresolution segmentation algorithm (MSA). Then, for the derivation of LST, a single channel (2009-2012) and Split- Window Algorithm (SWA) (2013-2019) was applied to derive the LST. In order to validate the results, air temperature data were obtained from Met Malaysia and MODIS data. Then this study determined the LULCC across forest types according to the forest type-elevations. Results have shown a significant rise in both agriculture and urban change where LULC change for agriculture nearly tripled in 10 years from 4.93% to 12.63%, while urban development increased from 7.48% to 9.12% between 2009 and 2019. This comes as a cost of a decline in primary forests by 59.44 km2 (8.87%) of total land area between 2009 and 2019. LST experienced an average increase of 2 °C between 2009 and 2019 for the overall study area, where hotspots were found to concentrate in the main towns of Ringlet, Brinchang and Tanah Rata. Our validation results proved successful as the accuracy of LULC, and LST outputs achieved 94.6% and 80.0%, respectively. The forest type most affected by deforestation is the upper dipterocarp forest, reducing 232.54 km2 to 207.38 km2. This is where most urban and agricultural land is located. A further study of LULC on slopes had shown an expansion of agriculture and urban development onto slopes above 35°, prevailing in 2014-2019. The sensitive upper dipterocarp forests saw an interannual temperature variation of +/- 5 °C with a gradual incline until 2019. This study provides a novel and essential fundamental research finding for Cameron Highland. Thus, government bodies, land planners, and environmentalists benefited to understand the impacts of LULC on LST. This study can be helpful in highland planning and development and control deforestation expansion to conserve forests and environmental sustainability in the mountainous region. On the other hand, this study could evaluate to a level where ecosystems and social systems can support the development of the REDD+ policy and approach achieving a low carbon credit value in this country.

Download File

[img] Text
FPAS 2021 5 IR.pdf

Download (2MB)

Additional Metadata

Item Type: Thesis (Doctoral)
Subject: Deforestation - Malaysia
Subject: Land use - Remote sensing
Subject: Remote-sensing images
Call Number: FPAS 2021 5
Chairman Supervisor: Professor Gs. Ts. Mohd Hasmadi bin Ismail, PhD
Divisions: Faculty of Forestry and Environment
Depositing User: Mas Norain Hashim
Date Deposited: 11 Aug 2022 00:28
Last Modified: 11 Aug 2022 00:28
URI: http://psasir.upm.edu.my/id/eprint/98280
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item