Page 1 of 10
European Journal of Applied Sciences – Vol. 11, No. 3
Publication Date: June 25, 2023
DOI:10.14738/aivp.113.14769.
Mamba, M. P., Mkhonta, S. V., Vilane, B. R. T., Mkhwanazi, M. M., & Hlanze, D. K. (2023). An Assessment of Land Use and Land
Cover Change for Lubovane Reservoir Sub-Catchment in Eswatini. European Journal of Applied Sciences, Vol - 11(3). 233-242.
Services for Science and Education – United Kingdom
An Assessment of Land Use and Land Cover Change for Lubovane
Reservoir Sub-Catchment in Eswatini
Mamba, M. P.
Department of Agricultural and Biosystems Engineering,
Faculty of Agriculture, University of Eswatini, Luyengo
Campus P. O. Luyengo M205, Kingdom of Eswatini
Mkhonta, S. V.
Ifa Lethu Technologies CC, P. O. Box 360, Badplaas, 1190
Vilane, B. R. T.1
Department of Agricultural and Biosystems Engineering
Faculty of Agriculture, University of Eswatini, Luyengo
Campus, P. O. Luyengo M205, Kingdom of Eswatini
Mkhwanazi, M. M.
Department of Agricultural and Biosystems Engineering,
Faculty of Agriculture, University of Eswatini, Luyengo
Campus P. O. Luyengo M205, Kingdom of Eswatini
Hlanze, D. K.
Centre for Financial Inclusion, Plot 2176 First Floor Lilunga House,
Somhlolo Road P. O. Box 6805, Mbabane, Kingdom of Eswatini
ABSTRACT
The increasing population in rural areas, increased livestock densities and
extensive deforestation have been reported as the main drivers of land degradation
in Eswatini. Land degradation, along with biodiversity loss and climate change
presents serious challenges to the environment, economy and the country’s
development agenda. This study was conducted to assess the land use land cover
(LULC) changes within the Lubovane reservoir catchment. Landsat 4-5 TM images
were used for mapping LULC changes for 1995, 2000, 2005, 2009 and a Landsat 8
image was used for mapping 2015 LULC. A supervised LULC classification was
conducted using 6 classes (water, settlements, irrigation, cultivation, shrubs and
forests, as well as bare land) in ArcGIS version 10.3.1. The classification was
validated using a confusion matrix and the results reflected that water, irrigation,
cultivation, forests and shrubs were well classified. The LULC assessment results
indicated that there was low coverage of water bodies observed from 1996 – 2005,
while a 3% increase was observed in 2009. Water coverage decreased to 1.9% in
2015 due to the El-Niño induced drought that hit Southern Africa, resulting in low
inflow to the dam. A reduction of shrubs and forest cover was experienced in 2000
1 Corresponding author.
Page 2 of 10
Services for Science and Education – United Kingdom 234
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 3, June-2023
due to conversion of forested areas into settlements for resettled households.
However, a slight increment of shrubs and forest was observed from 2009 to 2015.
A reduction in the concentration of forests cover around the reservoir, an increase
of settlements and bare land were also observed.
Keywords: Land use, land cover, catchment, reservoir, vegetation
INTRODUCTION
Land use land cover (LULC) influences global environmental changes. Therefore, it is crucial to
undertake LULC assessment in order to understand trends of on-going changes ([17] et al.
2021). Land use is defined as the purpose for which land is utilized, whereas land cover refers
to the physical features of the land surface [16]. Changes in LULC may contribute to significant
effects on various environmental, ecological and hydrological systems [13]. The changes of
LULC are caused by multiple interacting factors which could be anthropogenic or natural
processes occurring at different spatiotemporal scales, globally these has been characterized
by gains in agricultural land and reduction of forests [8]. The drivers of LULC changes comprise
a combination of demographics, politics, socioeconomic systems, institutions, as well as nature,
technology and culture. Therefore, due to the diversity of the drivers, LULC change does not
comply to a strict model and is non-deterministic [14]. These challenges have direct impacts to
the livelihoods of the community and subsequently the population both in urban and rural
areas.
The increasing human population densities in rural areas, increased livestock population on
poorly managed rangelands and extensive deforestation have been reported as the main
drivers for land degradation in the Kingdom of Eswatini [6]. Additionally, agricultural
expansion and settlements have led to intensification of land use and adoption of unsustainable
practices, which include loss of natural resources, changes in natural habitats and ecosystems,
biodiversity loss, decrease in water quality and quantity, as well as reduction in productivity of
arable and rangelands in Eswatini [7]. Land degradation, along with biodiversity loss and
climate change presents serious challenges to the environment, economy and the country’s
development agenda [4]. As, such in order to achieve sustainable development, the country has
to respond to these challenges.
In response to these challenges, the government of Eswatini implemented the Lower Usuthu
Smallholder Irrigation Project - Global Environment Facility (LUSIP-GEF)-Sustainable Land
Management (LUSLM) project. Through this project the government intended to reduce land
degradation, biodiversity loss and mitigate climate change in the Lower Usuthu River Basin
area through the application of sustainable land management practices [6]. The LUSLM project
was implemented within the Lubovane reservoir catchment, which also contributed towards
improved management of the watershed draining into the reservoir. This study was
undertaken to assess the LULC changes within the Lubovane reservoir catchment.
The application of remote sensing (RS) and geographic information systems (GIS) theories has
been widely recognized as accurate and highly efficient tools for mapping, characterizing and
monitoring changes in LULC [18]; [12]. RS data also provides valuable information regarding
the processes, location, rate, trend, nature, pattern and magnitude of LULC changes while GIS
Page 3 of 10
235
Mamba, M. P., Mkhonta, S. V., Vilane, B. R. T., Mkhwanazi, M. M., & Hlanze, D. K. (2023). An Assessment of Land Use and Land Cover Change for
Lubovane Reservoir Sub-Catchment in Eswatini. European Journal of Applied Sciences, Vol - 11(3). 233-242.
URL: http://dx.doi.org/10.14738/aivp.113.14769.
enables mapping and analysing the patterns captured in the remotely sensed data thus
enhancing the interpretation and understanding of LULC dynamics [11].
Description of the Study Area
The Lubovane reservoir, is located between latitude 26°46'57.60"S and 26°43'46.28"S and
longitudes 31°38'42.52"E and 31°42'54.45"E (Figure 1). The reservoir occupies a total surface
area of 13.9 km2 at the downstream end of the Mhlathuzane River catchment. The Lubovane
reservoir catchment sits in the Lower Usuthu sub-basin and the Usuthu Basin within Eswatini
covering about 12 700 km2 [10]. The LUSIP began in the late 90s, hence the study covered the
period prior to construction (1995, 2000) and the period after construction (2005, 2009 and
2015) of the reservoir. These years were chosen so that mapping could cover the time period
before the introduction of the project up to 2015, when the project was almost fully
implemented and activities in the project area were highly intensified.
Figure 1. Location of the Lubovane reservoir and its catchment area
The catchment area of the Lubovane reservoir is 524 km2 with a Mean Annual Runoff (MAR) of
70 x 106 m3 as well as a slightly high coefficient of variation (98%) [2]. The average temperature
range for Lubovane Reservoir catchment is 190 C to 300C, with maximum temperatures
reaching 400C, usually around December and January (Figure 2). The annual rainfall in the basin
ranges from 600 to 1000 mm, with the lower parts of the basin receiving the least rains [9].
Page 4 of 10
Services for Science and Education – United Kingdom 236
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 3, June-2023
Figure 2. Average monthly rainfall for Lubovane reservoir Catchment (1981 – 2015)
The dominant land uses in the Lubovane dam catchment were agriculture and rural
settlements. There were various farms both for subsistence and commercial purposes within
the catchment. The areas around Kubuta were characterized by large fruit tree plantations with
the main products being banana and oranges. There were noticeable spots of sand mining both
for coarse grained sand from the streams and also fine-grained plaster sand on the mainland.
The mining posed a risk of soil erosion in the micro-catchment and subsequently
sedimentation, as the sand could be left bare and uncovered.
MATERIALS AND METHODS
Data Requirements and Acquisition
Data required for this study included rainfall data (1995 - 2015), slope and elevation maps, land
cover/land use changes from the year 1995 to 2015 as well as data on the type of soils in the
study area. This period was chosen in order to cover the period before the construction of the
reservoir, resettlement period which saw a lot of deforestation around the water body, during
the construction and after the reservoir was commissioned. Data for rainfall from the period
1995 to 2015 was acquired from the Eswatini Meteorological Services for three stations (Big
Bend, Sithobela and Khubutha) falling in and closest to the study area. Coordinates for the
stations were collected using a GPS and then the stations were overlain on the map for the study
area to show their spatial distribution. The Thiessen polygon method in GIS was used to
determine the areas of influence of each station and the rainfall for each location was the
resultant rainfall from the influence of the interpolated stations.
ASTER Digital Elevation Model (DEM) covering the study area was retrieved from the Global
ASTER GDEM website (http://gdem.ersdac.jspacesystems.or.jp/). This was used to create a
sub-catchment map of the Lubovane reservoir to delineate the area that drains into the
reservoir. The elevation and slope factors were also determined from the DEM. Cloud free
0
20
40
60
80
100
120
140
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Average mothly rainfall (mm)
Bigbend Sithobela Kubuta
Page 5 of 10
237
Mamba, M. P., Mkhonta, S. V., Vilane, B. R. T., Mkhwanazi, M. M., & Hlanze, D. K. (2023). An Assessment of Land Use and Land Cover Change for
Lubovane Reservoir Sub-Catchment in Eswatini. European Journal of Applied Sciences, Vol - 11(3). 233-242.
URL: http://dx.doi.org/10.14738/aivp.113.14769.
Landsat TM images were downloaded from the United States Geological Survey website
(http://www.glovis.usgs.gov) for the years 1995, 2000, 2009 and 2015. The land cover maps
were necessary for mapping land use changes in the reservoir catchment from 1995 to 2015.
The DEM hydro-processing for catchment delineation was performed using the Integrated
Land and Water Information System (ILWIS) in order to obtain the boundary of the Lubovane
catchment area which enclosed all points draining into the reservoir.
Land Use and Land Cover Classification
Landsat 4-5 TM images were used for mapping land cover changes for 1995, 2000, 2005 and
2009. A Landsat 8 image was used for mapping land cover for 2015. This period was preferred
for the study because it covers the both the situation prior and post construction of the
Lubovane reservoir. The LULC classification was done using ArcGIS version 10.3.1. Since the
researcher was familiar with the study area, the supervised land cover classification method of
classification was opted for this study. Additionally, this land cover classification method
provides a high level of accuracy than other existing image classification methods [3].
Before beginning the classification, the dynamic range adjustment (DRA) function under image
analysis was used to enhance visualization of the image. Classification was done using a
signature of points extracted from the image and assigned classes according to the researcher’s
knowledge of the area. Six land cover classes were selected namely; water, settlements, forest
and shrubs, irrigated area, cultivated area and bare land.
A confusion matrix was used for assessing the accuracy of the land cover classification. The
confusion matrix was also conducted using ArcGIS version 10.3.1. The use of confusion matrix
in accuracy assessment is based on the assumption that each pixel can be allocated to a single
class in both the ground and map data sets, and that these two data sets have the same spatial
resolution and are perfectly registered [15]. Figure 3 shows a layout of a typical confusion
matrix and the equations for computing the producer’s and the user’s accuracy.
Page 6 of 10
Services for Science and Education – United Kingdom 238
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 3, June-2023
Figure 3 Layout of a typical confusion or error matrix, showing computation of user’s and
producer’s accuracies [15]
RESULTS AND DISCUSSION
Validation of the Land Cover Classification
Land cover classification, even though supervised, may possess some errors in the classification
where pixels are assigned a class that they do not represent in the real image [1]. The accuracy
of land cover assessment is considered to be poor below 50%; fair above 50% and good above
70% [5] ; [15] and [19]. The Land cover classification performed in this study was good
according to the above-mentioned classification. Table 1 and Table 2 presents the results of the
confusion matrix used to determine the accuracy of land cover classification for the study.
Table 1. Confusion matrix for Land cover classification
W SL IRR CU F&S BL SUM Producers’
accuracy
Water (W) 52 0 1 0 0 0 53 98.1
Settlements (SL) 8 24 1 8 19 17 77 31.2
Irrigation (IRR) 0 0 59 0 0 0 59 100.0
Cultivation (CU) 0 2 0 14 0 0 16 87.5
Forest and Shrubs
(F&S) 0 0 0 5 24 0 29 82.8
Bare land (BL) 0 3 0 3 0 6 12 50.0
SUM 60 29 61 30 43 23 246
User's accuracy 86.7 82.8 96.7 46.7 55.8 26.1 72.4
Page 7 of 10
239
Mamba, M. P., Mkhonta, S. V., Vilane, B. R. T., Mkhwanazi, M. M., & Hlanze, D. K. (2023). An Assessment of Land Use and Land Cover Change for
Lubovane Reservoir Sub-Catchment in Eswatini. European Journal of Applied Sciences, Vol - 11(3). 233-242.
URL: http://dx.doi.org/10.14738/aivp.113.14769.
The results reflected that the water, irrigation, cultivation, forests and shrubs were well
classified, whilst bare land and settlements were poorly classified. Water, irrigation and
cultivation were probably easy to identify and classify, especially in the 2015 image. This was
because the reservoir and the large areas under sugarcane production downstream of the
Lubovane reservoir fell in the same tile. The Mhlatuzane catchment was predominantly
composed of rural settlements, which had sparse homesteads, as opposed to clustered high
density settlement patterns. It is worth noting that some of the homesteads were made out of
thatched, stick and mud houses that had the same reflectance as bare land. That made it difficult
to distinguish between most settlements and bare land, hence the poor classification between
the two classes.
Table 2. Class and Overall Accuracy
Class Accuracy (%) Overall Accuracy (%)
Water 92.4
Settlements 57.0
Irrigation 98.4
Cultivation 67.1
Forest and Shrubs 69.3
Bare land 38.0
Average 70.4 72.8
Land Use/Land Cover Changes
The results in Figure 4 and Figure 5 indicated that there were some noticeable changes in the
land cover and land use patterns over the mapped period (1995, 2000, 2005, 2009 and 2015).
The changes observed included forest coverage, changes in bare land and changes in farming
activities. These changes have a direct impact on soil erosion and subsequently sedimentation.
On the other hand, the land area covered by water bodies was low during the period 1995 –
2005. Significant amount of land area covered by water started showing in the 2009
assessment. It is worth noting that this was the period when the dam started receiving water.
The areal coverage of water increased from 0.05% coverage in 2005 to 3% coverage in 2009.
Page 8 of 10
Services for Science and Education – United Kingdom 240
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 3, June-2023
Figure 4. Land use land cover changes in the Lubovane catchment
Figure 5. Land use and land cover changes in Lubovane reservoir catchment
0
10
20
30
40
50
60
Water Settlements Irrigation Cultivation Forest and
Shrubs
Bareland
Areal coverage (%)
1995 2000 2005 2009 2015
Page 10 of 10
Services for Science and Education – United Kingdom 242
European Journal of Applied Sciences (EJAS) Vol. 11, Issue 3, June-2023
Report.
7. IFAD, (2016) Lower Usuthu Sustainable Land Management (LUSLM) Project: Reducing land degradation,
biodiversity loss and mitigating climate change in the Lower Usuthu River Basin through the application of
sustainable land management practices. Rome.
8. Leta, M. K. and Demissie, T. A. (2021) Modeling and Prediction of Land Use Land Cover Change Dynamics
Based on Land Change Modeler ( LCM ) in Nashe Watershed, Upper Blue Nile Basin , Ethiopia. Sustainability
13:24.
9. Manyatsi, A. and Brown, R. (2009) IWRM Survey and Status Report : Swaziland. Mbabane, Kingdom of
Eswatini.
10. Mhlanga, N.; Matondo, J.; Nobert, J. and Salam, A. (2012) Evaluation of the Impact of Climate Change on the
Inflow To Lubovane Reservoir. J Sustain Dev Africa 14:96–116.
11. Munthali, M. G.; Botai, J. O.; Davis, N. and Adeola, A. M. (2019) Multi-temporal analysis of land use and land
cover change detection for dedza district of Malawi using geospatial techniques. Int J Appl Eng Res 14:1151–
1162.
12. Nde, S. C.; Bett, S.; Okpara, E. C. et al (2020) An Assessment of Land Use and Land Cover Changes and Its
Impact on the Surface Water Quality of the Crocodile River
13. Patil, N.S. and Nataraja, M. (2020) Effect of land use land cover changes on runoff using hydrological model:
a case study in Hiranyakeshi watershed. Model Earth Syst Environ 6:2345–2357.
https://doi.org/10.1007/s40808-020-00808-8 Accessed May, 2023.
14. Ryu, S. M.; Yang, H. S. and Shon, O. J. (2018) Staged treatment of bicondylar tibial plateau fracture (Schatzker
type v or vi) using temporary external fixator: Correlation between clinical and radiological outcomes. Knee
Surg Relat Res 30:261–268. https://doi.org/10.5792/ksrr.17.008. Accessed May, 2023.
15. Strahler, A. H.; Boschetti, L. Foody, G. M. et al (2006) Global Land Cover Validation: Recommendations for
Evaluation and Accuracy Assessment of Global Land Cover Maps. Luxembourg.
16. Tizora, P.; Roux, A. Le.; Mans, G. and Cooper, A. (2015) Land Use and Land Cover Change in the Western
Cape Province : Quantification of Changes & Understanding of Driving Factors.
17. Wnęk, A.; Kudas, D, and Stych, P. (2021) National level land-use changes in functional urban areas in Poland,
Slovakia, and Czechia. Land 10:1–16. https://doi.org/10.3390/land10010039 Accessed May, 2023.
18. Zhang, B.; Zhang, Q,. Feng, C. et al (2017) Understanding land use and land cover dynamics from 1976 to
2014 in Yellow River Delta. Land 6:1–20. https://doi.org/10.3390/land6010020. Accessed May, 2023.
19. Zhang, C.; Dong, J. and Ge, Q. (2022) Quantifying the accuracies of six 30-m cropland datasets over China: A
comparison and evaluation analysis. Comput Electron Agric 197:106946.
https://doi.org/10.1016/j.compag.2022.106946. Accessed May, 2023.