Browsing by Author "Gebreslasie, Michael Teweldemedhin."

Now showing 1 - 20 of 21

Analysis of geographical and temporal patterns of malaria transmission in Limpopo Province, South Africa using Bayesian geo-statistical modelling.
(2013) Mgabisa, Aphelele Ronnie.; Gebreslasie, Michael Teweldemedhin.
South Africa is at the southern fringe of sub-Saharan African countries which persist in experiencing malaria transmission. The purpose of the study is to analyse the geographical and temporal patterns of malaria transmission from 2000 to 2011 using Bayesian geostatistical modelling in Limpopo Province, South Africa. Hereafter, develop malaria case data-driven spatio-temporal models to assess malaria transmission in Limpopo Province. Malaria case data was acquired from the South African Medical Research Council (MRC). Population data was acquired from AfriPopo; and Normalised Difference Vegetation Index (NDVI), Land Surface Temperature (LST) and Land Cover data were acquired from MODerate-resolution Imaging Spectro-radiometer (MODIS). Rainfall, Altitude and distance to water bodies’ data were acquired from African Data Dissemination Service (ADDS), United States Geological Survey (USGS) and Environmental Systems Research Institute (ESRI), respectively. Bayesian spatio-temporal incidence models were formulated for Gibbs variable selection and models were fitted using the best set of environmental factors. Modelbased predictions were obtained over a regular grid of 1 x 1km. spatial resolution covering the entire province and expressed as rates of per 1 000 inhabitants for the year 2010. To assess the performance of the predicted malaria incidence risk maps, the predictions and field observations were compared. The best set of environmental factors selected by variable selection was Altitude and the night temperature of two months before the case was reported. The environmental factors were then used for model fitting and all of the covariates were important on malaria risk. Predictions were done using all the environmental factors. The predictions showed that Vhembe and Mopani district municipalities have high malaria transmission as compared to other district municipalities in Limpopo Province. Assessment of predictive performance showed scatter plots with the coefficient of determination ( R² ). The values representing the statistical correlation represented by the coefficient of determination ( R² ) were 0.9798 (January), 0.8736 (February), 0.8152 (March), 0.8861 (April), 0.9949 (May), 0.3838 (June), 0.7794 (July), 0.9235 (September), 0.8966 (October), 0.9834 (November) and 0.8958 (December). August had two values reported and predicted which resulted in R² of 1. The numbers of the The produced malaria incidence maps can possibly be considered as one of the baselines for future malaria control programmes. The results highlighted the risk factors of malaria in Limpopo Province which are the most important characteristics of malaria transmission.
Analysis of the geographical patterns of malaria transmission in KwaZulu-Natal, South Africa using Bayesian spatio-temporal modelling.
(2013) Ndlovu, Noluthando.; Gebreslasie, Michael Teweldemedhin.; Vounatsou, Penelope.
Malaria is one of the most important public health issues that is still affecting millions of people around the world, especially in Africa. Africa accounted for 80% of the 216 million cases worldwide and 91% of deaths. It poses serious economic burdens on communities and countries at large. However, through temporal and spatial mapping of the disease populations at risk can be identified timeously and resources distributed accordingly. Since malaria is a climatic disease geostatistical approaches can be utilised in modelling its spatial distribution. Bayesian geostatistical methods enable the mathematical descriptions of the environment-disease association. Significant environmental predictors of malaria transmission can be identified which can also allow for the development of a malaria epidemic prediction model. This model can serve as a surveillance system for early detection and containment of the disease. Therefore, it is crucial to understand the complex dynamics of malaria transmission so malaria control programmes can be more effective and efficient in managing this public health issue. In South Africa, malaria is transmitted in 3 provinces: KwaZulu-Natal, Mpumalanga and Limpopo. Although malaria is highly seasonal in these areas and KwaZulu-Natal has experienced tremendous achievements in decreasing morbidity and mortality due to malaria, it still remains in an unstable condition that needs constant control and surveillance. The aim of this study was to investigate which environmental/climatic variables are drivers of malaria incidence in KwaZulu-Natal and subsequently develop methods to produce risk maps using Bayesian spatio-temporal modelling. It emerged from the research that the main environmental/climatic drivers of malaria incidence in KwaZulu-Natal were the day temperature of the previous month, altitude and forest land cover type. This was due to the different ways these three factors affect the three-way interaction of the vector, the parasite and the human host. The predicted risk maps showed that incidence rates ranged from 0.2 to 5 per 1000 inhabitants in the study area. This prediction was based on only the climatic factors, however, non-climatic factors also affect malaria transmission through vector control strategies like Indoor Residual Spraying among others.
Assessing changes in land use and land cover using remote sensing : a case study of the Umhlanga Ridge sub-place.
(2015) Kercival, Nadira.; Gebreslasie, Michael Teweldemedhin.
Land has proven to be a key component in the development of the human population and is viewed as one of the most significant natural resources currently available. This observation brings into question the recent debates surrounding the pressures people tend to impose on the land, which have resulted in transformations in its physical landscape and usage. Mans impact on the earth in terms of the transformations in land cover and land use have rapidly increased over the years. However, as a result of these continuous land transformations, planning and designing sustainable urban development has become challenging due to the additional fact that the available mapped resources of the land can be outdated or of very poor quality. One of the main methods of depicting the significant changes in land cover or land usage is through the utilization of remote sensing and its key application of change detection. Change detection enables the user to analyse the transformations of land use and land cover as it is able to provide consistent coverage at short intervals. One of Durban’s greatest cases of land transformations is change which has occurred in Umhlanga and its surrounding areas. Umhlanga started out as an area of tremendous agricultural value to the South African economy by producing substantial amount of sugar from its vast lands of sugarcane. Over time however, Umhlanga began to develop its coastline and gradually it expanded the transition until it became a central hub of social and urban development. Therefore, this research endeavour focuses on depicting this above mentioned land transformation from fields of sugarcane to the presently expanding area of suburban development in Umhlanga Ridge. The aim of this study is to assess and analyse the land use and land cover changes that have occurred in Umhlanga Ridge using remotely sensed data and to further understand the socio – economic implications of these changes. Utilising the change detection method of image differencing, the remotely sensed data provided by the South African National Space Agency (SANSA) was analysed to identify the changes that have occurred between the years 2006 and 2012. The results yielded from this research endeavour have proved that within Umhlanga and specifically that of Umhlanga Ridge, major land use and land cover transformations have occurred. The most dominant and evident change was found in the larger extent of the Transport and Urban land cover classes by the year 2012. While classes such as the forest and woodland, cultivated Land, grassland and the barren land cover experienced significant drops in their land cover extent, the results generated from the analysis showed that most of these land covers were taken over by the development of urban features. Furthermore, this study reported profound socio – economic implications which occurred due to widespread land use and land cover changes. While many implications were documented, one of the main implications of this nature was found to be the significant number of employment opportunities that became available as a result of the expanding urban landscape. As the urban landscape of this area is continuing to change and expand, it is important to highlight that as a direct consequence of this action, the extent of the naturally occurring environment is being depleted. Therefore, with the urban development now being Umhlanga’s dominant land cover class, additional and supporting data has revealed that the Tongaat Hulett Company and Development sector strives to maintain a balanced with its ecological surroundings by ensuring that suitable sustainable methods are used in the development process and in the maintenance of the area. To conclude, in accordance to the research produced from this study, it is evident that the urban development in Umhlanga and Umhlanga Ridge has grown tremendously and will continue to expand at a steady rate in the future, with the intention of meeting the demands of the area’s visiting tourists and permanent residents. However, while these continuous land use and land cover changes are taking place and expanding, it is imperative that a balanced relationship between man and nature be taken into consideration.
An assessment of the heavy metal concentrations in the water and sediment of the uMgeni estuary, using visible and near-infrared reflectance spectroscopy.
(2021) Ramcharan, Suvasha.; Gebreslasie, Michael Teweldemedhin.
Abstract available in PDF.
Climatic variable selection using random forests regression for malaria transmission modelling in Mpumalanga Province, South Africa.
(2015) Kapwata, Thandi.; Gebreslasie, Michael Teweldemedhin.
Malaria is one of the wold’s most prevalent vector borne diseases with sub-Saharan Africa bearing the highest burden of reported cases. Climate is one of the major determinant factors of malaria transmission as it influences the spatial and temporal pattern of transmission. It is therefore important to be able to understand the relationship between climatic variables and malaria transmission because an understanding of the interactions between them at a local level is an important part in potential outbreaks, targeting vector control strategies, and developing malaria early warning systems. This study covered the Ehlanzeni district of Mpumalanga province in South Africa. It was aimed at determining the climatic variable importance of temperature, lag temperature, rainfall, lag rainfall, humidity, altitude and NDVI in relation to malaria transmission. The random forest algorithm was used to relate the climatic variables extracted from remote sensing imagery and malaria case data collected from health facilities in order to establish individual measures of variable importance and to develop a spatial and temporal prediction models. In this study altitude appeared to be the most responsible variable for malaria transmission because it was most frequently selected as one of the top variables with the highest variable importance followed by NDVI and temperature. The combination of climatic variables that produced the highest coefficient of determination values was altitude, NDVI, and temperature. This suggests that these 3 variables have high predictive capabilities and as a result they should be selected for spatial and temporal modelling of malaria. Furthermore, it was expected that the predictive models generated by the random forest algorithm could be used as an operational malaria early warning system using forecast climatic variable identified in this study in order to assist in containing any potential reoccurrence of malaria after elimination.
Climatic, environmental and socio-economic factors for malaria transmission modelling in KwaZulu-Natal, South Africa.
(2018) Ebhuoma, Osadolor Obiahon.; Gebreslasie, Michael Teweldemedhin.
Sub-Saharan Africa (SSA) largely bears the burden of the global malaria disease, with the transmission and intensity influenced by the interaction of a variety of climatic, environmental, socio-economic, and human factors. Other factors include parasitic and vectoral factors. In South Africa (SA) in general and KwaZulu-Natal (KZN) in particular, the change of the malaria control intervention policy in 2000, may be responsible for the significant progress over the past two decades in reducing malaria case report to near zero. Currently, malaria incidence in KZN is less than 1 case per 1000 persons at risk placing the province in the malaria elimination stage. To meeting the elimination target, it is necessary to study the dynamics of malaria transmission in KZN employing various analytical/statistical models. Thus, the aim of this study was to explore the factors that influence malaria transmission by employing different analytical models and approaches in a setting with low malaria endemicity and transmission. This involves a sound appraisal of the existing literature on the contribution of remote sensing technology in understanding malaria transmission, evaluation of existing malaria control intervention; delineation of empirical map of malaria risk; provide information on the climatic, environmental and socio-economic factors that influences malaria risk and transmission; and formulation of a relevant malaria forecast and surveillance models. The investigator started with a systemic review of studies in chapter two. The studies were aimed at identifying significant remotely-sensed climatic and environmental determinants of malaria transmission for modelling malaria transmission and risk in SSA via a variety of statistical approaches. Normalised difference vegetation index (NDVI) was identified as the most significant remotely-sensed climatic/environmental determinants of malaria transmission in SSA. Majority of the studies employed the generalised linear modelling approach compared to the Bayesian modelling approach. In the third chapter, malaria cases from the endemic areas of KZN with remotely-sensed climatic and environmental data were used to model the climatic and environmental determinants of malaria transmission and develop a malaria risk map in KZN. The spatiotemporal zero inflated Poisson model formulated indicates that at 95% Bayesian credible interval (BCI) NDVI (0.91; 95% BCI = 0.71, -1.12), precipitation (0.11; 95% BCI = 0.08, 0.14), elevation (0.05; 95% BCI = 0.032, 0.07) and night temperature (0.04; 95% BCI = 0.03, 0.04) are significantly related to malaria transmission in KZN, SA. The area with the highest risk of malaria morbidity in KZN was identified as the north-eastern part of the province. The fourth chapter was to establish the socio-economic status (SES) that influence malaria transmission in the endemic areas of KZN, by employing a Bayesian inference approach. The obtained posterior samples revealed that, significant association existed between malaria disease and low SES such as illiteracy, unemployment, no toilet facilities and no electricity at 95% BCI Lack of toilet facilities (odds ration (OR) =12.54; 95% BCI = 0.63, 24.38) exhibited the strongest association with malaria and highest risk of malaria disease. This was followed by no education (OR =11.83; 95% BCI = 0.54, 24.27) and lack of electricity supply (OR =10.56; 95% BCI = 0.43, 23.92) respectively. In the fifth chapter, the seasonal autoregressive integrated moving average (SARIMA) intervention time series analysis (ITSA) was employed to model the effect of the malaria control intervention, dichlorodiphenyltrichloroethane (DDT) on confirmed monthly malaria cases. The result is an abrupt and permanent decline of monthly malaria cases (w0= −1174.781, p-value = 0.003) following the implementation of the intervention policy. Finally, the sixth chapter employed a SARIMA modelling approach to predict malaria cases in the endemic areas of KZN. Three plausible models were identified, and based on the goodness of fit statistics and parameter estimation, the SARIMA (0,1,1) (0,1,1)12 model was identified as the best fit model. The SARIMA (0,1,1)(0,1,1)12 model was used to forecast malaria cases during 2014, and it was observed to fit closely with the reported malaria cases during January to December 2014. The models generated in this study demonstrated the need for the KZN malaria program, relevant policy makers and stakeholders to further strengthen the KZN malaria elimination efforts. The required malaria elimination fortification are not limited to the implementation of additional sustainable developmental approach that combines both improved malaria intervention resources and socio-economic conditions, strengthening of existing community health workers, and strengthening of the already existing cross-border collaborations. However, more studies in the area of statistical modelling as well as practical applications of the generated models are encouraged. These can be accomplished by exploring new avenues via cross-sectional survey to understand the impact of community and social related structures in malaria burden; strengthening of existing community health workers; knowledge, attitude and practices in malaria control and intervention; and the likely effects of temporal/seasonal and spatial variations of malaria incidence in neighbouring endemic countries should be explored.
Designing and implementing a GIS-based cadastral database for land administration in the city of Asmara, Eritrea.
(2004) Gebreslasie, Michael Teweldemedhin.; Ahmed, Fethi B.
The knowledge and application of land information and GIS in Eritrea are very limited and as a result there is a shortage of sufficient, reliable, up-to-date and modem land information for decision-making. This study aimed to develop and design a GIS-based cadastral database for land administration in the city of Asmara, Eritrea. The two components of the cadastral data, the spatial and non-spatial were collected and processed in a GIS environment. GIS-based cadastral data was created to store the data. A recently acquired IKONOS image and existing Auto CAD data were the main sources of data for the study. Geo-rectification, conversion, ,building of topology, spatial adjustment, and digitizing were used' for creating the parcels and roads of Asmara city. A key of relation was created to link both the spatial and non-spatial components of the cadastre. The study used illustrated practical examples to show how GIS-based cadastral data could support land administration as practiced by the different divisions of the Municipal Office of Asmara city. The importance and usage of the cadastral database for urban planners and property valuators were detailed. Spatial and network analysis were used to develop bands for creating a banded property appraisal system for assessing the service catchment area~ of hospitals and the only fire station. Further, a location allocation model was <,I:lso developed to recommend suitable sites for new fire stations in the city of Asmara. It is recommended that the Municipal Office of Asmara adopts the designed GIS-based database. It is also recommended there the major cities in Eritrea follow similar methodology to design and implement cadastre database for their administration.
The estimation of Eucalyptus plantation forest structural attributes using medium and high spatial resolution satellite imagery.
(2008) Gebreslasie, Michael Teweldemedhin.; Ahmed, Fethi B.; van Aardt, Jan.
Sustaining the socioeconomic and ecological benefits of South African plantation forests is challenging. A more systematic and rapid forest inventory system is required by forest managers. This study investigates the utility of medium (ASTER 15 m) and high (IKONOS 1-4 m) spatial resolution satellite imageries in an effort to improve the remote capture of structural attributes of even-aged Eucalyptus plantations grown in the warm temperate climatic zone of southern KwaZulu-Natal, South Africa. The conversion of image data to surface reflectance is a pre-requisite for the establishment of relationships between satellite remote sensing data and ground collected forest structural data. In this study image-based atmospheric correction methods applied on ASTER and IKONOS imagery were evaluated for the purpose of retrieving surface reflectance of plantation forests. Multiple linear regression and canonical correlation analyses were used to develop models for the prediction of plantation forest structural attributes from ASTER data. Artificial neural networks and multiple linear regression were also used to develop models for the assessment of plantation forests structural attributes from IKONOS data. The plantation forest structural attributes considered in this study included: stems per hectare, diameter at breast height, mean tree height, basal area, and volume. In addition, location based stems per hectare were determined using high spatial resolution panchromatic IKONOS data where variable and fixed window sizes of local maxima were employed. The image-based dark object subtraction (DOS) model was better suited for atmospheric correction of ASTER and IKONOS imagery of the study area. The medium spatial resolution data were not amenable to estimating even-aged Eucalyptus forest structural attributes. It is still encouraging that up to 64 % of variation could be explained by using medium spatial resolution data. The results from high spatial resolution data showed a promising result where the ARMSE% values obtained for stems per hectare, diameter at breast height, tree height, basal area and volume are 7.9, 5.1, 5.8, 8.7 and 8.7, respectively. Results such as these bode well for the application of high spatial resolution imagery to forest structural assessment. The results from the location based estimation of stems per hectare illustrated that a variable window size approach developed in this study is highly accurate. The overall accuracy using a variable window size was 85% (RMSE of 189 trees per hectare). The overall findings presented in this study are encouraging and show that high spatial resolution imagery was successful in predicting even-aged Eucalyptus forest structural attributes in the warm temperate climates of South Africa, with acceptable accuracy.
How green is green? : a socio-spatial analysis of the status of green spaces within the eThekwini Municipality.
(2014) Pillay, Sarushen.; Naidoo, Sershen.; Bob, Urmilla.; Gebreslasie, Michael Teweldemedhin.
Globally, urbanisation is occurring at an alarming rate and urban green spaces are increasingly recognised as essential components in the quest to achieve sustainable urban landscapes. This study, which involved a socio-spatial analysis of the status of green spaces within the eThekwini Municipality (located in KwaZulu-Natal, South Africa), offers a unique opportunity in terms of urban conservation research. The objectives of the study were to examine the socio-economic characteristics and the perspectives of residents on the use and value of green spaces within the eThekwini Municipality using areas surrounding the Bluff Conservancy (all situated within the SDA) as illustrative examples; to develop a spatial representation of the quality/ integrity of selected green spaces within the eThekwini Municipality in relation to land-use patterns; to examine the appropriateness of the typology presently used by the eThekwini Municipality to describe the status of green spaces and to compare the same with Adapted typologies in order to determine the level of deviation; and lastly, to generate recommendations on the conservation and management of these green spaces. A variety of socio-spatial analysis methods were used to collect and analyse primary data. Data was obtained using Geographical Information System mapping and a questionnaire in order to ascertain resident perceptions towards their surrounding green spaces. Thereafter, secondary spatial data acquired from the eThekwini Municipality was processed and subjected to a range of analyses to evaluate the efficacy of the typology presently used by the Municipality to assess the quality/ integrity of green spaces. Six random green space types (settlement, tree crops, woodland, forest, grassland and thicket) were selected and first examined using the eThekwini typology and thereafter with the Adapted typology, developed as part of this study. The results suggested that almost all respondents (75.50%) frequently utilised green spaces in their community, with most respondents favouring the use of recreational and social green spaces (for example, parks, sports field and the golf course). However, respondents also identified numerous challenges associated with accessing and using green spaces; crime, pollution and lack of maintenance in particular, were shown to hamper the optimal use and integrity of a number of green spaces. Additionally, it was found that respondents use of green spaces was not dependent on their gender and income but was significantly influenced by their education. Furthermore, though most respondents indicated that they frequently engage in environmentally-friendly practices, only a small proportion of respondents (9.75%) were aware of the Durban Metropolitan Open Space System (which is a programme that formally allows for the creation and preservation of green spaces). In terms of the spatial analyses, the results revealed that selected green spaces within the Municipality when classified using a more discriminatory typology (Adapted typology), can be shown to contain micro-habitats that are either more degraded or more intact than that reflected by the typology presently used by the eThekwini Municipality. It was found that the five thicket green space sites assessed using the eThekwini typology collectively deviated by approximately 60% from that assessed using the more discriminatory Adapted typology. Overall, it was evident that quality based land cover differed minimally to moderately when selected green space types were compared using the two typologies. This resulted in some green micro-habitats within larger green spaces being potentially misclassified in terms of their ecological integrity when using the eThekwini typology and, possibly not being prioritised for conservation and/ or restoration. The combination of social and spatial results obtained and interpreted in this study was used to generate recommendations for the conservation and management of green spaces within the eThekwini Municipality. Evidence from the social survey clearly showed that respondents expressed a willingness and desire to have and use green spaces. Therefore, it is recommended that the eThekwini Municipality increase the number of green spaces, preferably within densely populated communities as well as improve existing greenery within the Municipality. In addition, these areas should be made more accessible and useable and have value added benefits to communities who are intrinsically supporting them. Furthermore, it was found that the current typology used for the classification of green spaces within the eThekwini Municipality is not discriminative enough to allow for effective management and conservation. This suggests the need for a more nuanced classification of green spaces within the Municipality which ensures that quality characteristics are adequately incorporated into the assessment of these environments.
Identifying opportunities for low carbon emission zones in South Africa : a case study of Durban.
(2015) Jagarnath, Meryl.; Thambiran, Tirusha.; Gebreslasie, Michael Teweldemedhin.
There is increasing attention on emissions reduction strategies that also deliver developmental co-benefits (i.e. low carbon development), especially in developing cities, thus research on the links between emissions, spatial planning, and urban development are emerging. The majority of studies on emissions inventories lack integration with strategic spatial planning, which is critical for place-based mitigation strategies. In response to this gap, a bottom-up methodological framework for the spatial representation of emissions was developed, based on the consumption perspective, to identify high emission zones and assess their urban development goals. The framework was applied to Durban (eThekwini Municipality), which aims to become a low carbon city and is also representative of a developing city. The total emissions calculated for Durban in 2013, was 12 219 118 tCO₂e, of which the road transport sector contributed the most to total emissions (43%), followed by industry electricity consumption (30%) A high emissions zone was identified along the coast, from Durban south, through the central business district (CBD) and the north to Umhlanga. Specifically, the areas with the highest emissions activities are from energy-intensive manufacturing industries in south Durban, and road transport, specifically private passenger cars, in central and north Durban. Furthermore, the highest emitting area, Prospecton, (767 172 tCO₂e), emitted ~ 6.5 times more than the Durban ward average (118 632 tCO₂e). Furthermore, Prospecton is highlighted for further port, fuel, chemical and petrochemicals, transport equipment manufacturing, and logistics development. The lowest emissions were from the rural edges, where the neighbourhoods emitted ~11 times less than the Durban average, which are also the areas with the most developmental needs, therefore highlighting the spatial disparity in emissions contribution within the city. A three-pronged approach of specific mitigation measures are recommended to simultaneously reduce emissions and achieve development: (i) manufacturing industries in south Durban must invest in carbon offset projects in the rural periphery to ensure that the development of those areas are not associated with increasing emissions, (ii) the implementation of car-free roads in central and north Durban to reduce distances travelled by private cars and to also ensure the widespread use of the Integrated Rapid Public Transport Network and other eco-mobility options, (iii) limit industrial expansion in south Durban and commercial and residential developments in north Durban which do not have a low carbon plan. Thus, the spatially-resolved emissions inventory generated emissions profiles which identified suitable mitigation strategies to assist with the transition to a low carbon city.
An investigation into the co-benefits of climate change mitigation and adaptation for the waste sector in the eThekwini Municipality.
(2015) Ngwenya, Nomdeni Simphiwe.; Thambiran, Tirusha.; Gebreslasie, Michael Teweldemedhin.
Abstract available in PDF file.
Long-term and climatological studies on sulphur dioxide (SO²) using ground based and space-borne measurements over South Africa.
(2018) Venkataraman, Sangeetha.; Gebreslasie, Michael Teweldemedhin.; Wright, Caradee Yale.
Abstract is available in the PDF file.
Mapping natural forest cover, tree species diversity and carbon stocks of a subtropical Afromontane forest using remote sensing.
(2021) Gyamfi-Ampadu, Enoch.; Gebreslasie, Michael Teweldemedhin.
Natural forests cover about a third of terrestrial landmass and provides benefits such as carbon sequestration, and regulation of biogeochemical cycles. It is essential that adequate information is available to support forest management. Remote Sensing imageries provide data for mapping natural forests. Hence, our study aimed at mapping the Nkandla Forest Reserve attributes with Remote Sensing imageries. Quantitative information on the forest attributes is non-existent for many of these forests, including the sub-tropical Afromontane Nkandla Forest Reserve. This does not support scientific and evidence based natural forest management. A review of literature revealed that progress has been made in Remote Sensing monitoring of natural forest attributes. The Random Forest (RF) and Support Vector Machine (SVM) were applied to Landsat 8 in classifying the land use land cover (LULC) classes of the forest. Each of the algorithms produced higher accuracy of above 95% with the SVM performing slightly better than the RF. The SVM, Markov Chain and Multi-Layer Perceptron Neural Network (MLPNN) were adopted for a spatiotemporal change detection over the last 30 years at decadal interval for the forest. There were consistent changes in each of the four LULC classes. The study further conducted a forecasting of LULC distribution for 2029. Aboveground carbon (AGC) estimation was carried out using Sentinel 2 imagery and RF modelling. Four models made up smade of Sentinel 2 products could successfully map the AGC with high accuracies. The last two studies focused on tree species diversity with the first evaluating the influence of spatial and spectral resolution on prediction accuracies by comparing the PlanetScope, RapidEye, Sentinel 2 and Landsat 8. Both the spatial and spectral resolution were found to influence accuracies with the Sentinel 2 emerging as the best imagery. The second aspect focused on identifying the best season for the prediction of tree species diversity. Summer imagery emerged as the best season and the winter being the least performer. Overall, our study indicates that Remote Sensing imageries could be used for successful mapping of natural forest attributes. The outputs of our studies could also be of interest to forest managers and Remote Sensing experts.
Modelling biomass of the rehabilitation forest around the Buffelsdraai landfill site using remote sensing data, Durban, South Africa.
(2017) Mkhabela, Nozipho Nokubongwa.; Gebreslasie, Michael Teweldemedhin.
Forests have important roles in ecosystem service provisions and maintenance of the global carbon cycle hence they are one of the main subjects of the Intergovernmental Panel on Climate Change which recommends strategies to stabilize greenhouse gas emissions. Remote sensing is an advancing science whose data products keep improving spectrally and spatially with time which makes them worth exploitation for broad scientific uses including forest-related studies such as biomass estimations. These are important for understanding of carbon sequestration potential of trees which informs monitoring and forest cover enhancement strategies across various environments. This study investigated the potential of optical data, Landsat 8 Operational Land Imager (OLI) to achieve biomass estimation in a secondary indigenous forest that buffers the Buffelsdraai landfill site. Image processing types used included extraction of spectral reflectance bands, vegetation indices and texture parameters. A Partial Least Squares analysis was performed to determine a significant set of independent variables that could predict aboveground biomass of the Buffelsdraai rehabilitation forest. The findings indicated that the Partial Least Squares models of bands and vegetation indices were rather weak in biomass prediction as only 11.22% and 30.88% biomass variation was explained, respectively. Models inclusive of texture extractions, however, performed much better and demonstrated an improved 77.33% variation explanation of above-ground biomass. Overall, the results indicate that texture parameters derived from Landsat 8 OLI optical data are effective to achieve improved biomass estimation. The development of allometric equations built directly from the species found in the rehabilitation zone and national instilment of environmental responsibility within society for improved local waste management were the major recommendations provided which would assist in the stabilization of greenhouse gas emissions in Buffelsdraai and South Africa.
Modelling terrain roughness using LiDAR derived digital terrain model in eucalyptus plantation forests, in KwaZulu-Natal, South Africa.
(2017) Munsamy, Roxanne.; Gebreslasie, Michael Teweldemedhin.; Ismail, Riyad Abdool Hak.
South African commercial plantation forests are established primarily to meet both the local and global demands of industries that require direct raw materials such as pulpwood or timber. Consequently, the commercial forest industry in South Africa is held in high esteem as it makes up one of the largest economic forces within the country. For this reason, individuals responsible for implementing strategies pertaining to silvicultural and harvesting operations within commercial plantations require up to date and detailed multi-forest inventory datasets to ensure that optimal yields are guaranteed and that sites are well maintained. Despite this, various drawbacks within commercial plantations exist: steep slopes, high elevations, and other forms of topographic irregularities, can affect the productivity of the site and impact mechanical silvicultural and harvesting operations. In lieu of making more informed and efficient decision-making protocols, forest researchers are often tasked with implementing and utilising alternative technologies such as remote sensing to determine if specific methodologies can be used for gathering multi-forest inventory data that also incorporate terrain information. Light Detection and Ranging (LiDAR), a recent remote sensing technology, has demonstrated that it is highly robust and can lend itself towards providing highly accurate vertical forest structural attributes and horizontal topographic derivatives. This study employs the use of a LiDAR derived Digital Terrain Model (DTM) (1 m x 1 m spatial resolution) to create terrain indices that are representative of the horizontal features within the commercial forest sites of interest. In addition, a machine learning approach using a random forest (RF) ensemble classifier was adopted to determine how much of the variation in forest structural attributes: mean dominant height, mean height, pulpwood volumes and diameter at breast height can be attributed to terrain when using the LiDAR derived DTM terrain variables. The overall findings presented in this study are encouraging and show that a LiDAR derived DTM can be successfully used for creating highly accurate terrain indices and can be used for predicting variability within even-aged Eucalyptus forest structural attributes within commercial plantation forests in KwaZulu-Natal, South Africa, with an acceptable level of accuracy.
An overview of the soil organic matter content present within the Emakhosaneni area, KwaZulu-Natal, South Africa using remote sensing technologies.
(2021) Sewpersad, Tessnika.; Gebreslasie, Michael Teweldemedhin.
Soil Organic Matter (SOM) is one of the fundamental constituents of soil and plays significant roles in the overall fertility, productivity, and quality of soil. It consists of decaying plant and animal material at various stages of decomposition, substances released by plant roots, and soil organisms. Additionally, it is responsible for supporting many physical, chemical, and biological functions within the soil. And, these functions influence the provision of ecosystem services to humans, plants, and animals. However, this SOM is under threat as 25% of the earth’s surface has become degraded, with 12 million hectares of topsoil being lost every year and, hence SOM. South Africa is one of those countries that are impacted by arable soil loss. Therefore, accurate measurement of SOM at different spatial scales is vital in providing information for planning a recovery strategy. However, traditional methods of soil analysis can be time-consuming, costly, and labour extensive. On the other hand, remote sensing is an efficient method that is time effective, low-cost, non-destructive, and has rapid data acquisition. Thus, offering an alternative to traditional methods of soil analysis. Hence, this research aimed to examine the SOM content within the Emakhosaneni area, KwaZulu-Natal, South Africa, by measuring the reflectance of soil using laboratory and remote sensing analysis. The first study examined the percentage of SOM content present within the study area, and its four major land uses using a laboratory technique. The results indicated that the area has a relatively low average percentage of SOM content (2.79%) present within its soils. Also, out of the four major land use types the agricultural land use had the highest average percentage of SOM content, followed by rangeland, built-up, and eroded land uses. These results further indicated the influence of land use activities on the SOM content within the study area. Overall, this study revealed the SOM content within the area is very low. It also highlighted the severity and consequences of the depleting SOM content within the area’s soils. The second study examined the relationship between the SOM content and spectral reflectance of soils within the study area using laboratory spectroscopy. Results suggested that the spectra obtained was influenced by soil colour and thus, established a relationship between SOM content and reflectance of soil samples as SOM content is linked to soil colour. Further assessment of this relationship by Partial Least Squares Regression analysis revealed a fair performance. With the models created using the pre-processed spectra and SOM content performing better than those with no pre-processed spectra. Overall, this study highlighted the importance and capability of Visible and Near Infrared (400 nm to 2400 nm) spectroscopy in examining SOM content compared to conventional laboratory approaches and its influence on the spectral reflectance of the soils. The third study assessed the relationship between SOM content and Sentinel-2 satellite remote sensing data of soils in the area. Both geostatistical methods and hybrid geostatistical methods were used to predict SOM content. Results showed that the hybrid geostatistical methods performed better than the geostatistical methods in predicting SOM content due to the contribution of auxiliary information. Therefore, this study emphasizes the potential of auxiliary remote sensing data such as Sentinel-2 imagery in predicting SOM content in KwaZulu-Natal, South Africa, while also making critical inferences regarding the spatial and temporal variability of SOM within an area. The overall findings presented in this research are encouraging and show that different remote sensing techniques can be successfully used in the estimation, assessment, and prediction of SOM content, especially within the Emakhosaneni area, KwaZulu-Natal, South Africa, with accuracy levels that are acceptable.
Quantitative comparison of the aerosol optical properties over Durban using ground and satellite based instrumentation.
(2017) Singh, Priyanka.; Venkataraman, Sivakumar.; Gebreslasie, Michael Teweldemedhin.
Aerosols are ubiquitous constituents in the atmosphere and are important for atmospheric processes. This is due to their ability to scatter and absorb solar radiation and influence cloud microphysics. This study will focus on discerning trends in aerosol optical properties in Durban (29.8587° S, 31.0218° E), a coastal city on the east coast of South Africa, using the preliminary results from the sun-photometer located at the University of KwaZulu-Natal. These results will also be compared to the well-established Skukuza sun-photometer. Skukuza is a rural agricultural area in the north eastern parts of South Africa. The Aerosol Optical Depth (AOD), Angstrom Exponent (α440–870), Columnar Water Vapour (CWV), Volume Size Distribution (VSD), Single Scattering Albedo (SSA), Asymmetry parameter (ASP), Real and Imaginary parts of the complex refractive index were studied for Durban and Skukuza. Analysis of the aerosol optical properties suggested that various sources of aerosols were identified for Durban, such as biomass burning, urban industrial aerosols and marine aerosols. Biomass burning aerosols impacted Skukuza during spring. There was a high extent of fine mode aerosols present throughout the year for Skukuza, indicating that urban industrial emissions from the South African Highveld region can also contribute to aerosol loads in the region. Preliminary results from the ground-based Durban sun-photometer was used to compare aerosol optical depth at 550 nm (AOD) to the satellite Moderate Resolution Imaging Spectroradiometer (MODIS) for the Aqua, Terra and Aqua and Terra combined (average of both) datasets for the dark target (DT) and deep blue (DB) retrieval algorithms to validate satellite retrievals. The results gave way to moderate correlations between MODIS Terra and the Durban sun-photometer for both DB (R2 = 0.70) and DT (R2 = 0.60), and between MODIS Aqua and the Durban sun-photometer for DB (0.68). Good correlations were observed for MODIS Terra and Aqua merged for both DB (0.79) and DT (0.74). The ability of MODIS to predict AOD was noted as dependent on the season and location. HYSPLIT 720 hour–backward trajectory analysis, AOD and α440–870 from the Durban sun-photometer, a Lidar profile and satellite imagery were used to determine if air mass from the Calbuco volcanic eruption in Chile in April 2015 reached Durban. Trajectory analysis found that only during May 2015, was air masses arriving from South America, within the 20 km altitude. This led to the assumption that stratospheric aerosols from the Calbuco volcano, travelled to Durban. Analysis of the AOD found that only during 2015 was a constant phenomenon driving AOD in Durban and this was attributed to the eruption. Lidar observations coupled with the backward trajectory analysis allowed for the identification of air masses in Durban arriving from the Calbuco volcano in Chile.
Review of environmental training practices in selected businesses in Durban.
(2015) Sennoga, Dianne.; Gebreslasie, Michael Teweldemedhin.; Ahmed, Fethi B.
Environmental management has moved from a policy concept to a proactive strategy defining business responsiveness to stakeholder and market-related pressures for more environmentally sustainable business practices. Paradoxically, the financial benefits accrued to businesses at the often externalised expense of environmental goods and services, is the very advantage that best positions it to respond to the environmental crisis. The importance of a systematic and proactive environmental response from the business community is compelled by the fact that environmental impacts are predominantly caused by errant pollutant and non-compliant business activities which is increasingly regulated through South African environmental legislation. The business response through corporate sustainability and environmental management is considered a sweeping change to business as usual. Increasing environmental regulations make the adoption of environmental management systems such as ISO 14001 more commonplace. In adapting to these changes in the workplace, it makes environmental training and awareness of employees a material avenue of investigation which further directs the aim of this study. In applying the ISO 14001 certification criterion, through a purposive and nonprobable sampling technique, twenty-four (24) Durban businesses have participated in this study. Similarly, in addition, fifteen (15) employees undergoing environmental training along with five (5) other role-players and stakeholders that relevantly bear on environmental training practices participated in this research, which was conducted through the use of survey questionnaires. The extent of adoption of environmental training and its effective reach across company structures has been assessed against seven (7) developed environmental training principles of this study. The selected businesses and other respondents in Durban show keen awareness, attitudes and perceptions regarding environmental training. Environmental training is a widely practiced activity across all the businesses sampled with topic coverage focussed predominantly on waste management, hazardous chemicals, and environmental auditing. The environmental training activities are largely combined with other Safety and Health priorities. Whilst this has no perceived negative impact on the content of environmental training, there is an indication that environmental training budget allocations are not effectively prioritised in combination with other training activities. The implementation of training across the company tiers shows executive levels in need of greater exposure to this activity. While the respondents predominantly showed limited satisfaction with environmental training received, various areas of improvement became clear such as greater management commitment, greater institutional assistance for clarity of training standards, course offerings and inter-industry collaboration in environmental training.
Socio-economic and environmental deteminants of malaria in four malaria endemic provinces of Zambia.
(2015) Shimaponda, Nzooma Munkwangu.; Mukaratirwa, Samson.; Gebreslasie, Michael Teweldemedhin.; Tembo-Mwase, Enala.
A large fraction of the global malaria burden occurs in sub-Saharan Africa and its endemicity depends on the interaction of environmental factors, vectors, parasites and the host. In Zambia, the negative effect of the break in interventions experienced in the late 2000s varied by regions. Therefore, it was necessary to determine the malaria determinants through the study of: statistical models that have been employed; knowledge of the community in malaria management and control; prevalence of malaria and presence of social and community-related factors influencing malaria control in selected communities; contribution of other social and environmental determinants of malaria from the household point of view; and also socio-economic and climatic determinants of malaria at provincial level, in Zambia. This work was achieved through a number of methods beginning with a systematic review of studies that have identified socio-economic and eco-environmental determinants of malaria through the use of statistical models in malaria burden determination and prediction in southern Africa. We also conducted a cross-sectional survey employing a simple random sampling technique to administer questionnaires to 584 household heads from selected communities, on the following components: knowledge, attitude and practices in malaria control; the role of social and community-related structures in malaria burden and control; and water sources and practices as well as housing structures in relation to self-reported malaria infections. Malaria testing was also performed using a rapid diagnostic test (RDT) in 756 individuals sampled from the 584 households. The household-level data was analysed in STATA and WinBUGS whereas the provincial-level malaria cases, government socio-economic and remotely-sensed climatic data were analysed in STATA, WinBUGS and also in R- integrated Nested Laplace (R-INLA) The focus of the studies conducted in southern Africa reviewed, has mainly been on malaria determinants related to intervention strategies and climatic factors. Additionally, the use of Bayesian statistical modelling was quite low (29.2%) in the studies reviewed. The community knowledge study showed that although knowledge levels in malaria were high they were not interrelated with attitudes and practices. In the malaria testing survey, a higher infection rate was seen in children and the highest RDT malaria prevalence was recorded in communities of Luapula province. Relating malaria burden with the role of community health workers (CHWs) in malaria control, malaria prevalence was inversely related with CHWs presence in Western Province. On the other hand, relating malaria burden with water practices and housing structures, “river” as a water source was the main predictor. The Bayesian hierarchical (or Generalised Linear mixed model) and R-INLA based models showed that region on one hand and region, time and precipitation on the other, were strong predictors of malaria incidence. More research in the area of statistical modeling as well as in other areas such as behaviour change, strengthening of existing CHW and exploring new avenues with regards to community social structures and ecological and climatic factors by locality is a great need.
A statistical approach for modelling forest structural attributes using multispectral remote sensing data within a commercial forest plantation.
(2017) Reddy, Nicole.; Gebreslasie, Michael Teweldemedhin.; Ismail, Riyad Abdool Hak.
Abstract available in PDF file.