Bathymetry is the science to study and measure the depths of the ocean floor. The differences in water depth, underwater slope and ocean floor structure were investigated using a geographic information system (GIS). This article investigates changes to the hydrodynamic sedimentation processes in Saldanha Bay as a result of the harbour constructions that took place in the early 1970s. The construction of the harbour included the massive dredging operations and resultant relocation of 30 million m3 of soil. Bathymetric data from Saldanha Bay before (1957) and after (1977) the construction of the harbour in the early seventies were analysed and compared in ArcGIS. It was found that the overall depth of the inner part of Saldanha Bay increased with about 1.4 m and that the bottom and slopes were smoothed. The physical structure that was constructed for the harbour had a serious impact on the hydrodynamic sedimentation processes. It was found that these interventions altered the depth profile of the Bay and the article finally draws conclusions on coastal and beach formation, based on the change in the hydrodynamic sedimentation processes of Saldanha Bay.
{"title":"Determining the change in the bathymetry of Saldanha Bay due to the harbour construction in the seventies","authors":"Ivan Henrico, J. Bezuidenhout","doi":"10.4314/sajg.v9i2.16","DOIUrl":"https://doi.org/10.4314/sajg.v9i2.16","url":null,"abstract":"Bathymetry is the science to study and measure the depths of the ocean floor. The differences in water depth, underwater slope and ocean floor structure were investigated using a geographic information system (GIS). This article investigates changes to the hydrodynamic sedimentation processes in Saldanha Bay as a result of the harbour constructions that took place in the early 1970s. The construction of the harbour included the massive dredging operations and resultant relocation of 30 million m3 of soil. Bathymetric data from Saldanha Bay before (1957) and after (1977) the construction of the harbour in the early seventies were analysed and compared in ArcGIS. It was found that the overall depth of the inner part of Saldanha Bay increased with about 1.4 m and that the bottom and slopes were smoothed. The physical structure that was constructed for the harbour had a serious impact on the hydrodynamic sedimentation processes. It was found that these interventions altered the depth profile of the Bay and the article finally draws conclusions on coastal and beach formation, based on the change in the hydrodynamic sedimentation processes of Saldanha Bay.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"46129192","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Feature selection techniques are often employed for reducing data dimensionality, improving computational efficiency, and most importantly for selecting a subset of the most important features for model building. The present study explored the utility of a Filter-Wrapper (FW) approach for feature selection using terrestrial hyperspectral remote sensing imagery. The efficacy of the FW approach was evaluated in conjunction with the Random Forest (RF) and Extreme Gradient Boosting (XGBoost) classifiers, to discriminate between water-stressed and non-stressed Shiraz vines. The proposed FW approach yielded a test accuracy of 80.0% (KHAT = 0.6) for both RF and XGBoost, outperforming the more traditional Kruskal-Wallis (KW) filter by more than 20%. The FW approach was also less computationally expensive when compared with the more commonly used Sequential Floating Forward Selection (SFFS) wrapper. Additionally, we examined the effect of hyperparameter optimisation on classification accuracy and computational expense. The results showed that RF marginally outperformed XGBoost when using all wavebands (p = 176) and optimised hyperparameter values. RF yielded a test accuracy of 83.3% (KHAT = 0.67), whereas XGBoost yielded a test accuracy of 81.7% (KHAT = 0.63). Our results further show that optimising hyperparameter values yielded an overall increase in test accuracy, ranging from 0.8% to 5.0%, for both RF and XGBoost. Overall, the results highlight the effect of feature selection and optimisation on the performance of machine learning ensembles for modelling vineyard water stress.
{"title":"A feature selection approach for terrestrial hyperspectral image analysis","authors":"Kyle Loggenberg, Nitesh K. Poona","doi":"10.4314/sajg.v9i2.20","DOIUrl":"https://doi.org/10.4314/sajg.v9i2.20","url":null,"abstract":"Feature selection techniques are often employed for reducing data dimensionality, improving computational efficiency, and most importantly for selecting a subset of the most important features for model building. The present study explored the utility of a Filter-Wrapper (FW) approach for feature selection using terrestrial hyperspectral remote sensing imagery. The efficacy of the FW approach was evaluated in conjunction with the Random Forest (RF) and Extreme Gradient Boosting (XGBoost) classifiers, to discriminate between water-stressed and non-stressed Shiraz vines. The proposed FW approach yielded a test accuracy of 80.0% (KHAT = 0.6) for both RF and XGBoost, outperforming the more traditional Kruskal-Wallis (KW) filter by more than 20%. The FW approach was also less computationally expensive when compared with the more commonly used Sequential Floating Forward Selection (SFFS) wrapper. Additionally, we examined the effect of hyperparameter optimisation on classification accuracy and computational expense. The results showed that RF marginally outperformed XGBoost when using all wavebands (p = 176) and optimised hyperparameter values. RF yielded a test accuracy of 83.3% (KHAT = 0.67), whereas XGBoost yielded a test accuracy of 81.7% (KHAT = 0.63). Our results further show that optimising hyperparameter values yielded an overall increase in test accuracy, ranging from 0.8% to 5.0%, for both RF and XGBoost. Overall, the results highlight the effect of feature selection and optimisation on the performance of machine learning ensembles for modelling vineyard water stress.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43361580","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
T. Idowu, R. Waswa, K. Lasisi, M. Nyadawa, V. Okumu
In the wake of the burgeoning population, socio-economic and environmental issues facing coastal areas, LULC change detection has become an essential tool for environmental monitoring towards achieving sustainable development. In this study, an object-based image analysis approach using post-classification comparison technique was applied for assessing the LULC of the coastal city of Lagos from 1986 to 2016. The study describes how satellite imagery from different sources (Landsat and SENTINEL 2A) can be successfully integrated for Land use Land cover change detection. The results show that between 1986 and 2016, there were net increases in bare areas, built-up areas, and shrublands and a general decline in forestlands, waterbodies and wetlands. Over 60,000ha cover (approx. 190% increase) was converted into built-up areas while 83,541ha (835.4km2) of forestland were lost, suggesting high rates of urbanization and corresponding deforestation. About 60% loss of wetlands was also observed in the same time period. The decrease in water bodies and a steady increase in bare and built-up areas are possibly due to the prevalent land reclamation activities in the study area. Higher rates of deforestation and increase in bare areas were observed from 2001 to 2016 in comparison to 1986 to 2001. The observed trends are likely to continue, and for future management actions, predictive studies are suggested to provide more empirical evidence.
{"title":"Object-based land use/land cover change detection of a coastal city using Multi-Source Imagery: a case study of Lagos, Nigeria","authors":"T. Idowu, R. Waswa, K. Lasisi, M. Nyadawa, V. Okumu","doi":"10.4314/sajg.v9i2.10","DOIUrl":"https://doi.org/10.4314/sajg.v9i2.10","url":null,"abstract":"In the wake of the burgeoning population, socio-economic and environmental issues facing coastal areas, LULC change detection has become an essential tool for environmental monitoring towards achieving sustainable development. In this study, an object-based image analysis approach using post-classification comparison technique was applied for assessing the LULC of the coastal city of Lagos from 1986 to 2016. The study describes how satellite imagery from different sources (Landsat and SENTINEL 2A) can be successfully integrated for Land use Land cover change detection. The results show that between 1986 and 2016, there were net increases in bare areas, built-up areas, and shrublands and a general decline in forestlands, waterbodies and wetlands. Over 60,000ha cover (approx. 190% increase) was converted into built-up areas while 83,541ha (835.4km2) of forestland were lost, suggesting high rates of urbanization and corresponding deforestation. About 60% loss of wetlands was also observed in the same time period. The decrease in water bodies and a steady increase in bare and built-up areas are possibly due to the prevalent land reclamation activities in the study area. Higher rates of deforestation and increase in bare areas were observed from 2001 to 2016 in comparison to 1986 to 2001. The observed trends are likely to continue, and for future management actions, predictive studies are suggested to provide more empirical evidence.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43469870","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
C. D. Villiers, C. Munghemezulu, G. Chirima, Philemon Tsele, Zinhle Mashaba
Invasive alien plants (IAPs) are responsible for loss in biodiversity and the depletion of water resources in natural ecosystems. Prosopis species are IAPs previously introduced by farmers to provide shade and fodder for livestock. In the Northern Cape, Prosopis spp. invasions are associated with the loss of native species resulting in overgrazing and degrading rangelands. Mapping Prosopis glandulosa is essential for management initiatives to assist the government in minimising the spread and impact of IAPs. This study aims to evaluate the performance of two machine learning algorithms i.e., Support Vector Machine (SVM) and Random Forest (RF) to map the spatial dynamics of P. glandulosa in Prieska. The spatial invasion extent of P. glandulosa was mapped using multitemporal Landsat data spanning the period from 1990 to 2018. Validation of the results was done through an estimated error matrix with the use of the proportion of area and the estimates of overall accuracy, user’s accuracy and producer’s accuracy with a 95% confidence interval. The performance of the SVM and RF classifiers showed similar results in the overall accuracy and Kappa statistics throughout the years. These methods showed an overall increase of at least 3.3% of the area invaded by P. glandulosa from 1990 to 2018. The study indicates the importance of Landsat imagery for mapping historical and current land cover change of IAPs. The spread of P. glandulosa was confirmed by an increase in the total area of invasion, which enables decision-makers to improve monitoring and eradication initiatives.
{"title":"Machine learning algorithms for mapping Prosopis glandulosa and land cover change using multi-temporal Landsat products: a case study of Prieska in the Northern Cape Province, South Africa","authors":"C. D. Villiers, C. Munghemezulu, G. Chirima, Philemon Tsele, Zinhle Mashaba","doi":"10.4314/SAJG.V9I2.13","DOIUrl":"https://doi.org/10.4314/SAJG.V9I2.13","url":null,"abstract":"Invasive alien plants (IAPs) are responsible for loss in biodiversity and the depletion of water resources in natural ecosystems. Prosopis species are IAPs previously introduced by farmers to provide shade and fodder for livestock. In the Northern Cape, Prosopis spp. invasions are associated with the loss of native species resulting in overgrazing and degrading rangelands. Mapping Prosopis glandulosa is essential for management initiatives to assist the government in minimising the spread and impact of IAPs. This study aims to evaluate the performance of two machine learning algorithms i.e., Support Vector Machine (SVM) and Random Forest (RF) to map the spatial dynamics of P. glandulosa in Prieska. The spatial invasion extent of P. glandulosa was mapped using multitemporal Landsat data spanning the period from 1990 to 2018. Validation of the results was done through an estimated error matrix with the use of the proportion of area and the estimates of overall accuracy, user’s accuracy and producer’s accuracy with a 95% confidence interval. The performance of the SVM and RF classifiers showed similar results in the overall accuracy and Kappa statistics throughout the years. These methods showed an overall increase of at least 3.3% of the area invaded by P. glandulosa from 1990 to 2018. The study indicates the importance of Landsat imagery for mapping historical and current land cover change of IAPs. The spread of P. glandulosa was confirmed by an increase in the total area of invasion, which enables decision-makers to improve monitoring and eradication initiatives.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43790294","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Spatial Data Infrastructure (SDI) provides a platform for spatial data sharing and is a key for sustainable development. Developing countries, including Tanzania, are at different stages of implementing SDIs. The importance and advantage of implementation lie in the fact that considerable funds can be saved by avoiding duplication of data, and improving quality of decisions making as well as public services. However, SDI is very complex in nature, including many influencing factors and different stakeholders. This paper investigates the possibilities of using Agent-Based Modelling (ABM) for simulating an SDI development process in Tanzania, for better understanding and making better planning. The roles and actions of organizations were identified through interviews, and the results were analysed. The behaviour of individual organizations (stakeholders) while interacting with the system were observed and analysed. The growth results in terms of data availability, standards, and data sharing for each organization were plotted and priority tables were generated. The model was evaluated for consistency and the results were judged to be within a reasonable range. The ABM simulation depicted the main attributes of agents, their roles and their interactions while pursuing SDI development in Tanzania. The results will help SDI planners and stakeholders to understand the roles of partners and prioritize activities and actions for successful SDI implementation.
{"title":"Investigating an agent based modelling approach for SDI planning: A case study of Tanzania NSDI development","authors":"Alex Lubida, M. Rajabi, P. Pilesjö, A. Mansourian","doi":"10.4314/SAJG.V9I2.14","DOIUrl":"https://doi.org/10.4314/SAJG.V9I2.14","url":null,"abstract":"Spatial Data Infrastructure (SDI) provides a platform for spatial data sharing and is a key for sustainable development. Developing countries, including Tanzania, are at different stages of implementing SDIs. The importance and advantage of implementation lie in the fact that considerable funds can be saved by avoiding duplication of data, and improving quality of decisions making as well as public services. However, SDI is very complex in nature, including many influencing factors and different stakeholders. This paper investigates the possibilities of using Agent-Based Modelling (ABM) for simulating an SDI development process in Tanzania, for better understanding and making better planning. The roles and actions of organizations were identified through interviews, and the results were analysed. The behaviour of individual organizations (stakeholders) while interacting with the system were observed and analysed. The growth results in terms of data availability, standards, and data sharing for each organization were plotted and priority tables were generated. The model was evaluated for consistency and the results were judged to be within a reasonable range. The ABM simulation depicted the main attributes of agents, their roles and their interactions while pursuing SDI development in Tanzania. The results will help SDI planners and stakeholders to understand the roles of partners and prioritize activities and actions for successful SDI implementation.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48142849","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2020-07-23DOI: 10.20944/preprints202007.0560.v1
T. Idowu, R. Waswa, K. Lasisi, Kenneth Mubea, M. Nyadawa, J. Kiema
The most extensive urban growths in the next 30 years are expected to occur in developing countries. Lagos, Nigeria - Africa’s second most populous megacity- is a prime example. To achieve more sustainable and resilient cities, there is a need for modeling the urban growth patterns of major cities and analyzing their implications. In this study, the urban growth of Lagos state was modeled using the Multi-Layer Perceptron (MLP) neural network for the transition modeling and the Markov Chain analysis for the change prediction, achieving a model accuracy of 81.8%. An innovative visual validation of the model results using the ArcGIS was combined with kappa correlation statistics. The results show that by 2031, built-up areas will be the most spatially extensive LULC class in the study area with percentage coverage of 34.1% as opposed to 9% in 1986. The coverage of bare areas is also expected to increase by 53% between 2016 and 2031. Conversely, 24.9% and 68.3% loss of forestlands and wetlands respectively, are expected between 2016 and 2031. In view of the 11th goal of SDGs which focuses on achieving sustainable cities and communities, the objectives of African Union’s Agenda 2063, and based on the urban growth trends observed, the study recommends a prioritization of vertical expansion as opposed to the current horizontal urban growth trends in the study area.
{"title":"Towards Achieving Sustainability of Coastal Environments: Urban Growth Analysis and Prediction of Lagos, State Nigeria","authors":"T. Idowu, R. Waswa, K. Lasisi, Kenneth Mubea, M. Nyadawa, J. Kiema","doi":"10.20944/preprints202007.0560.v1","DOIUrl":"https://doi.org/10.20944/preprints202007.0560.v1","url":null,"abstract":"The most extensive urban growths in the next 30 years are expected to occur in developing countries. Lagos, Nigeria - Africa’s second most populous megacity- is a prime example. To achieve more sustainable and resilient cities, there is a need for modeling the urban growth patterns of major cities and analyzing their implications. In this study, the urban growth of Lagos state was modeled using the Multi-Layer Perceptron (MLP) neural network for the transition modeling and the Markov Chain analysis for the change prediction, achieving a model accuracy of 81.8%. An innovative visual validation of the model results using the ArcGIS was combined with kappa correlation statistics. The results show that by 2031, built-up areas will be the most spatially extensive LULC class in the study area with percentage coverage of 34.1% as opposed to 9% in 1986. The coverage of bare areas is also expected to increase by 53% between 2016 and 2031. Conversely, 24.9% and 68.3% loss of forestlands and wetlands respectively, are expected between 2016 and 2031. In view of the 11th goal of SDGs which focuses on achieving sustainable cities and communities, the objectives of African Union’s Agenda 2063, and based on the urban growth trends observed, the study recommends a prioritization of vertical expansion as opposed to the current horizontal urban growth trends in the study area.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-07-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43579694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
N. Mudau, W. Mapurisa, Thomas Tsoeleng, Morwapula Mashalane
This study investigated the automation of the building extraction using SPOT 6 satellite imagery. The proposed methodology uses variance textural information derived from 1.5m panchromatic image to detect built-up areas from non-built-up areas. Once detected, detailed segmentation is performed on built-up class to create individual building objects. Canny edges, SAVI and spectral properties of the objects were used to classify building structures from other land use features using a thresholding technique. The methodology was tested in different areas including formal, rural village and informal and new development settlement types without modifying segmentation and classification parameters. The proposed methodology successfully detected built-up from non built-up areas in all different settlement types. The detection of individual structures achieved more than 70% in formal, rural village and new development areas while less than 50% of building structures in informal settlement were detected. The proposed method can contribute towards monitoring of human settlement developments over a larger area which is vital during spatial planning, service delivery and environmental management. This work will contribute towards the development of a National Human Settlement Layer developed and maintained by SANSA.
{"title":"Towards development of a national human settlement layer using high resolution imagery: a contribution to SDG reporting","authors":"N. Mudau, W. Mapurisa, Thomas Tsoeleng, Morwapula Mashalane","doi":"10.4314/sajg.v9i1.1","DOIUrl":"https://doi.org/10.4314/sajg.v9i1.1","url":null,"abstract":"This study investigated the automation of the building extraction using SPOT 6 satellite imagery. The proposed methodology uses variance textural information derived from 1.5m panchromatic image to detect built-up areas from non-built-up areas. Once detected, detailed segmentation is performed on built-up class to create individual building objects. Canny edges, SAVI and spectral properties of the objects were used to classify building structures from other land use features using a thresholding technique. The methodology was tested in different areas including formal, rural village and informal and new development settlement types without modifying segmentation and classification parameters. The proposed methodology successfully detected built-up from non built-up areas in all different settlement types. The detection of individual structures achieved more than 70% in formal, rural village and new development areas while less than 50% of building structures in informal settlement were detected. The proposed method can contribute towards monitoring of human settlement developments over a larger area which is vital during spatial planning, service delivery and environmental management. This work will contribute towards the development of a National Human Settlement Layer developed and maintained by SANSA.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48797197","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Grassland biomes are valuable socio-economic and ecological resources. However, the invasion of grasslands by alien plant species has emerged as one of the biggest threats to their sustainability, management and conservation. Timely, cost-effective and accurate determination of invasive alien plant spatial distribution is paramount for mitigating the adverse effects of alien plants on natural grasslands. Whereas literature on use of optical bands for invasive alien plants detection and mapping is abound, there is paucity in literature on the integration of Vegetation Indices (VIs) and optical reflectance bands in invasive species mapping. Specifically, there is need to test the efficacy of improved and freely available sensors like Sentinel-2 in understanding landscape invasion. Hence, this study sought to assess the efficacy of Sentinel-2’s optical bands and VIs for improving the mapping of American Bramble (Rubus cuneifolius) within a grassland biome. Variable Importance in the Projection (VIP) was used to identify the most influential reflectance bands and VIs, which were then fused at a feature level to determine Bramble spatial distribution. To determine the optimal season for Bramble mapping, seasonal classification accuracies were executed in Support Vector Machine (SVM) learning algorithm and accuracies for Spring, Summer, Autumn and Winter seasons compared. Results show that although the highest overall accuracy was achieved using only optical bands, fused imagery increased overall classification accuracies during spring and autumn i.e. 70% to 73% and 63% to 65%, respectively. However, the fused imagery failed to improve on the benchmark of optical imagery during summer and winter. Findings from this study underline the efficacy of complementing VIs and optical bands in determining the distribution of invasive species within grasslands at specific seasons. Furthermore, this study advocates for the adoption and fusion of freely available new generation satellite imagery such as Sentinel-2 as a cost effective option in landscape mapping.
{"title":"Assessing the synergistic potential of Sentinel-2 spectral reflectance bands and derived vegetation indices for detecting and mapping invasive alien plant species","authors":"J. Odindi, O. Mutanga, Perushan Rajah","doi":"10.4314/sajg.v9i1.6","DOIUrl":"https://doi.org/10.4314/sajg.v9i1.6","url":null,"abstract":"Grassland biomes are valuable socio-economic and ecological resources. However, the invasion of grasslands by alien plant species has emerged as one of the biggest threats to their sustainability, management and conservation. Timely, cost-effective and accurate determination of invasive alien plant spatial distribution is paramount for mitigating the adverse effects of alien plants on natural grasslands. Whereas literature on use of optical bands for invasive alien plants detection and mapping is abound, there is paucity in literature on the integration of Vegetation Indices (VIs) and optical reflectance bands in invasive species mapping. Specifically, there is need to test the efficacy of improved and freely available sensors like Sentinel-2 in understanding landscape invasion. Hence, this study sought to assess the efficacy of Sentinel-2’s optical bands and VIs for improving the mapping of American Bramble (Rubus cuneifolius) within a grassland biome. Variable Importance in the Projection (VIP) was used to identify the most influential reflectance bands and VIs, which were then fused at a feature level to determine Bramble spatial distribution. To determine the optimal season for Bramble mapping, seasonal classification accuracies were executed in Support Vector Machine (SVM) learning algorithm and accuracies for Spring, Summer, Autumn and Winter seasons compared. Results show that although the highest overall accuracy was achieved using only optical bands, fused imagery increased overall classification accuracies during spring and autumn i.e. 70% to 73% and 63% to 65%, respectively. However, the fused imagery failed to improve on the benchmark of optical imagery during summer and winter. Findings from this study underline the efficacy of complementing VIs and optical bands in determining the distribution of invasive species within grasslands at specific seasons. Furthermore, this study advocates for the adoption and fusion of freely available new generation satellite imagery such as Sentinel-2 as a cost effective option in landscape mapping.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44700150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
LiDAR intensity and texture features have reported high accuracies for discriminating forest species, particularly with the utility of the random forest (RF) algorithm. To date, limited studies has utilized LiDAR-derived forest gap information to assist in forest species discrimination. In this study, LiDAR intensity and texture features were extracted from forest canopy gaps to discriminate Eucalyptus grandis and Eucalyptus dunnii within a forest plantation. Additionally, LiDAR intensity and texture information was extracted for both canopy gaps and forest canopy and utilized for species discrimination. Using LiDAR intensity and texture information extracted for both canopy gap and forest canopy, resulted in a model accuracy of 94.74% (KHAT = 0.88). Using only canopy gap information, the RF model obtained an overall accuracy of 90.91% (KHAT = 0.81). The results highlight the potential for using canopy gap information for commercial species discrimination and mapping.
{"title":"Modelling forest species using LiDar-derived metrics of forest canopy gaps","authors":"L. Lombard, R. Ismail, Nitesh K. Poona","doi":"10.4314/sajg.v9i1.3","DOIUrl":"https://doi.org/10.4314/sajg.v9i1.3","url":null,"abstract":"LiDAR intensity and texture features have reported high accuracies for discriminating forest species, particularly with the utility of the random forest (RF) algorithm. To date, limited studies has utilized LiDAR-derived forest gap information to assist in forest species discrimination. In this study, LiDAR intensity and texture features were extracted from forest canopy gaps to discriminate Eucalyptus grandis and Eucalyptus dunnii within a forest plantation. Additionally, LiDAR intensity and texture information was extracted for both canopy gaps and forest canopy and utilized for species discrimination. Using LiDAR intensity and texture information extracted for both canopy gap and forest canopy, resulted in a model accuracy of 94.74% (KHAT = 0.88). Using only canopy gap information, the RF model obtained an overall accuracy of 90.91% (KHAT = 0.81). The results highlight the potential for using canopy gap information for commercial species discrimination and mapping.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43099224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The production of land cover maps using supervised classification algorithms is one of the most common applications of remote sensing. In this study, the effectiveness of supervised classification algorithms in land cover classification using ASTER data was evaluated in the Mankweng Area and its environs. The false colour composite image generated from combination of band 1, 2 and 3 in red, green and blue, respectively, was used to generate training classes for six land cover types (waterbody, forest, vegetation, Duiwelskloof leucogranite, Turfloop granite and built-up land). These were used to construct land cover maps using eight supervised classification algorithms: Maximum Likelihood, Minimum Distance, Support Vector Machine, Mahalanobis Distance, Parallelepiped, Neural Network, Spectral Angle Mapper and Spectral Information Divergence. To evaluate the effectiveness of the algorithms, the land cover maps were subjected to accuracy assessment to determine precision of the algorithms in accurately classifying the land cover types and level of confidence that can be attributed to the land cover maps. Most algorithms poorly performed in classifying spatially overlapping land cover types without abrupt boundaries. This indicates that the environmental conditions and distribution of land cover types can affect the performance of certain classification algorithms, and thus need to be considered prior to selection of algorithms. However, Support Vector Machine and Minimum Distance proved to be the two most effective algorithms as they provided better producer’s and user’s accuracy in the range of 80-100% for all land cover types, which represent good classification.
{"title":"Evaluation of effectiveness of supervised classification algorithms in land cover classification using ASTER images-A case study from the Mankweng (Turfloop) Area and its environs, Limpopo Province, South Africa","authors":"Nndanduleni Muavhi","doi":"10.4314/sajg.v9i1.5","DOIUrl":"https://doi.org/10.4314/sajg.v9i1.5","url":null,"abstract":"The production of land cover maps using supervised classification algorithms is one of the most common applications of remote sensing. In this study, the effectiveness of supervised classification algorithms in land cover classification using ASTER data was evaluated in the Mankweng Area and its environs. The false colour composite image generated from combination of band 1, 2 and 3 in red, green and blue, respectively, was used to generate training classes for six land cover types (waterbody, forest, vegetation, Duiwelskloof leucogranite, Turfloop granite and built-up land). These were used to construct land cover maps using eight supervised classification algorithms: Maximum Likelihood, Minimum Distance, Support Vector Machine, Mahalanobis Distance, Parallelepiped, Neural Network, Spectral Angle Mapper and Spectral Information Divergence. To evaluate the effectiveness of the algorithms, the land cover maps were subjected to accuracy assessment to determine precision of the algorithms in accurately classifying the land cover types and level of confidence that can be attributed to the land cover maps. Most algorithms poorly performed in classifying spatially overlapping land cover types without abrupt boundaries. This indicates that the environmental conditions and distribution of land cover types can affect the performance of certain classification algorithms, and thus need to be considered prior to selection of algorithms. However, Support Vector Machine and Minimum Distance proved to be the two most effective algorithms as they provided better producer’s and user’s accuracy in the range of 80-100% for all land cover types, which represent good classification.","PeriodicalId":43854,"journal":{"name":"South African Journal of Geomatics","volume":" ","pages":""},"PeriodicalIF":0.5,"publicationDate":"2020-02-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43041444","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：