Discrimination and classification of mangrove forests using EO-1 Hyperion data: a case study of Indian Sundarbans

Kumar, Tanumi; Mandal, A.B.; Dutta, Dibyendu; Nagaraja, R.; Dadhwal, V. K.

doi:10.1080/10106049.2017.1408699

Cited by 42 publications

(15 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…From the existing research, as well as the study conducted by Sarker et al 2016 [145], it is known that degradation is happening in the Sundarbans. Therefore, a more important index consideration, such as landscape change index (LCI) proposed by Krajewski et al [146], our provided index (used in this study) of SFHQ with ranks and CGP and SFD based on FDI ranking (user friendly and can be modified), along with drone-based forest surveys with other satellite sensors [147] with high-resolution satellite images, such as Sentinel 2 MSI, EO-1 Hyperion [148] and UAV [149], might be effective in near-future research.…”

Section: Limitations and Future Scope Of The Studymentioning

confidence: 99%

Applying Multi-Temporal Landsat Satellite Data and Markov-Cellular Automata to Predict Forest Cover Change and Forest Degradation of Sundarban Reserve Forest, Bangladesh

Hasan

Nath

Sarker

et al. 2020

Forests

View full text Add to dashboard Cite

Overdependence on and exploitation of forest resources have significantly transformed the natural reserve forest of Sundarban, which shares the largest mangrove territory in the world, into a great degradation status. By observing these, a most pressing concern is how much degradation occurred in the past, and what will be the scenarios in the future if they continue? To confirm the degradation status in the past decades and reveal the future trend, we took Sundarban Reserve Forest (SRF) as an example, and used satellite Earth observation historical Landsat imagery between 1989 and 2019 as existing data and primary data. Moreover, a geographic information system model was considered to estimate land cover (LC) change and spatial health quality of the SRF from 1989 to 2029 based on the large and small tree categories. The maximum likelihood classifier (MLC) technique was employed to classify the historical images with five different LC types, which were further considered for future projection (2029) including trends based on 2019 simulation results from 1989 and 2019 LC maps using the Markov-cellular automata model. The overall accuracy achieved was 82.30%~90.49% with a kappa value of 0.75~0.87. The historical result showed forest degradation in the past (1989–2019) of 4773.02 ha yr−1, considered as great forest degradation (GFD) and showed a declining status when moving with the projection (2019–2029) of 1508.53 ha yr−1 and overall there was a decline of 3956.90 ha yr−1 in the 1989–2029 time period. Moreover, the study also observed that dense forest was gradually degraded (good to bad) but, conversely, light forest was enhanced, which will continue in the future even to 2029 if no effective management is carried out. Therefore, by observing the GFD, through spatial forest health quality and forest degradation mapping and assessment, the study suggests a few policies that require the immediate attention of forest policy-makers to implement them immediately and ensure sustainable development in the SRF.

show abstract

Section: Limitations and Future Scope Of The Studymentioning

confidence: 99%

Applying Multi-Temporal Landsat Satellite Data and Markov-Cellular Automata to Predict Forest Cover Change and Forest Degradation of Sundarban Reserve Forest, Bangladesh

Hasan

Nath

Sarker

et al. 2020

Forests

View full text Add to dashboard Cite

show abstract

“…ETLAND plants act as sentinels for ecological changes and provide early signs of physical or chemical degradation, such as reduction of wetland area and wetland plant diversity, decreased and damaged rare and endangered waterfowl habitats, water eutrophication, as well as decreased of organic matter accumulation and primary productivity [1,2]. And once their health and functions are disturbed, the effects are detrimental on all life forms supported by wetlands [3].…”

Section: Introductionmentioning

confidence: 99%

“…Although the spatial resolution of some satellite images has improved significantly (e.g., Quickbird and Worldview images), limitations in spectral resolution mean that the classification of wetland plant communities may be unsatisfactory [10,11]. Hyperspectral images can capture the subtle spectral differences among ground objects because of the continuous spectral information, thereby allowing different wetland plant communities to be identified accurately [1,12]. In recent years, hyperspectral data have been commonly used to map wetland plant communities by means of diverse algorithms [13,14].…”

Section: Introductionmentioning

confidence: 99%

Mapping Wetland Plant Communities Using Unmanned Aerial Vehicle Hyperspectral Imagery by Comparing Object/Pixel-Based Classifications Combining Multiple Machine-Learning Algorithms

Mao

Wang

et al. 2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Understanding the spatial patterns of plant communities is important for sustainable wetland ecosystem management and biodiversity conservation. With the rapid development of unmanned aerial vehicle (UAV) technology, UAV-borne hyperspectral data with high spatial resolution have become ideal for accurate classification of wetland plant communities. In this study, four dominant plant communities (Phragmites australis, Typha orientalis, Suaeda glauca, and Scirpus triqueter) and two unvegetated cover types (water and bare land) in the Momoge Ramsar wetland site were classified. This was achieved using UAV hyperspectral images and three object-and pixel-based machine-learning classification algorithms [random forest (RF), convolutional neural network (CNN), and support vector machine (SVM)]. First, spectral derivative analysis, logarithmic analysis, and continuum removal analysis identified the wavelength at which the greatest difference in reflectance occurs. Second, dimensionality reduction of hyperspectral images was conducted using principal component analysis. Subsequently, an optimal feature combination for community mapping was formed based on data transformation (spectral features, vegetation indices, and principal components). Image objects were

show abstract

“…There are studies which did focus on index-based mapping where mangrove-specific indices were applied to different inputs bands and satellite data. (Zhang and Tian, 2013;Winarso et al, 2014;Gupta et al, 2018;Jia et al, 2019;Kumar et al, 2019;Baloloy et al, 2020). The latest mangrove extent estimate in the Philippines were generated using Sentinel-2 images and the derived Mangrove Vegetation Index developed by Baloloy et al (2020).…”

Section: Introductionmentioning

confidence: 99%

Development of a Rapid Mangrove Zonation Mapping Workflow Using Sentinel 2-Derived Indices and Biophysical Dataset

Baloloy

Blanco

Sharma

et al. 2021

Front. Remote Sens.

View full text Add to dashboard Cite

Moderate to high resolution satellite imageries are commonly used in mapping mangrove cover from local to global scales. In addition to extent information, studies such as mangrove composition, ecology, and distribution analysis require further information on mangrove zonation. Mangrove zonation refers to unique sections within a mangrove forest being dominated by a similar family, genus, or species. This can be observed both in natural and planted mangrove forests. In this study, a mapping workflow was developed to detect zonation in test mangrove forest sites in Katunggan-It Ibajay (KII) Ecopark (Aklan), Bintuan (Coron), Bogtong, and Sagrada (Busuanga) in the Philippines and Fukido Mangrove Park (Ishigaki, Japan) using Sentinel-2 imagery. The methodology was then applied to generate a nationwide mangrove zonation map of the Philippines for year 2020. Combination of biophysical products, water, and vegetation indices were used as classification inputs including leaf area index (LAI), fractional vegetation cover (FVC), fraction of photosynthetically-active radiation (FAPAR), Canopy chlorophyll content (Cab), canopy water content (Cw), Normalized Difference Vegetation Index (NDVI), modified normalized difference water index (MNDWI), modified chlorophyll absorption in reflectance index (MCARI), and red-edge inflection point (REIP). Mangrove extents were first mapped using either the Maximum Likelihood Classification (MLC) algorithm or the Mangrove Vegetation Index (MVI)-based methodology. The biophysical and vegetation indices within these areas were stacked and transformed through Principal Component Analysis (PCA). Regions of Interest (ROIs) were selected on the PCA bands as training input to the MLC. Results show that mangrove zonation maps can highlight the major mangrove zones in the study sites, commonly limited up to genera level only except for genera with only one known species thriving in the area. Four zones were detected in KII Ecopark: Avicennia zone, Nypa zone, Avicennia mixed with Nypa zone, and mixed mangroves zones. For Coron and Busuanga, the mapped mangrove zones are mixed mangroves, Rhizophora zone and sparse/damaged zones. Three zones were detected in Fukido site: Rhizophora stylosa-dominant zone, Bruguiera gymnorrhiza-dominant zone, and mixed mangrove zones. The zonation maps were validated using field plot data and orthophotos generated from Unmanned Aerial System (UAS) surveys, with accuracies ranging from 75 to 100%.

show abstract

Discrimination and classification of mangrove forests using EO-1 Hyperion data: a case study of Indian Sundarbans

Cited by 42 publications

References 44 publications

Applying Multi-Temporal Landsat Satellite Data and Markov-Cellular Automata to Predict Forest Cover Change and Forest Degradation of Sundarban Reserve Forest, Bangladesh

Applying Multi-Temporal Landsat Satellite Data and Markov-Cellular Automata to Predict Forest Cover Change and Forest Degradation of Sundarban Reserve Forest, Bangladesh

Mapping Wetland Plant Communities Using Unmanned Aerial Vehicle Hyperspectral Imagery by Comparing Object/Pixel-Based Classifications Combining Multiple Machine-Learning Algorithms

Development of a Rapid Mangrove Zonation Mapping Workflow Using Sentinel 2-Derived Indices and Biophysical Dataset

Contact Info

Product

Resources

About