uMngeni Catchment Research Articles

Evaluating Machine-Learning Algorithms for Mapping LULC of the uMngeni Catchment Area, KwaZulu-Natal

Analysis of land use/land cover (LULC) in catchment areas is the first action toward safeguarding freshwater resources. LULC information in the watershed has gained popularity in the natural science field as it helps water resource managers and environmental health specialists develop natural resource conservation strategies based on available quantitative information. Thus, remote sensing is the cornerstone in addressing environmental-related issues at the catchment level. In this study, the performance of four machine learning algorithms (MLAs), namely Random Forests (RFs), Support Vector Machines (SVMs), Artificial Neural Networks (ANNs), and Naïve Bayes (NB), were investigated to classify the catchment into nine relevant classes of the undulating watershed landscape using Landsat 8 Operational Land Imager (L8-OLI) imagery. The assessment of the MLAs was based on a visual inspection of the analyst and commonly used assessment metrics, such as user’s accuracy (UA), producers’ accuracy (PA), overall accuracy (OA), and the kappa coefficient. The MLAs produced good results, where RF (OA = 97.02%, Kappa = 0.96), SVM (OA = 89.74%, Kappa = 0.88), ANN (OA = 87%, Kappa = 0.86), and NB (OA = 68.64%, Kappa = 0.58). The results show the outstanding performance of the RF model over SVM and ANN with a significant margin. While NB yielded satisfactory results, its sensitivity to limited training samples could primarily influence these results. In contrast, the robust performance of RF could be due to an ability to classify high-dimensional data with limited training data.

Exploring social processes in transformation: the case of a collaborative water partnership in South Africa

ABSTRACT We explore the social processes supporting transformation towards collaborative water governance in the uMngeni catchment, South Africa. Using Holling’s adaptive cycle as a heuristic of phases (conservation, release, reorganisation and exploitation) present during transformation of social-ecological systems, we consider the role of learning, power, agency and structure during each phase of the evolution of the uMngeni Ecological Infrastructure Partnership (UEIP). The UEIP is a partnership between government, research institutions, and civil society groups that facilitates broader and more collaborative participation in water management. During the conservation phase, strong control power and institutional structure (denoted by a hierarchical governance mode embodying control and regulation by the State) limited the introduction of new ideas and reinforced single-loop learning. The release phase was triggered by a shock which weakened control power and permitted the introduction of new ideas thereby enabling double-loop learning. The changing conditions gave rise to protean power (defined as results of practices of agile actors coping with uncertainty) which enhanced the agency of key actors who began to mobilise others in a rapid phase of re-organisation. Triple-loop learning was evident in the exploitation phase as new collaborative institutions, that were better able to accommodate innovative ideas, began to emerge. We found the adaptive cycle helpful for delineating phases of change, while the four multi-faceted processes of learning, power, agency and structure proved useful in illuminating dynamics of change. This understanding may help to inform actions to steer transformations towards more sustainable and collaborative water governance in South Africa and elsewhere.

Pandora's Box: assessing the current trends and challenges of IWRM in the uMngeni catchment

AbstractThe uMngeni catchment in KwaZulu-Natal faces numerous challenges that threaten the availability and quality of water resources. To understand the prevalent issues, the purpose of the study was to assess the institutional aspects that may or may not have facilitated Integrated Water Resource Management (IWRM). Twenty-one semi-structured interviews were conducted, and development planning and environmental management tools were analysed. Water User Associations (WUAs) are statutory bodies and have not been established at the local level. Moreover, an operational Catchment Management Agency (CMA) at the regional level is non-existent. Consequently, the implementation of IWRM has been very limited. The establishment of the uMngeni Ecological Infrastructure Partnership (UEIP) has facilitated the integration of role-players in the absence of an operational CMA. Most of the spatial planning and environmental management tools feature water resource planning except for the integrated waste management plans. As a result, poor solid waste management contributes to the poor water quality in the uMngeni catchment. The challenges remaining are the poor implementation of plans due to a lack of human and financial resources. Therefore, the gap created by a non-existent operational CMA means catchment management activities will continue to negatively affect water resources and the degree to which water resource management is integrated.

South Africa needs a hydrological soil map: a case study from the upper uMngeni catchment

Accurate hydrological modelling to evaluate the impacts of climate and land use change on water resources is pivotal to sustainable management. Soil information is an important input in hydrological models but is often not available at adequate scale with appropriate attributes for direct parameterisation of the models. In this study, conducted in three quaternary catchments in the midlands of KwaZulu-Natal, three different soil information sets were used to configure SWAT+, a revised version of the Soil and Water Assessment Tool (SWAT). The datasets were: (i) the Land Type database (currently the only soil dataset covering the whole of South Africa), (ii) disaggregation of the Land Type database using digital soil mapping techniques (called DSMART), and (iii) a dataset where DSMART were complemented by field observations and interpretations of the hydropedological behaviour of the soils (DSMART+). Simulated streamflow was compared with measured streamflow at three weirs with long-term measurements, and the impact of the soil datasets on water balance simulations was evaluated. In general, the simulations were acceptable when compared to other studies, but could be improved through calibration and including small reservoirs in the model. The DSMART+ dataset yielded more accurate simulations of streamflow in all three catchments with Nash-Sutcliffe efficiencies increasing by between 9% and 67% when compared to the Land Type dataset. The value of the improved soil maps is, however, highlighted through the enhanced spatial detail of streamflow generation mechanisms and water balance components. The internal catchment processes are represented more accurately, and we argue that South Africa needs a detailed hydrological soil map for effective water resource management.

Can Sentinel-2 be used to detect invasive alien trees and shrubs in Savanna and Grassland Biomes?

The impact of alien plant invasions on ecosystems translates into major economic losses for nations across the globe. Freely available satellite imagery can be used to provide relevant data for effective management of alien plant invasions, especially for developing nations with limited budgets. However, detecting alien herbaceous and shrubby vegetation using freely available imagery has not been widely tested. The aim of this study was to detect invasive alien plants in the Savanna and Grassland Biomes of the uMngeni Catchment. The novelty of this study is the focus on not only alien trees, but also alien shrubs, which tend to be more sparsely spread and are harder to detect. Using the free Google Earth Engine Platform for repeatability, we explore the efficacy of data fusion of Sentinel-2, Sentinel-1 and landform data, various feature selection techniques and two powerful non-parametric classifiers. We found that feature selection techniques result in lower classification accuracies and are therefore not recommended for similar applications. Data fusion, such as combining Sentinel-2, Sentinel-1 and landform data, did not improve classification results. Support Vector Machine shows superiority over the Random Forest classifier. The final classifications indicate, with 80–90 % accuracy, that for 2019 about 17.5 % (736.15–763.75 km2) of the uMngeni Catchment was occupied by invasive alien trees and shrubs (e.g., gums, wattles, pines and Bugweed), either in the form of plantations or invasions. This translates to between 88.3 million and 173 million m3 of water per year in evapotranspiration losses, or 23 % or the current system yield. At a spatial resolution of 20 × 20 m, these outputs are useful for regional or local planning and prioritization. Furthermore, the repeatable approach of this study is a significant advance compared to past efforts that have produced publicly available datasets for the region, and could be employed by other data-scarce developing nations.

Use of biological and water quality indices to evaluate conditions of the Upper uMngeni Catchment, KwaZulu-Natal, South Africa

Urban and agricultural land uses have the potential to severely compromise the quality of impoundments, if ineffectively managed and operated. A case in point is the upper uMngeni Catchment, including Midmar Dam, which is integral to the freshwater supply infrastructure in KwaZulu-Natal, South Africa. Monitoring sites were established in varying land use types in three subcatchments of the upper uMngeni Catchment to assess water quality and ecosystem health impacts of current land uses. Conclusions about water quality were drawn using pairing of SASS5 and spot water quality data. Water quality and ecological condition were highest in commercial plantations and upstream of a high-density settlement where natural land cover and sparse settlement occurred. Although marked declines in water quality and ecological condition were observed under commercial agriculture. The most notable declines in water quality and ecological condition were observed downstream of the settlement with elevated nutrient loads. Shifts in aquatic biota were highly correlated with seasonal shifts in water quality, influenced markedly by land use. The cumulative effects of current land use activities, urban development and agriculture on Midmar Dam’s water quality should be viewed with concern. Additional development in the form of additional social housing projects may exacerbate impacts.

An assessment of the ecological condition of a wetland on the Lions River floodplain based on soil and vegetation parameters, South Africa

Wetlands are increasingly exposed to human activities, resulting in degradation. As the wetland degrades, it loses functionality. In South Africa, wetlands can play an important water-regulating role. This study aims to establish the ecological condition of a historically utilised wetland on the Lions River floodplain in the uMngeni catchment, to provide a guide for the planning and implementation of rehabilitation interventions. A comprehensive assessment of the wetland’s structure was undertaken in 2014, using vegetation and soil parameters, mapped and compared with changes in landuse on the wetland from historical aerial photographs. The study concluded that the wetland’s ecological condition had declined, as a result of historical cultivation and commercial forestry. The wetland vegetation present is still generally aligned with the prevailing hydro-edaphic gradient and soil water regime; however, some areas of the wetland showed a mismatch in the soil water indicators and vegetation wetness indicators (the level of wetness based on vegetation being higher than that showed by the soil), suggesting localised drying out. A moderately high abundance of ruderal indigenous and invasive alien species in 61% of the wetland, particularly the drier areas of the wetland, contributed to the altered ecological condition.

Long-term trends and variability in the microclimates of the uMngeni Catchment, KwaZulu-Natal, South Africa and potential impacts on water resources

The past few decades have produced a wealth of research investigating the extent and severity of global climate change. Patterns in climate variability and change have been studied at global, continental and regional scales. Trends observed at these larger spatial scales are expected to be observable at smaller scales, although research in this regard has been limited in South Africa. This study presents an investigation into the microclimatic trends of a key catchment in the KwaZulu-Natal province, South Africa, with particular focus on water resources. Long-term (1980–2017) historical daily meteorological data were analysed using the RClimDex software package as well as other multi-temporal regression and time series analyses. Results indicate an overall warming of the study sites between 1 and 4 °C century−1. Atmospheric water vapour pressure was found to be increasing significantly at three of the sites. Total annual rainfall, an important variable for water resources of the catchment, did not display significant change in the catchment but high levels of inter-annual variability were observed. The results of the study indicate that the catchment microclimate has changed, and is changing, but at different rates from global and continental estimates. The study emphasises the importance of catchment scale analyses using ground-based historical meteorological data.

Identifying hotspots for investment in ecological infrastructure within the uMngeni catchment, South Africa

In recent times, there has been a growing recognition of the role which ecological infrastructure (EI) can play within the water resources management arena. However, practice has generally lagged conjecture regarding the integration of EI in the water resources management decision-making process. In this study, we aim to demonstrate how the state-of-the-art in ecosystem service modelling can be implemented to guide decision making with regards to investments in EI to improve the delivery of specific hydrological ecosystem services (HES) within the uMngeni catchment. For this purpose, the Resource Investment Optimization System (RIOS) and Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) models were applied to identify priority areas for investment and to evaluate the HES benefits that can potentially be achieved. Several global and local data sets were used to provide the requisite inputs to perform alternative land management scenario simulations. The results of these simulations demonstrated that potential investments in EI within the uMngeni catchment can lead to substantial reductions in sediment export (≈47%). Although, this may be accompanied by marginal decreases in the surface water yield (≈1.40%), there is a net benefit associated with reducing sediment export. Despite, these investigations being limited in their representation of HES due to inter alia the lack of localized data sets and inherent model limitations, the study successfully demonstrated how the RIOS and InVEST models can be collectively applied to guide decision-making regarding investments in EI.

Mapping of water-related ecosystem services in the uMngeni catchment using a daily time-step hydrological model for prioritisation of ecological infrastructure investment – Part 1: Context and modelling approach

South Africa is a semi-arid country which frequently faces water shortages, and experienced a severe drought in the 2016 and 2017 rainfall seasons. Government is under pressure to continue to deliver clean water to the growing population at a high assurance of supply. Studies now show that the delivery of water may be sustained not only through built infrastructure such as dams and pipelines, but also through investment in ecological infrastructure (EI). A daily time-step hydrological model was used to map areas which should be prioritised for protection or rehabilitation to sustain the delivery of water-related ecosystem services within the uMngeni catchment. We focused on three water-related ecosystem services, i.e.: water supply, sustained baseflow, erosion control/avoidance of excessive sediment losses. The two key types of degradation were modelled, namely, overgrazing and the invasion of upland areas by Black Wattle (Acacia mearnsii). This, Part 1 of a paper in 2 parts, provides a discussion on the role of EI in delivering water-related ecosystem services, describes the motivation for the study, and the methods used in modelling and mapping the catchment. The results of this modelling exercise are presented in Part 2, which also explores and illustrates the potential hydrological benefits of rehabilitation and protection of EI in the uMngeni Catchment.

Mapping of water-related ecosystem services in the uMngeni catchment using a daily time-step hydrological model for prioritisation of ecological infrastructure investment – Part 2: Outputs

South Africa is a semi-arid country which frequently faces water shortages, and experienced a severe drought in the 2016 and 2017 rainfall seasons. Government is under pressure to continue to deliver clean water to the growing population at a high assurance of supply. Studies now show that the delivery of water may be sustained not only through built infrastructure such as dams and pipelines, but also through investment in ecological infrastructure (EI). Part 1 of this paper in 2 parts concentrated on the role of EI in delivering water-related ecosystem services, as well as the motivation for this study, and the methods used in modelling and mapping the catchment. Part 2 explores and illustrates the current level of delivery of water-related ecosystem services in different parts of the catchment, with potential hydrological benefits of rehabilitation and protection of EI in the uMngeni catchment. The Mpendle, Lions River, Karkloof, Inanda and Durban sub-catchments are important areas for the generation of streamflows which accumulate downstream (i.e. water yield in the catchment) when annual totals are considered. Modelled annual sediment yield (in tonnes) from naturally vegetated areas is most severe in the lower catchment areas with steeper slopes such as Inanda, and in the high-altitude areas which have both steeper slopes and higher rainfall. The central and eastern parts of the uMngeni catchment were found to contribute the greatest yield of sediment from degraded areas with low protective vegetation cover. This combined modelling and mapping exercise highlighted areas of priority ecosystem service delivery, such as higher altitude grassland areas, which could be recommended for formal conservation, or protection under private partnerships. Generally, these areas confirm the intuitive sense of catchment stakeholders, but provide a robust and more defendable analysis through which water volumes are quantifiable, and potential investment into catchment interventions are justified.

Sensitivity analysis for water quality monitoring frequency in the application of a water quality index for the uMngeni River and its tributaries, KwaZulu-Natal, South Africa

Water quality indices are commonly used to provide summary information from water quality monitoring programmes to stakeholders. However, declining funding and changing mandates often result in reduced monitoring frequencies which could affect the accuracy of information provided. Thus, this study aimed to assess the effect of water sampling frequency on water quality index reporting using the the upper uMngeni catchment as a study site. A 28-year time series of water quality data from 11 sampling stations was assessed for pH, electrical conductivity, temperature, turbidity, total suspended solids, Escherichia coli counts, NH4-N, NO3-N, PO4-P and total phosphorus. Statistical packages were used to process the data and water quality indices (WQIs) for eutrophication and recreational water were calculated and their sensitivity to input parameters analysed. It was found that the higher the monitoring frequency, the lower the WQI calculated at all sites. This suggests that water quality, due to a declining monitoring frequency, is poorer than reported in the uMngeni catchment. The findings showed that Escherichia coli and turbidity are the most influential variables affecting the recreational and eutrophication WQIs, respectively. Although WQIs are considered a useful tool for monitoring the changes in water quality across space and over time in the uMngeni Catchment, their use should complement, and not substitute for, other, more comprehensive, water quality management tools.

Effects of land use and land cover changes on water quality in the uMngeni river catchment, South Africa

Land use and land cover change are major drivers of water quality deterioration in watercourses and impoundments. However, understanding of the spatial and temporal variability of land use change characteristics and their link to water quality parameters in catchments is limited. As a contribution to address this limitation, the objective of this study is to assess the linkages between biophysico-chemical water quality parameters and land use and land cover (LULC) classes in the upper reaches of the uMngeni Catchment, a rapidly developing catchment in South Africa. These were assessed using Geographic Information Systems tools and statistical analyses for the years 1994, 2000, 2008 and 2011 based on changes over time of eight LULC classes and available water quality information. Natural vegetation, forest plantations and cultivated areas occupy 85% of the catchment. Cultivated, urban/built-up and degraded areas increased by 6%, 4.5% and 3%, respectively coinciding with a decrease in natural vegetation by 17%. Variability in the concentration of water quality parameters from 1994 to 2011 and an overall decline in water quality were observed. Escherichia coli (E. coli) levels exceeding the recommended guidelines for recreation and public health protection was noted as a major issue at seven of the nine sampling points. Overall, water supply reservoirs in the catchment retained over 20% of nutrients and over 85% of E. coli entering them. A relationship between land use types and water quality variables was found. However, the degree and magnitude of the associations varies between sub-catchments and is difficult to quantify. This highlights the complexity and the site-specific nature of relationships between land use types and water quality parameters in the catchment. Thus, this study provides useful findings on the general relationship between land use and land cover and water quality degradation, but highlights the risks of applying simple relationships or adding complex relationships in the management of the catchment.

Evaluation of uncertainty in capturing the spatial variability and magnitudes of extreme hydrological events for the uMngeni catchment, South Africa

Downscaled General Circulation Models (GCMs) output are used to forecast climate change and provide information used as input for hydrological modelling. Given that our understanding of climate change points towards an increasing frequency, timing and intensity of extreme hydrological events, there is therefore the need to assess the ability of downscaled GCMs to capture these extreme hydrological events. Extreme hydrological events play a significant role in regulating the structure and function of rivers and associated ecosystems. In this study, the Indicators of Hydrologic Alteration (IHA) method was adapted to assess the ability of simulated streamflow (using downscaled GCMs (dGCMs)) in capturing extreme river dynamics (high and low flows), as compared to streamflow simulated using historical climate data from 1960 to 2000. The ACRU hydrological model was used for simulating streamflow for the 13 water management units of the uMngeni Catchment, South Africa. Statistically downscaled climate models obtained from the Climate System Analysis Group at the University of Cape Town were used as input for the ACRU Model. Results indicated that, high flows and extreme high flows (one in ten year high flows/large flood events) were poorly represented both in terms of timing, frequency and magnitude. Simulated streamflow using dGCMs data also captures more low flows and extreme low flows (one in ten year lowest flows) than that captured in streamflow simulated using historical climate data. The overall conclusion was that although dGCMs output can reasonably be used to simulate overall streamflow, it performs poorly when simulating extreme high and low flows. Streamflow simulation from dGCMs must thus be used with caution in hydrological applications, particularly for design hydrology, as extreme high and low flows are still poorly represented. This, arguably calls for the further improvement of downscaling techniques in order to generate climate data more relevant and useful for hydrological applications such as in design hydrology. Nevertheless, the availability of downscaled climatic output provide the potential of exploring climate model uncertainties in different hydro climatic regions at local scales where forcing data is often less accessible but more accurate at finer spatial scales and with adequate spatial detail.

Use of ACRU, a distributed hydrological model, to evaluate how errors from downscaled rainfall are propagated in simulated runoff in uMngeni catchment, South Africa

ABSTRACTThe hypothesis of no significant difference in errors between observed (historical) and downscaled global circulation model (GCM) rainfall and corresponding errors in simulated runoff was tested. The percentage difference in mean and standard deviation, normalized errors and normalized bias metrics were used. The ACRU hydrological model was used to simulate runoff. Results indicated that errors in rainfall lead to amplified errors in simulated runoff. A 10% error magnitude in mean rainfall was amplified three times in mean runoff. Rainfall variability was amplified by twice as much from rainfall to simulated runoff. These findings indicate that uncertainty in input downscaled rainfall is amplified in simulated runoff, hence the quality of input rainfall is a strong determining factor of the simulated runoff. Ultimately, there is a need for continuous improvement in the GCM downscaling process, particularly model process description, so as to minimize uncertainties due to GCM model description.EDITOR A. Castellarin; ASSOCIATE EDITOR S. Kanae

Projected impacts of urbanisation on hydrological resource flows: A case study within the uMngeni Catchment, South Africa

Significant land use changes from natural/agricultural to urban land uses have been proposed within the Mpushini/Mkhondeni sub-catchments of the uMngeni Catchment in South Africa. A better understanding of the influences which such land use changes are likely to have on hydrological flows, is required, in order to make informed land use decisions for a sustainable future. As a point of departure, an overview of linkages between urbanisation and hydrological flow responses within this sub-humid study area is given. The urban characteristics of increased impervious areas and the potential return flows from transfers of potable water from outside the catchment were identified as being important in regard to hydrological flow responses. A methodology was developed to model urban response scenarios with urban characteristics as variables, using the daily time-step process based ACRU model. This is a hydrological multi-process model and not an urban hydraulic model and it addresses the landscape as well as the channel components of a catchment, and in addition to runoff components includes evaporation and transpiration losses as outputs. For the study area strong links between proposed urbanisation and hydrological resource flow responses were found, with increases in stormflows, together with increased and more regulated baseflows, and with impacts varying markedly between dry or wet years and by season. The impacts will depend on the fractions of impervious areas, whether or not these are connected to permeable areas, the amount of imported water and water system leaks. Furthermore, the urban hydrological impacts were found to be relatively greater in more arid than humid areas because of changes in the rainfall to runoff conversion. Flow changes due to urbanisation are considered to have important environmental impacts, requiring mitigation. The methodology used in this paper could be used for other urbanising areas.

An evaluation of how downscaled climate data represents historical precipitation characteristics beyond the means and variances

Precipitation is the main driver of the hydrological cycle. For climate change impact analysis, use of downscaled precipitation, amongst other factors, determines accuracy of modelled runoff. Precipitation is, however, considerably more difficult to model than temperature, largely due to its high spatial and temporal variability and its nonlinear nature. Due to such qualities of precipitation, a key challenge for water resources management is thus how to incorporate potentially significant but highly uncertain precipitation characteristics when modelling potential changes in climate for water resources management in order to support local management decisions. Research undertaken here was aimed at evaluating how downscaled climate data represented the underlying historical precipitation characteristics beyond the means and variances. Using the uMngeni Catchment in KwaZulu-Natal, South Africa as a case study, the occurrence of rainfall, rainfall threshold events and wet dry sequence was analysed for current climate (1961–1999). The number of rain days with daily rainfall>1mm,>5mm,>10mm,>20mm and>40mm for each of the 10 selected climate models was, compared to the number of rain days at 15 rain stations. Results from graphical and statistical analysis indicated that on a monthly basis rain days are over estimated for all climate models. Seasonally, the number of rain days were overestimated in autumn and winter and underestimated in summer and spring. The overall conclusion was that despite the advancement in downscaling and the improved spatial scale for a better representation of the climate variables, such as rainfall for use in hydrological impact studies, downscaled rainfall data still does not simulate well some important rainfall characteristics, such as number of rain days and wet-dry sequences. This is particularly critical, since, whilst for climatologists, means and variances might be simulated well in downscaled GCMs, for hydrologists, downscaled climate data still needs to represent the underlying historical precipitation properties, such as consecutive wet days, number of rain days and their seasonal and monthly distribution during the downscaling process. This then calls for an improvement in the downscaling process by incorporating rainfall drivers such as cyclones, so as to capture better, these rainfall characteristics important to hydrologists.

Effect of spatial resolution on remote sensing estimation of total evaporation in the uMngeni catchment, South Africa

This study evaluated the effect of two readily available multispectral sensors: the newly launched 30 m spatial resolution Landsat 8 and the long-serving 1000 m moderate resolution imaging spectroradiometer (MODIS) datasets in the spatial representation of total evaporation in the heterogeneous uMngeni catchment, South Africa, using the surface energy balance system model. The results showed that sensor spatial resolution plays a critical role in the accurate estimation of energy fluxes and total evaporation across a heterogeneous catchment. Landsat 8 estimates showed better spatial representation of the biophysical parameters and total evaporation for different land cover types, due to the relatively higher spatial resolution compared to the coarse spatial resolution MODIS sensor. Moreover, MODIS failed to capture the spatial variations of total evaporation estimates across the catchment. Analysis of variance (ANOVA) results showed that MODIS-based total evaporation estimates did not show any significant differences across different land cover types (one-way ANOVA; F1.924=1.412, p=0.186). However, Landsat 8 images yielded significantly different estimates between different land cover types (one-way ANOVA; F1.993=5.185, p<0.001). The validation results showed that Landsat 8 estimates were more comparable to eddy covariance (EC) measurements than the MODIS-based total evaporation estimates. EC measurement on May 23, 2013, was 3.8 mm/day, whereas the Landsat 8 estimate on the same day was 3.6 mm/day, with MODIS showing significantly lower estimates of 2.3 mm/day. The findings of this study underscore the importance of spatial resolution in estimating spatial variations of total evaporation at the catchment scale, thus, they provide critical information on the relevance of the readily available remote sensing products in water resources management in data-scarce environments.

Estimating spatial variations of total evaporation using multispectral sensors within the uMngeni catchment, South Africa

Total evaporation is of importance in assessing and managing long-term water use, especially in water-limited environments. Therefore, there is need to account for water utilisation by different land uses for well-informed water resources management and future planning. This study investigated the feasibility of using multispectral Landsat 8 and moderate resolution imaging spectroradiometer (MODIS) remote sensing data to estimate total evaporation within the uMngeni catchment in South Africa, using surface energy balance system. The results indicated that Landsat 8 at 30 m resolution has a better spatial representation of total evaporation, when compared to the 1000 m MODIS. Specifically, Landsat 8 yielded significantly different mean total evaporation estimates for all land cover types (one-way ANOVA; F4.964 = 87.011, p < 0.05), whereas MODIS failed to differentiate (one-way ANOVA; F2.853 = 0.125, p = 0.998) mean total evaporation estimates for the different land cover types across the catchment. The findings of this study underscore the utility of the Landsat 8 spatial resolution and land cover characteristics in deriving accurate and reliable spatial variations of total evaporation at a catchment scale.