Maximum Likelihood Classification Algorithm Research Articles

Assessing the performance of the multi-morphological profiles in urban land cover mapping using pixel based classifiers and very high resolution satellite imagery

Improved spatial and spectral resolution from recent sensor advancements provides opportunities for detailed and enhanced accuracies in the classification of heterogeneous urban landscapes. However, to date, such classifications have remained a challenge for most pixel based techniques. Whereas object based techniques have proved effective in classifying heterogeneous urban landscapes, by providing an effective framework for analysis of high spatial resolution images, challenges such as under and over-segmentation and non-robust statistical estimation impede their optimum performance in the often complex urban landscapes. Therefore, it is imperative that an effective classification approach is identified for effective utilisation of both spatio-spectral characteristics of image objects. Morphological techniques, especially multi-morphological profiles (MMP), provide an effective framework for the analysis of both spectral and spatial information from very high-resolution satellite imagery by performing image analysis based on features such as geometric, texture and contrast. In this study, using Support Vector Machine (SVM) and Maximum Likelihood (ML) classification algorithms, we compare the classification accuracies based on MMP as feature vector against those without MMP as a feature vector. Results from this study indicate that the use of MMP as a feature vector produced significantly higher classification accuracies of 84.8% and 82.2%, compared to 75.77% and 77.6% without MMP as a feature vector for SVM and ML, respectively. The study concludes that MMP can be used as a feature vector to increase the classification accuracy of a heterogeneous urban land use land cover.

Spatio-temporal Analysis of Urban Growth and Its Effects on Wetlands in Rwanda: The Case of Rwampara Wetland in the City of Kigali

This study aimed at analyzing the spatio-temporal patterns of urban growth and its effects on Rwampara wetland, located in the City of Kigali, Rwanda. First, the study was based on the application of remote sensing technology, where 4 Landsat images (1987, 1999, 2009 & 2018) were classified using maximum likelihood classification algorithm. This helped in analyzing the Land Use and Land Cover (LULC) trends in the study area. Secondly, it used the existing LULC data for the years 1990, 2000, 2010 and 2018 in order to investigate the overall changes in LULC in Kigali City. Finally, semi-structured interviews were used to collect data from local people and decision-makers about their past and future management strategies of Rwampara wetland. In this regard, 30 local communities (mainly natives from the study area), 15 local government authorities at sector and district levels as well as 4 senior government authorities in the central administration were interviewed. The findings revealed that over the past 4 decades, urban growth in Kigali City has rapidly increased at the expense of resource degradation in Rwampara wetland. Specifically, there has been an increase of about 77% of the built-up area over the last 31 years (1987-2018) which has led to the decrease of the wetland surface area from 24 ha in 1987 to only 7.7 ha in 2018. The results also revealed that demographic factors (i.e. a high population growth rate and high population densities) were mainly responsible for urban growth and degradation of wetland resources in the area under investigation.
 Keywords: urban growth, wetland, wetland resources, wetland degradation, wetland management

Predicting Forest Land Cover Changesin Ba Be National Park of Vietnam

The study is conducted in Ba Be National Park which is the specially protected natural area (SPNA) of Vietnam. The paper analyzes changes and makes prediction of forest land cover in National Park. For these purposes satellite images (Landsat-5 and Landsat-8) acquired for the year of 1992, 2003 and 2019 were used. The Normalized Difference Vegetation Index (NDVI) was used to identify vegetation quality. Forest land cover was classified by 5 categories using maximum likelihood classifier algorithm. In order to detect and evaluate forest land cover change, supervised classification and image differencing method are applied. Then, Cellular Automata and Markov Chain model is employed predict of forest land cover in this area. The results of the study indicate that forest land cover change is being transformed in Ba Be National Park. According to our estimate, from 1992 to 2019, the area covered by woody vegetation increased by 1.1%. By 2035, the area of broad-leaved forests will increase by 9.7%, due to a decrease in areas of meadows and shrubs. The increase in forest cover protected areas is explained by the measures taken by the Vietnamese government to expand the forest area in the country.

Mapping the spatial distribution of Lantana camara using high-resolution SPOT 6 data, in Mpumalanga communal areas, South Africa

Agroecosystems play a critical role in rural economics and contribute to the national economy. However, these areas are currently threatened by invasive species. Timely information on potential hotspots and affected areas is required for management purposes. We assess the spatial distribution patterns of Lantana camara invasion in savanna rangeland ecosystems before and after clearing (2014, 2016 and 2018), using high-resolution SPOT 6 data, in Agincourt, South Africa. A maximum likelihood classification algorithm was applied to SPOT 6 data to detect and map areas affected by this invasive species. Furthermore, different accuracy assessment measures (e.g. producer accuracy (PA), user accuracy (UA) and overall accuracy (OA)) were used to assess the validity of the classification results, namely built-up areas, bare areas, uninvaded areas and invaded areas. The results demonstrate that Lantana camara occurrence in rangelands ecosystems can be mapped with high accuracy. An overall accuracy of 81.29% was attained for 2014, 86.85% for 2016 and 88.69% for 2018. It was further observed that in 2016 Lantana camara occurrence was minimal when compared to 2014, but the results indicate regeneration in 2018 in previously cleared areas. Approximately 40.84% of the area was covered by Lantana camara in 2014, 25.19% in 2016 and 27.02% in 2018. Overall, the findings in this study underscore the need for continuous monitoring of affected areas, especially after clearing, to continuously advise on areas that are at risk for potential re-establishment of invasive weeds.

ASSESSMENT OF LAND USE/LAND COVER CHANGE MAPPING IN BANGUI CITY USING REMOTE SENSING AND GIS TECHNIQUES

Abstract. The security instability in the Central African Republic (CAR) forces the civilian population to flee the provinces to seek refuge in Bangui city, or in other countries. Human activity, which is very beneficial in the context of urbanization, is the main driving force of change in the city of Bangui, but also has a negative effect on the geoenvironment. Multispectral images data Landsat TM5, Landsat 7 ETM+ and Landsat-8 OLI of the years 1986, 2003 and 2020 was used to investigate Land use land cover (LULC) change of the city of Bangui. Maximum Likelihood (ML) classification algorithm was used to produce the map land use/land cover change detection in the study area. In Bangui city, four major classes have been identified, including vegetation, built-up, bare soil / rock and water. The analyses of the classified maps showed that Bangui city has been changed between 1986 and 2020, exceedingly area increased for built up (145.81%), vegetation (5.59%) and water (3.46%), it has however decreased for bare soil/ rock (40.60%). The overall accuracies and overall Kappa statistics achieved were 92.5%, 82.5% and 87.5%, and 0.90, 0.87 and 0.83 for 1986, 1999 and 2018 images, respectively.

Woodland Cover Change in the Central Rift Valley of Ethiopia

Woodlands, which are part of the landscape and an important source of livelihood for smallholders living in the environmentally vulnerable Central Rift Valley (CRV) of Ethiopia, are experiencing rapid changes. Detecting and monitoring these changes is essential for better management of the resources and the benefits they provide to people. The study used a combination of both quantitative and qualitative methods to analyze the extent and pattern of woodland cover changes from 1973 to 2013. Pixel-based supervised image classification with maximum likelihood classification algorithm was used for land cover classification and change detection analyses. Local peoples’ perceptions were used to explain the patterns of change and their possible reasons. Four major land cover classes were identified, with an overall accuracy of 88.3% and a Kappa statistic of 0.81 for the latest image. The analysis revealed a major land cover reversal, where woodland (92.4%) was the dominant land cover in 1973, while it was agriculture (44.7%) in 2013. A rapid reduction in woodland (54%) and forest (99%) covers took place between 1973 and 2013, with the majority of the conversions being made during the government transition period (1973 to 1986). Agriculture (3878%) and grassland (11,117%) increased tremendously during the 40-year period at the expense of woodlands and forests. Bare land increased moderately (40%). Thus, woodlands are under increasing pressure from other land uses, particularly agriculture, and declining faster. If the current trends of land cover change remain unabated it is likely that woodlands will disappear from the landscape of the area in the near future. Therefore, better forest policy and implementation tools, as well as better woodland management strategies and practices, need to be in place for woodlands to continue providing vital ecosystem goods and services to the local people, as well as to the environment.

Use of multispectral satellite datasets to improve ecological understanding of the distribution of Invasive Alien Plants in a water‐limited catchment, South Africa

AbstractInvasive Alien Plants (IAPs) pose major threats to biodiversity, ecosystem functioning and services. The availability of moderate resolution satellite data (e.g. Sentinel‐2 Multispectral Instrument and Landsat‐8 Operational Land Imager) offers an opportunity to map and monitor the occurrence and spatial distribution of IAPs. The use of two multispectral remote sensing data sets to map and monitor IAPs in the Heuningnes Catchment, South Africa, was therefore investigated using the maximum likelihood classification algorithm. It was possible to identify areas infested with IAPs using remote sensing data. Specifically, IAPs were mapped with a higher overall accuracy of 71%, using Sentinel‐2 MSI as compared to using Landsat 8 OLI, which produced 63% accuracy. However, both sensors showed similar patterns in the spatial distribution of IAPs within the hillslopes and riparian zones of the catchment. This work demonstrates the utility of the two multispectral data sets in mapping and monitoring the occurrence and distribution of IAPs, which contributes to improved ecological modelling and thus to improved management of invasions and biodiversity in the catchment.

Monitoring of Forest Land Cover Change in Binh Chau - Phuoc Buu Nature Reserve in Vietnam Using Remote Sensing Methods and GIS techniques

Binh Chau - Phuoc Buu Nature Reserve is located in the tropical rainforest zone of southeast Vietnam. For monitoring purposes of forest land cover change in Reserve satellite images (Landsat 5, Landsat 7 and Sentinel-2A) taken in 1988, 2003 and 2019 were used. NDVI was used for the estimation of vegetation quality. Forest land cover were classified by 5 categories using maximum likelihood classifier algorithm in ENVI 5.3. After classification, differentiation of forest land cover was analysed on every image. The results confirm that the area of meadows and shrubs increased from 16.3% in 1988 to 19.1% in 2019, and conversely, the area of broad-leaved forests decreased from 80.7% to 77.3%. Thus, there is a general degradation of forest vegetation.

Land Conversions and Forest Dynamics in a Riparian Forest Zone in South East Nigeria

The rate at which forest ecosystems are lost and modified across tropical landscapes are alarming, yet proper documentation and proactive mea sures to curtail this still remains a huge challenge in most areas. This research focused on elucidating the ongoing land use change patterns of a riparian forest landscape, its current impacts on the ecosystem and land surface temperature, as well as its likely future scenarios for the zone. LANDSAT images were downloaded for 1988, 2003 and 2018 and used to show the dynamics for the zone, its drivers and their varying temperatures. Maximum Likelihood Classification algorithm was used for the classification and the land-use classes were categorized as: Water body, Farms and Sparse Vegetation, Built-up Areas, Bare Surface, and Thick Vegetation. Furthermore, Markov Chain Analysis was employed for understanding the future patterns of land use change in the zone. Land use categories experienced changes over the three epochs, but among all, farmlands/ sparse vegetation and thick vegetation had the most significant changes from 7.70 to 58.67 percent and 73.56 to 20.58 percent, respectively; implying that much of the forestland use/cover (which constituted the bulk of the land initially; 73.56 percent) were converted to agricul tural land use. This same trend at which agriculture grew in the zone was seen to affect the land surface temperature for zone (Pearson correlation coefficient of 0.99 with p = 0.0058 at 0.05 level of significance). Future projection for the zone equally showed that agricultural land use will likely dominate the entire landscape in the coming years and a conse quent impact on the climate and ecosystem expected as well. On that note, intensive agricultural practices that seek to maximize allocated farm units were advocated. Such initiatives will help to ensure that agricultural growth is contained within delimited zones so that haphazard cultivations, reductions in ecological value of the forest landscape and consequent climatic impacts could be managed across the region.

Monitoring of Forest Land Use/Cover Change in Cat Tien National Park, Dong Nai Province, Vietnam Using Remote Sensing and GIS techniques

The Cat Tien National Park in the south of Vietnam represents a protected ecological area. It is located in tropical rainforest zone. For monitoring purposes of forest land use/cover change of the park satellite images (Landsat 5, Landsat 7 and Sentinel-2A) taken in 1988, 2003 and 2019 were used. NDVI was used for the estimation of vegetation quality. Forest land use/cover were classified by 5 categories using maximum likelihood classifier algorithm in ENVI 5.3. After classification, differentiation of forest land use/cover was analyzed on every image. Deforestation took place in 1988-2003 periods and broad-leaved forest areas have undergone reduction. Currently, new policies were introduced to reduce deforestation and therefore support free growth of tropical forest areas.

Land use changes and restraining environmental risks via mitigation approach, the case of Darakeh-Velenjak watershed in Iran

Widespread land use changes and unusual uses of highlands in Northern Tehran (Darakeh-Velenjak Watershed) have markedly caused to pose environmental risks and unsustainability. This research aims to study land use changes in Darakeh-Velenjak Watershed, to identify the key risks which can be posed by land use changes, and to present appropriate strategies using mitigation approach. To study land uses changes, satellite images were used (TM 1987, ETM+ 2000, and IRS 2018) and assessed by concentrating on four classes of land including vegetation, arid, roads and built-up areas. To extract the necessary information from the satellite images, Supervised Maximum Likelihood Classification algorithm was used, further complemented by visual interpretation methods. The sample population for conducting the Delphi technique consisted of 34 experts. The results of this research exhibited that from 1987 to 2018, vegetation had dramatically decreased, while arid, roads and built-up areas had increased. The results also revealed that in 1987, around 37.3% of the study area was covered with vegetation while in 2018, less than 14.54% of the area was covered with vegetation. In contrast, built-up, roads and arid areas which respectively covered 8.96%, 1.75%, and 51.99%, increased to 15.18%, 3.19% and 67.08% in 2018. In addition, the results of this study illustrated that key risks of land use changes in the Darakeh-Velenjak Watershed could be divided into natural and human risks. The human risks with an average score of 4.21 were more than natural risks. Finally, strategies have been presented using mitigation approach in the study area. Results revealed that avoidance strategies with an average score of 3.37 were situated in the first rank because unfortunately this watershed has been damaged by excessive land use changes in an aspect of developing buildings, roads, tourism infrastructures and etc.

ПРОГНОЗ СОСТОЯНИЯ РАСТИТЕЛЬНОГО ПОКРОВА ЛЕСНЫХ УГОДИЙ ЗАПОВЕДНИКА ДОНГ НАЙ ВЬЕТНАМА

The study was conducted in Dong Nai Reserve, specially protected natural area (SPNA), Vietnam. It aims to analyze and forecast forest land cover in the Reserve. For these purposes were studied satellite images (Landsat 5, Landsat 7 and Landsat 8) taken in 2003, 2011 and 2019. The Normalized Difference Vegetation Index (NDVI) was used to identify vegetation quality. Forest land cover was divided into 5 categories using maximum likelihood classifier algorithm. In order to detect and evaluate forest land cover changes, supervised classification and image differencing method are applied. Then, Cellular Automata and Markov Chain model was employed for making forecast of forest land cover in this area. The results of the study indicate that forest land cover change is being transformed in Dong Nai Reserve. According to our estimation, from 2003 to 2019, the area covered by woody vegetation increased by 7.0 %. By 2035, the area of broad-leaved forests will increase by 1.6 %, due to a decrease in areas of meadows and shrubs. The dynamics of increasing forest land is explained by the measures taken by the Vietnamese government to expand the area of forests in SPNA.

Environmental Assessment of Land Use Land Cover Using Remote Sensing and GIS: A Case Study of Khirthar National Park

National Park area is a natural reserve with an outstanding beauty having rich biodiversity, natural eco-system and sustainable environment is a gift from nature to conserve. It is extremely important to monitor the sustainability of National Park by extracting land use land cover change information over the period of time using satellite imagery which is the main application of change detection. The aim of this research study is to assess the land use land cover change at Khirthar National Park by acquiring multispectral Land sat satellite imagery to detect decadal information from 1992 to 2019 using supervised classification Maximum Likelihood classifier. The Land sat satellite data used in this research study was almost of the summer season to assess land use land cover change using supervised classification Maximum Likelihood Classifier and results showed increasing trend in agriculture while decreasing trend in natural vegetation at Khirthar National Park which is a clear indication of increasing anthropogenic activities. The supervised classification Maximum Likelihood Algorithm used to detect land use land cover change at Khirthar National Park results overall accuracy in the year 1992, 2000, 2010 and 2019 was 99.7446%, 99.4536%, 99.6495% and 96.7530% while the kappa statistics results achieved in the year 1992, 2000, 2010 and 2019 was 0.9951, 0.9879, 0.9648 and 0.9503. These increasing human activities will not only disturbs the ecology of National Park but also will impact on the local flora and fauna of the park, therefore it is highly recommended to prevent further loss of biodiversity, natural eco-system and the sustainable management for such natural heritage.

Monitoring changes in land use and distribution of mangroves in the southeastern part of the Mekong River Delta, Vietnam

Mangroves are one of the most valuable and productive coastal ecosystems. Previous studies show a severe loss of mangroves around the world over the last several decades due to natural and anthropogenic activities. Mangroves located in the southeastern part of the Mekong River Delta (MRD) are also affected by these activities. Shrimp farming is considered as one of the main drivers causing the rapid loss and degradation of mangroves. The goal of this study is to assess the spatiotemporal changes in land use and distribution of mangroves in the Soc Trang and Bac Lieu provinces of the southeastern part of the MRD. Multi-temporal Landsat data were used for land use and land cover (LULC) classification using the maximum-likelihood classification algorithm. The changes in mangrove forest areas were monitored using medium spatial resolution (Landsat-5 Thematic Mapper and Landsat-8 Operational Land Imager) satellite imageries from 1988 to 2018. In the study area, there were seven major LULC types namely, dense mangroves, sparse mangroves, aquaculture farms, arable land with crop cover, arable land without crop cover, settlements, and water bodies. The overall accuracies of the LULC maps in 1988, 1998, 2008, and 2018 were 81.2%, 83.3%, 78.3%, and 81.9%, respectively. This study reveals that dense and sparse mangrove forests have decreased by 90% from 5495 hectares (ha) to 515 ha and by 55% from 14,105 to 6289 ha, respectively from 1988 to 2018. On the other hand, the aquaculture farm has increased at the rate of 5024 ha/year for a period of 30 years. This rapid growth of aquaculture farming activities caused the rapid loss and degradation of mangroves in the MRD. Quantitative information about mangrove change obtained by this study is considered to be useful for future coastal management and relevant policies in the MRD.

Land use/land cover change and surface urban heat island intensity: source–sink landscape-based study in Delhi, India

Urbanization-induced rapid land use/land cover change (LULCC) modifies the thermal characteristics of a region at local scale and often creates urban heat island (UHI) effects. Delhi also experiences such enhancing UHI scenario. In present study, LULCC was linked with the land surface temperature (LST), and surface urban heat island intensity (SUHII) was measured at the sub-district level. Landsat imageries and MODIS LST product for 2001, 2009 and 2017 were main data sources. LULC classification and LST were derived using maximum likelihood classification and single-channel algorithm techniques, respectively. Proposed LULC classification scheme was built-up, current fallows land, bare land (or source landscape); and crop land, vegetation and water bodies (or sink landscape). SUHII was calculated using the contribution index of both source and sink landscapes and landscape index (LI). Results from LI also supported with temperature vegetation feature space index, transformed difference vegetation index, enhanced vegetation index, diurnal temperature range, population density and transect analysis. The result showed that sub-districts of North, North-East, East and West Delhi were more prone to SUHII due to the highly dense built-up area and industrial area. SUHII was comparatively low in the sub-districts of South and South-West Delhi because of less built-up area as well the presence of greenery. Not only built-up but also fallow land and barren land contributed significantly in some places, i.e. South and South-West Delhi. Besides, urban green space and green crop field reduced the LST sufficiently in some areas, i.e. Najafgarh and New Delhi. Again, this study could help to understand UHI in Delhi at sub-district level.

Land cover mapping using Sentinel-1 SAR and Landsat 8 imageries of Lagos State for 2017.

For several years, Landsat imageries have been used for land cover mapping analysis. However, cloud cover constitutes a major obstacle to land cover classification in coastal tropical regions including Lagos State. In this work, a land cover appearance for Lagos State is examined using Sentinel-1 synthetic aperture radar (SAR) and Land Satellite 8 (Landsat 8) imageries. To this aim, a Sentinel-1 SAR dual-pol (VV+VH) Interferometric Wide swath mode (IW) data orbit for 2017 and a Landsat 8 Operational Land Imager (OLI) for 2017 over Lagos State were acquired and analysed. The Sentinel-1 imagery was calibrated and terrain corrected using a SRTM 3Sec DEM. Maximum likelihood classification algorithm was performed. A supervised pixel-based imagery classification to classify the dataset using training points selected from RGB combination of VV and VH polarizations was applied. Accuracy assessment was performed using test data collected from high-resolution imagery of Google Earth to determine the overall classification accuracy and Kappa coefficient. The Landsat 8 was orthorectified and maximum likelihood classification algorithm also performed. The results for Sentinel-1 include an RGB composite of the imagery, classified imagery, with overall accuracy calculated as 0.757, while the kappa value was evaluated to be about 0.719. Also, theLandsat 8 includes a RBG composite of the imagery, classified imagery, but an overall accuracy of 0.908 and a kappa value of 0.876. It is concluded that Sentinel 1 SAR result has been effectively exploited for producing acceptable accurate land cover map of Lagos State with relevant advantages for areas with cloud cover. In addition, the Landsat 8 result reported a high accuracy assessment values with finer visual land cover map appearance.

Land Use and Land Cover Dynamics in Eastern Pastoral Rangelands of Somali Region, Ethiopia

Drastic changes have occurred in Harshin district rangeland management over nearly the last two decades, due to rapid spread of fencing of 'private' grazing areas, contrary to the communal ownership pattern that informal institutions govern. In Harshin district, the land use change over time and space and temporal trends rangeland condition have never been studied relative to the effects of long term management changes. This study analyzed land use and land cover (LULC) change dynamics since 1980s. Three dates, 1984, 1998 and 2014, Landsat images were used for classification and analysis of the various LULC. The three images were geo-referenced, re-sampled and processed for classification, using the maximum likelihood classifier algorithm. The best Kappa hat statistic of classification accuracy was 85%. The results of the classification over the three periods showed that settlement and bare land increased from 7325 hectare (ha) in 1984 to 18,720 ha in 1998. Grassland decreased by 8010 ha over the same period and increased by 53,230 ha by the 2014; shrub land also decreased by 9471 ha and 31,196 ha in 1998 and 2014, respectively. Woodland increased by 6176 ha in 1998, however; it decreases by 33,175 ha in 2014. The study findings have shown important changes in the LULC patterns in the district. The bare land, coupled with shrub land reduction between 1984 and 1998, substantial increments in bare land, settlement land uses, and grassland coverage, while a substantial decrease in woodland coverage were found between 1998 and 2014. These trends are certainly the characteristics of pastoral way of life turn to settlement. This suggests that major changes in the socio-ecological driving forces affecting landscape dynamics have occurred in the last two decades or so. Keywords: Harshin district, Land use land cover dynamics, Rangeland, Remote sensing DOI : 10.7176/JEES/9-6-03 Publication date :June 30 th 2019

A Comparative Evaluation of Different Techniques of Supervised Classification in Landuse/Landcover Mapping of Awka South L.G.A, Anambra State, Nigeria.

The aim of this study is to compare the different techniques of supervised classification using Awka South LGA, of Anambra State as a case study. The techniques considered include: Maximum Likelihood (MLC), Minimum Distance, Mahalanobis Distance, Spectral Angle Mapper and Parallelepiped. Landsat 7 ETM+ (2000 and 2007) and Landsat 8 OLI/TIRS (2015) were acquired. The images were pre-processed. The scan-line effect present in the Landsat 7 image was corrected using the analysis tool of Quantum GIS (QGIS) 2.18 software. To compensate for atmospheric effects, Fast Line-of-site Atmospheric Analysis of Hypercube (FLAASH) Atmospheric Module of ENVI software was used. Image enhancement was carried out on the images. The images were classified using the different techniques and the results compared. Change detection was also carried out to determine the rate of changes between 2000 and 2015. Error matrices of the various techniques were calculated to determine the accuracy level of the algorithms and to judge which is the better choice. It can be deduced from the results that Maximum Likelihood (99.63%) produced the best result, followed closely by Mahalanobis Distance (98.54%), Spectral Angle (89.28%), Minimum Distance (84.42%) and Parallelepiped (85.00%). The study recommends Maximum Likelihood Classification algorithm for supervised classification. Key words: Classification, Maximum Likelihood, Algorithm, Land cover land use DOI : 10.7176/JEES/9-5-10 Publication date :May 31 st 2019

The impact of land used changes on mangrove forest and shoreline dynamic in Muara Gembong, Bekasi, West Java

Muara Gembong used to a large area of mangroves at the northern tip of Bekasi Regency, West Java. Over the last 4 decades, this area has experienced significant changes in land use and loss of environmental and ecosystem services provided by mangroves. Past economic activities have resulted in the conversion of thousands of hectares of mangrove into pond areas in pursuit of fisheries production, which at the time became a national program. This change has resulted in ecological disasters and caused economic losses to coastal communities such as shoreline changes (causing the loss of several coastal villages) and sea water intrusion (damaging pond areas and resulting crop failure). This study assessed the impact of land used changes on mangrove dynamic and shoreline erosion rate along the 5 coastal villages of Muara Gembong over 42 years. Multi-temporal Landsat data set and ground truth survey were applied. The imageries have been classified digitally using Maximum Likelihood Classification (MLC) algorithm and have been validated through the process of accuracy assessment. Digital Shoreline Analysis System (DSAS) in ArcGIS environment was used to create transects and statistical analyses for the shoreline. The results show that since 1976 the area of mangrove forests has decreased by 55% and most of it has changed to become fishponds (increased to 35%). The land conversion has resulted in shoreline changes and the biggest erosion occurred in the Pantai Bahagia village with an erosion rate of 15.46 m/year and lost an area of 330,460 Ha.

Soil salinity mapping in Everglades National Park using remote sensing techniques and vegetation salt tolerance

Everglades National Park (ENP) is a hydro-ecologically enriched wetland with varying salinity contents, which is a concern for terrestrial ecosystem balance and its nearby urban sustainability. In this study, spatio-temporal soil salinity maps are created using remote sensing techniques, coupled with literature review to understand vegetation salt tolerance properties, and field assessments entailing insitu electric conductivity (EC) measurements. The mapping first entailed the execution of a supervised machine learning technique—the maximum likelihood classification algorithm—to delineate seven vegetation-based land cover classes for the area, namely, mangrove forest, mangrove scrub, low-density forest, sawgrass, prairies and marshes, barren lands with woodland hammock and water, for the years 1996, 2000, 2006, 2010 and 2015. The classifications for 1996–2010 yielded accuracies of 82%–94%, and the 2015 classification was supported through ground truthing. Afterwards, EC tolerance thresholds for each vegetation class were established, which yielded soil salinity maps comprising four soil salinity classes—i.e., the non- (EC = 0–2 dS/m), low- (EC = 2–4 dS/m) and high-saline (EC ≥4 dS/m) areas. The soil salinity maps visualized the spatial distribution of soil salinity with no significant temporal changes. Furthermore, insitu EC measurements carried out at 23 sampling sites covering all land cover classes mostly validated (91% samples were tested within range) the estimated soil salinity ranges for the latest distribution. The approach of using land cover classes to sense salinity ranges in an urban-proximal ecosystem is pragmatic and application oriented, attributing to novel and useful study upshots considering the diversifying ecological context.