Assessment of landuse induced land surface temperature changes in the Greater Accra Region, Ghana
No Thumbnail Available
Date
2019-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
KNUST
Abstract
This study explored the application of geospatial technology in understanding change pattern of
Land use/Land cover (LULC) and associated Land Surface Temperature (LST) variability in the
Greater Accra region between 1986 and 2018. Landsat satellite imagery between the months of
December and February (dry season) were used for LULC analysis and LST Estimation. MODIS
LST data for 1986, 2002 and 2018 were also utilized for the validation of satellite derived LST.
The aim of this research was to investigate how LULC changes affects LST. The results of analysis
show that LULC in Greater Accra region has changed significantly over the 32-year period. Four
major LULC types identified were; vegetation, built-up, bare land and waterbody. Bare land cover
type indicated decrease from 1986 to 2018 of 60.22%. Water bodies experienced varying trend of
7.84% decrease representing 9.11 Km2
. Vegetation cover indicated 16.20% increase in 2002 and
28.02% decrease in 2018. This amounted to a total of 11.82% decrease in vegetation cover. On the
contrary, built-up increased by 654.03% (724.3 Km2
) between 1986 and 2018. LST was observed
to have increased by 3.05 over the 32-year period of this study across the Greater Accra region.
with the highest LST of 36.95 and the lowest of 20.19 , both recorded in 2002. The trend
however indicated a significant increase between 1986 and 2002 but a decrease in LST from 2002
to 2018. The increasing trend in LST over time was found to be consistent with the observed urban
expansion of the study area. High temperatures are also more apparent along the coastal belt of the
region stretching northwards as development spread in the same direction. Total change in LST
for the entire period of the study show increases of 3.53 , 3.04 , 2.78 and 2.65 for bare land,
vegetation, built-up and waterbody land use/land cover types respectively. LULC association with
LST shows that bare land experienced the highest negative change of 60.22% and this
corresponded to the highest increase in LST of 3.53 over the same period. Correlation analysis
and Coefficient of variation indicated largely that, MODIS LST data validated derived LST from
Landsat satellite imagery. Trend of both LST values follow identical pattern with few instances
where derived LST is greater than MODIS LST. Both NDBAI and NDBI exhibited a positive
correlation with LST. NDVI demonstrated a negative linear correlation while NDWI indicated the
strongest negative linear correlation with LST. Land use/Land cover change can be linked to rising
Land Surface
Temperature Changes in the Greater Accra region of Ghana. The research stresses the need to
ensure effective land use planning by relevant regulatory bodies to control urban development and
address anthropogenic causes of surface temperature changes.
Description
A Thesis submitted to the Department of Geomatic Engineering
in partial fulfilment of the requirements for the degree of MASTER OF PHILOSOPHY in geographic information systems