Introduction
Mangrove forests thrive in the inter-tidal zones along coastlines in tropical and semi-tropical regions, playing a crucial role in maintaining ecological balance (Tuan, Oanh et al., 2002) These highly productive ecosystems serve as vital habitats for diverse wildlife species, supporting biodiversity and ecological sustainability (Wolanski, Brinson et al.).
Mangroves are vital coastal ecosystems that provide natural protection against natural disasters, such as tsunamis and storms, as evidenced by their role during the 2004 Indian Ocean tsunami They stabilize coastlines, prevent erosion, enhance coastal accretion, and act as natural barriers against tidal bores, cyclones, and other destructive forces Additionally, mangroves support the socioeconomic well-being of coastal communities by supplying timber, firewood, building materials, charcoal, tannin, food, honey, herbal medicines, and other vital resources Importantly, they are among the most carbon-rich ecosystems in the tropics, playing a crucial role in climate change mitigation through "blue carbon" storage, making their conservation essential for environmental and climate stability.
Mangrove ecosystems hold immense socio-economic importance but are currently facing severe threats due to high population growth and increased migration to coastal regions, which escalate demand for their resources This decline is worsened by inadequate governance, poor planning, and uncoordinated economic development in coastal zones Globally, over 3.6 million hectares of mangroves have been lost since 1980, with Asia experiencing the most significant reduction of 1.9 million hectares (FAO 2007), highlighting the urgent need for conservation efforts.
Vietnam's mangrove areas have experienced significant decline, mirroring trends across Southeast Asia Once covering approximately 400,000 hectares in the early 20th century, mangrove forests in Vietnam have drastically reduced over the past 50 years (Tuan, Munekage et al., 2003) Between 1964 and 1997, northern regions from Mong Cai to Do Son saw a loss of 17,094 hectares of mangroves, while the Red River plain lost about 4,640 hectares during the same period This alarming reduction underscores the urgent need for conservation efforts to protect Vietnam’s vital mangrove ecosystems.
Between 1975 and 1991, mangrove forests in Vietnam experienced significant decline, with a further decrease of 7,430 hectares in 1993 (NEA 2003) Despite government and international efforts to restore mangroves during the 1990s, aquaculture activities have continued to contribute to the substantial reduction of these vital ecosystems along the Northern coast However, mangrove forests within protected zones are effectively managed through community-based forest management approaches, ensuring better conservation and sustainable use (Dat and Yoshino et al., 2013).
Hai Phong city, with 125 km of sea dykes, has significant potential for supporting local mudflats and mangrove ecosystems Between 1989 and 2013, the region experienced a net loss of 281 hectares of mangrove forests in the first decade, followed by a gain of approximately 355 hectares over the subsequent 24 years The annual rate of mangrove loss was around 23 hectares from 1989 to 2001, but from 2001 to 2013, mangrove areas increased by about 53 hectares per year This shift was influenced by Vietnam's Doi Moi economic reforms initiated in 1986, which transformed the economy into a market-oriented system and promoted export-oriented shrimp farming As a result, many mangrove areas were converted into shrimp aquaculture due to high export profits However, from 2001 onward, some coastal districts managed to conserve and restore mangroves through community-based forest management in collaboration with local authorities Notably, after three severe tropical cyclones, including Cyclone Washi in 2005, which caused damage in some areas, local communities recognized the vital role of mangroves in protecting dyke systems and livelihoods Consequently, they replanted mangroves in vulnerable zones to enhance natural defenses against future typhoons.
Mapping mangroves is essential for supporting coastal management and planning, enabling accurate inventory of mangrove ecosystems and aquaculture sites, as well as effective change detection Mangrove maps can be developed through on-site investigations or by analyzing remotely sensed images using GIS techniques, providing reliable data for sustainable coastal development.
Numerous studies have demonstrated the effectiveness of Landsat imagery for mangrove habitat mapping using various image processing techniques, including vegetation indices The most commonly employed index is the Normalized Difference Vegetation Index (NDVI), which utilizes near-infrared and visible red spectral bands to assess vegetation health (Ramsey & Jensen, 1996; Green et al., 1998; Wang et al., 2004; Campbel, 1996; Baret & Guyot, 1991; Ismail et al., 2010; Akamar et al., 2009) Several developed indices aim to minimize sensitivity to factors such as soil background, canopy architecture, and atmospheric conditions However, despite their widespread use, the accuracy of mangrove mapping relying solely on vegetation index models remains limited, highlighting the need for complementary methods to improve precision.
This study evaluates three key vegetation indices—SVI (Simple Vegetation Index), NDVI (Normalized Difference Vegetation Index), and SAVI (Soil-adjusted Vegetation Index)—to effectively map mangrove forests in Kien Thuy These indices are compared to determine their accuracy and suitability for detecting and monitoring mangrove vegetation The research aims to identify the most reliable vegetation index for precise mangrove forest mapping, supporting sustainable conservation efforts in the region.
Do Son coasts, Hai Phong, Vietnam using multi-temporal image.
Literature Review
Overview of using vegetation index for mangrove mapping
Since the launch of the first remote sensing satellite in the 1970s, significant progress has been made in mapping mangrove areas using advanced image processing techniques The development of vegetation indices models has been pivotal in vegetation mapping, as these indices combine spectral values to generate a single value that indicates vegetation vigor within a pixel (Ismail et al., 2010; Campbell, 1996) Vegetation indices (VIs) are more sensitive than individual spectral bands to parameters such as vegetation health and density (Baret and Guyot, 1991), although their accuracy can be influenced by two main factors.
Effects of soil background on VI:
Until the soil is fully covered by vegetation, the soil background will influence the
VI For incomplete canopies, the wetting of a previously dry soil can cause a change in VI The change is further complicated by the fact that transmission of light through vegetation is considerably greater in the NIR than in the Red band Both the ratio and the linear combination classes of VI rely on the existence of the soil baseline in Red and NIR wavelength space for soil normalization (Huete, 1998) The intercept of this line is close to, but does not pass through the origin and there is usually some scatter of soil points away from the principal soil line Such secondary soil influences are most noticeable with Red and yellow colored soils (Kauth and Thomas, 1976).These two factors affect the discrimination of low amounts of vegetation from bare soil, and are significant in arid regions and in the early stages of vegetation growth (Huete et al., 1984)
Effect of canopy architecture on VI:
The architecture of a vegetation canopy influences the direction of reflected radiation from plant surfaces Erectophile canopies, characterized by vertical elements, trap reflected radiation within the canopy, reducing the amount reflected vertically toward nadir-oriented sensors In contrast, planophile canopies, with horizontally oriented leaves, reflect more radiation in the vertical direction, resulting in less internal trapping Consequently, nadir-pointing sensors can detect 20-30% more reflected radiation from planophile canopies compared to erectophile ones, highlighting the impact of canopy structure on remote sensing measurements (Bunnik, 1978).
Research by Jackson and Pinter (1986) demonstrated that canopy architecture significantly influences vegetation indices (VI), with peak green vegetation densities resulting in the RVI being approximately 30% higher for planophile canopies Similarly, the Plant Vegetation Index (PVI) was about 30% higher in planophile canopies compared to erectophile canopies, indicating that different canopy geometries affect VI responses even when vegetation amounts are similar Although this complexity may initially seem to reduce the utility of VI, it actually provides a valuable perspective, offering potential insights into canopy architecture that were previously difficult to obtain Incorporating canopy structure into VI analysis enhances its effectiveness for remote sensing and vegetation monitoring.
Key vegetation indexes for coastal mangrove mapping
Understanding how vegetation indices are designed requires knowledge of soil influence, the soil line, and the vegetation iso-line Different types of vegetation indices have been developed over the years, including ratio-based indices that assume all vegetation iso-lines converge at a single point and measure the slope between this convergence point and the soil line Additionally, some indices treat the vegetation iso-line as parallel to the soil line, known as perpendicular vegetation indices, which enhance accuracy in vegetation analysis.
Ratio-based Vegetation Indices (VIs) are calculated using data from visible red and near-infrared bands, providing valuable insights into the health and abundance of green vegetation cover and biomass These indices serve as an essential first step in land cover classification, helping to differentiate vegetated areas from non-vegetated regions effectively By capturing the contrast between red and near-infrared reflectance, ratio-based VIs facilitate accurate delineation of vegetative zones, supporting applications in environmental monitoring and land management.
The Simple Vegetation Index (SVI) is a ratio vegetation index developed by Rose et al (1973) that effectively distinguishes green vegetation from soil background It is calculated by dividing the reflectance values in the near-infrared (NIR) band by those in the red band (R), providing a reliable measure of vegetation health and density This index is widely used in remote sensing applications for monitoring plant growth and assessing land cover changes.
The contrast between red and infrared bands for vegetated pixels is evident, with high index values resulting from low red reflectance due to chlorophyll absorption and high infrared reflectance caused by leaf structure A ratio value below 1.0 indicates non-vegetation, while a ratio above 1.0 signifies vegetated areas However, this method faces a major drawback when the red band pixel value is zero, leading to an infinite ratio To address this issue, the Normalized Difference Vegetation Index (NDVI) is calculated, providing a more reliable measure of vegetation health.
The Normalized Difference Vegetation Index (NDVI) effectively addresses the limitations of the Ratio method, such as division by zero issues, by providing a more reliable spectral vegetation index Introduced to differentiate green vegetation from background soil brightness, NDVI enhances the accuracy of vegetation monitoring and assessment in remote sensing applications (Rose et al.).
NDVI = (NIR - RED) / (NIR + RED)
The most commonly used Vegetation Index (VI) effectively minimizes topographic effects while providing a linear measurement scale from –1 to +1 Negative VI values indicate non-vegetated areas, whereas positive values represent vegetated regions, making it a reliable tool for remote sensing vegetation analysis.
Ratio Vegetation Index (RVI): The ratio vegetation index is the reverse of the standard simple ratio (Richerdson et al., 1977),
The range for RVI extends from 0 to infinity The ratio value less than 1.0 is taken as vegetation while value greater than 1.0 is considered as non-vegetation area
The Perpendicular Vegetation Index (VIS) aims to mitigate soil brightness effects in images with sparse vegetation, where pixels often contain a mix of green vegetation and soil background This index is based on the soil line intercept concept, which models the typical spectral signatures of bare soils in red and near-infrared bands The soil line is a hypothetical line derived through linear regression of the infrared band against the red band for soil pixels, helping differentiate between soil and vegetation Pixels located near the soil line are classified as soils, while those farther away are identified as vegetation, enabling more accurate vegetation analysis in mixed pixels.
Y2 = 0.985684x + 9.501355(infra-red band as independent variable)
The Perpendicular Vegetation Index (PVI) is a remote sensing metric that differentiates vegetation from non-vegetation in arid and semi-arid regions by calculating the perpendicular distance from each pixel to the soil line (Richerdson et al., 1977) Pixels located close to the soil line are classified as non-vegetation, while those farther away indicate the presence of vegetation To ensure accuracy across different dates, PVI values require atmospheric correction due to the index’s sensitivity to atmospheric variations.
PVI= sin (a) NIR – cos (a) Red
Where, a = angle between the soil line and the NIR axis
Where, (x1, y1) is the co-ordinate of the pixel and (x2, y2) is the coordinate of soil line point that is perpendicular to pixel co-ordinate
Perpendicular distance less than 7.0 is taken as non-vegetation area while greater than 7.0 is taken as vegetation area
The Perpendicular Vegetation Index 1 (PVI1) was developed to address the limitations of the original PVI equation, which is computationally intensive and fails to distinguish between pixels on either side of the soil line Since vegetation typically exhibits higher infrared reflectance compared to red reflectance, all vegetation pixels tend to fall to the right of the soil line in spectral space However, non-vegetation pixels, such as water bodies, may be located at similar distances from the soil line but fall on the left side, highlighting the need for a more efficient and discriminative index in remote sensing vegetation analysis.
In PVI analysis, water pixels are assigned high vegetation index values to distinguish them from vegetated areas PVI1 specifically assigns negative values to pixels that can be identified as non-vegetation, such as water bodies The mathematical equation for PVI1, as described by Perry et al (1984), facilitates accurate delineation between vegetated and non-vegetated pixels, enhancing remote sensing vegetation assessments.
Where, NIR: reflectance in the near infrared band, RED: reflectance in the red band , a : intercept of the soil line, b : slope of the soil line
In the regression analysis, the infrared band is used as the independent variable, while the red band serves as the dependent variable A perpendicular distance threshold of less than 6.5 is established to identify non-vegetation areas, whereas distances greater than 6.5 indicate vegetation regions This methodology effectively differentiates vegetation from non-vegetation based on spectral reflectance, supporting accurate land cover classification using remote sensing data.
The Perpendicular Vegetation Index 2 (PVI2) utilizes the red band as the independent variable and the infrared band as the dependent variable in regression analysis, emphasizing the significance of the red band in vegetation monitoring Based on the soil line intercept, PVI2 effectively isolates vegetation signals from soil background, making it a valuable remote sensing index Mathematically, PVI2 is represented through a specific formula that quantifies vegetation health by perpendicular distance from the soil line in the spectral space, enhancing accuracy in vegetation assessment.
√ Where, a: intercept of the soil line, b: slope of the soil line
Here, pixels having less than –95.0 are grouped as non-vegetation area
The Perpendicular Vegetation Index 3 (PVI3) is an enhanced version of the original PVI, utilizing the red band as an independent variable in regression analysis to improve accuracy This modification specifically aims to minimize negative results, ensuring more reliable vegetation monitoring PVI3 provides a refined measurement for assessing plant health and biomass, making it a valuable tool in remote sensing applications.
Where, a: intercept of the soil line, b: slope of the soil line, pNIR: reflectance in the near infrared band, pRED: reflectance in the visible red band
Difference Vegetation Index (DVI): DVI weigh up the near-infrared band by the slope of the soil line (Richerdson and Wiegand., 1977) and is given as :
Where, g: the slope of the soil line
Similar to the PVI1, with the DVI, a value of zero indicates bare soil, values less than zero indicate non vegetation and greater than zero indicates vegetation
Weighted Difference Vegetation Index (WDVI): Like PVI, WDVI is very sensitive to atmospheric variations (Richerdson et al., 1977) and can be presented as,
Where, NIR : reflectance of near infrared band, RED: reflectance of visible red band, g: slope of the soil line
WDVI is a simple yet highly effective vegetation index, comparable to most slope-based vegetation indices By weighting the red band with the slope of the soil line, WDVI enhances the vegetation signal in the near-infrared band This approach also reduces the influence of soil brightness, leading to more accurate vegetation analysis.
Soil reflectance spectra depend on type of soil The vegetation indices computed earlier assume that there is a soil line, where there is a single slope in red-NIR space
Soils with varying red-NIR slopes often appear within a single imagery, complicating accurate analysis Additionally, inaccuracies in the assumption of the iso-vegetation line—whether parallel or intersecting at the origin—can lead to erroneous vegetation index readings, especially when soil moisture variations cause shifts along these lines The impact of soil noise is particularly pronounced in areas with sparse vegetation cover To address these challenges, indices like SAVI are employed to minimize soil interference and improve the reliability of vegetation assessments.
TSAVI1, TSAVI2, MSAVI1, MSAVI2 attempt to reduce soil noise by altering the behavior of the iso-vegetation lines All of them are ratio-based, and the way that they
Significance of study site
Hai Phong, a coastal city with 125km of seacoast, benefits from its coastal mangroves by mitigating natural disaster impacts and providing substantial economic advantages to local communities The use of Remote Sensing and GIS technology can significantly improve the management and conservation of these vital mangrove ecosystems Currently, there is a lack of published research on applying vegetation indices to coastal mangroves in Hai Phong, especially in Kien Thuy and Do Son districts Therefore, this study offers valuable contributions toward advancing technical methods for mangrove exploration, mapping, and sustainable management in the region, supporting environmental protection and economic development.
Study goals, Objectives and Methodology
Study goals and Objectives
This study aims to identify an optimal vegetation index for accurate coastal mangrove mapping using multi-temporal Landsat images in Dai Hop and Bang La communes, Hai Phong The findings establish a solid scientific basis for improving mangrove rehabilitation and restoration efforts These insights are crucial for addressing the impacts of climate change and sea-level rise on Hai Phong's vital mangrove ecosystems.
In order to reach the main goal of study, three specific objectives are given as below:
Objective 1:To investigate the spatial distribution and structures of coastal mangrove species in Dai Hop and Bang La communes, Hai Phong
Objective 2: To determine different vegetation indices for classifying mangroves in Dai Hop and Bang La communes, Hai Phong
Objective 3: To quantify the spatial dynamics of coastal mangroves in study areas during period 2010 - 2014.
Study scope
With period of five months (May to October 2014), the study focuses on determining coastal mangrove classes in Dai Hop commune, Kien Thuy district and Bang
La ward , Do Son county, Hai Phong city.
Methodology
3.3.1 Investigation and determination of coastal mangrove species composition and its habitat
The field survey comprised two phases: calibration data were collected in May 2014, followed by accuracy assessments in September 2014 A total of 335 square sample sites (30m x 30m) were established to ensure consistent pixel size, with comprehensive data recorded at each site, including species composition, tree height, tree density, and crown diameter, providing a robust dataset for ecological and remote sensing analysis.
Species composition was visually estimated from a 5m 2 (2 *2.5 m) plot marked by a tape measure
Tree height and crown diameter were measured using a 5 m telescopic pole
Tree density was measured by counting the number of tree trunks at breast height
When a tree forked beneath breast height (1.3 m) each branch was recorded as a separate stem The location of each field site was determined using GPS with a probable circle error of 2-5 m
Canopy closure or coverage percentage is estimated using a densitometer, a device constructed from a 2.5 cm diameter duct pipe approximately 40 cm long, with cross-hairs at both ends created using fine wire The densitometer is held vertically to assess the proportion of the sky blocked by the canopy, with readings indicating whether the view is less than, equal to, or greater than 50% obscured by leaves, branches, or trunks These measurements are recorded every meter along a 100-meter transect within a 100 m² plot, resulting in approximately 100 data points for accurate canopy coverage estimation.
A habitat classification for the mangrove areas of Dai Hop and Bang La was developed using hierarchical agglomerative clustering with group-average sorting, based on calibrated data To ensure balanced contributions, data were transformed to weight tree height, crown diameter, and density more evenly with species composition, which otherwise varied by an order of magnitude The clustering analysis employed an 85% similarity threshold to define habitat categories, which were characterized by mean or median values of species percent composition, tree height, tree density, and crown diameter.
These habitat categories would use to direct the image classification of the Landsat data, and the collection of accuracy data in September 2014
3.3.2 Calculation of different types of vegetation indices used to classify mangroves
Vegetation indices are derived from the near Red and Infrared bands of Landsat images, enabling accurate identification of mangrove areas Mangrove classes were delineated and refined using field data, then consolidated into a single mangrove category for clarity The final steps include assessing map accuracy and conducting post-processing to ensure reliable results before exporting the finalized map.
GIS applications to perform VI‟s classification and mapping
Remote sensing data for study
Landsat satellite images are used to calculate vegetation indices and to classify coastal mangroves (Table 3.1)
Table.3.1 Landsat data used this study
Year Landsat image code Date Resolution Path/Row
Source: http://earthexplorer.usgs.gov
Convert DN values to TOA reflectance:
At this stage, atmospheric and radiance corrections were applied to the image for improving the quality of vegetation classification by using the Spatial Analyst Tools in
ArcToolbox => Spatial Analyst Tools => Map Algebra => Raster Calculator:
Wherre ML: Band-specific multiplicative rescaling factor from the metadata
(radiance_Mult_Band_x, where x is the band number)
AL: Band-specific additive rescaling factor from the metadata (Radiance_add_band_x, where x is the band number)
QCal: Quantized and calibrated standard product pixel values (DN)
Composite bands: The image was corrected to UTM zone 48 N, WGS 84 band and Green band of the Landsat Contrast, brightness and anomalies are verified before using as original image data
To accurately analyze the study area, it is essential to clip the image by isolating the specific region of interest This process involves using a file with the boundaries of the study area to extract the relevant portion from newspaper photographs or other images Clipping ensures focused analysis and improves the precision of spatial data for research purposes.
Fig 3.1: Clipped images of study sites in Hai Phong: (a) image in 2010, (b) image in 2013, (c) image in 2014
3.3.2.3 Calculating vegetation indices for mangrove classification
To calculate vegetation index, the study uses ArcGIS 10.1 with the extension of Spatial Analysis Tools
Step 1: Calculation vegetation index uses Red band (0.636 – 0.673 àm) and NIR band (0.851 – 0.879 àm) of the image
Calculation of NDVI: NDVI = (NIR - RED) / (NIR + RED)
Calculation of SVI: SVI = NIR / RED
SAVI = {(NIR - RED) / (NIR+RED+ L)} * (1+L), where L = 0.5 to minimize the soil influence
Because of vegetation indices image are generated with a range of values from -
NDVI and SAVI values typically range from +1 to -1, while SVI values extend from 0 to infinity To ensure all values have an equal probability distribution across the range 0-255, these indices are transformed and scaled accordingly This normalization process enhances the comparability of images and optimizes their suitability for various analysis tasks.
Vegetation indices were analyzed based on digital values ranging from 0 to 255, enabling precise assessment of plant health and density Mangrove habitat categories are determined through the classification of pixels according to their digital reflectance values, which correspond to specific vegetation conditions The reflectance range for each habitat category is estimated based on these digital values, facilitating accurate mapping and monitoring of mangrove ecosystems.
Estimating the accuracy of classified images against ground conditions is essential for comparing classification techniques and assessing potential errors in land cover analysis Accuracy is typically evaluated using an error matrix, which is a square table highlighting the classification's correctness; its main diagonal indicates overall agreement The error matrix clarifies omission errors, where true land cover is missed, and commission errors, where areas are misclassified User’s and producer’s accuracies provide detailed insights into these errors, ensuring reliable land cover classification assessments for further environmental and spatial analyses.
Test areas of the same size as the sample sites were used to determine the agreement of the classification with the ground reference data
The size of the sample (n) to be used in accuracy assessment can be estimated using equation :
Z = 2 (from the standard normal deviate of 1.96 for the 95% two-sided confidence level)
For example, the expected accuracy is 85% at an allowable error of 5% (i.e., it is 95% accurate), the number of sample sites necessary for reliable results is: N = a minimum of 203 sites
Accuracy assessment involves randomly sampling ground sites to verify true land cover types, with a values file recording the land cover class (by integer index) at each location This file is combined with a vector point file to create a raster image of true classes, which is then compared to the classified map An error matrix is generated to detail the relationship between actual land cover and mapped classes Overall accuracy is calculated from the comparison of classification outputs, ensuring precise evaluation of land cover classification quality.
After image classification, the procedure of post classification is taken to remove the salt and pepper of classified pixels
To determine the size of a map unit in meters, divide the map scale resolution by 1,000 The calculation involves using the formula: map resolution rate (meters) multiplied by 2 and then by 1,000 This method helps establish a clear relationship between the sample resolution and the actual ground measurement, ensuring accurate mapping data.
Remote sensing data with a spatial resolution of 30 meters was utilized in this study, resulting in an appropriate map scale of 1:60,000 based on the specified formula To create a comprehensive map, it is essential to include additional details such as projection systems, annotations, a ruler, and compass orientation, ensuring accurate and reliable spatial representation.
3.3.2.7 Assessing and comparing different kinds of vegetation indices for mangrove mapping
Percent of accuracy and number of mangrove habitat categories are compared for
Quantifying spatial dynamics of coastal mangroves in study areas during period 2010 – 2014
To quantify the dynamics of coastal mangroves in different periods, the spatial analyst tools in ArcGIS is used These steps are described as the following:
STUDY SITE, NATURAL AND SOCIOECONOMIC FEATURES
Natural characteristics
The study area encompasses two districts, Kien Thuy and Do Son, specifically within the boundaries of two communes: Dai Hop in Kien Thuy and Bang La in Do Son, as shown in Fig 4.1.
Fig.4.1 Study sites in Hai Phong where: (a) Viet Nam map, (b) Hai Phong city map (c) Selected sites as Kien Thuy District and Do Son County
As shown in Fig.4.1 that Bang La commune is geographically located at 20°42‟42‟‟N and 106°44‟43‟‟E, with 966.73 ha of natural area Contiguous zones include:
East (Ngoc Xuyen and Van Huong communes), West (Kien Thuy district), South (Gulf of Tonkin), North (Minh Duc commune and Kien Thuy district)
Dai Hop commune is situated at coordinates 20°41′45″N and 106°42′51″E, covering approximately 1,097.79 hectares of natural area The commune boasts a coastline stretching 4.2 kilometers, enhancing its maritime significance It shares borders with Bang La commune to the east, Doan Xa commune to the west, Van Uc seaport to the south, and neighboring areas to the north, positioning it as a key geographic location with strategic connectivity.
Tu Son commune Today, Dai Hop commune includes 4 villages: Dai Loc, Quan Muc, Dong Tac and Viet Tien
Study area is far from city center Topography is not flat, separated by river, wrapped by sea and along the northwest - southeast The soil has high acidity and salinity
Hai Phong has a subtropical climate characterized by four distinct seasons: spring, summer, autumn, and winter The region experiences seasonal variations in precipitation influenced by monsoon patterns, with the rainy season occurring from May to October and monsoon winds bringing rain and cooler temperatures during summer Conversely, the dry season spans from November to April, featuring cold and dry winter winds The average summer temperature is around 32.5°C, while winter averages approximately 20.3°C, with the annual mean exceeding 23.9°C Annual precipitation typically ranges between 1,600 to 1,800mm, and humidity levels remain high at 85-86%, peaking in June, July, and August Dai Hop and Bang La are annually threatened by 1-2 storms from June to September, highlighting the region's vulnerability to tropical typhoons during peak storm season.
River network in Hai Phong is dense Total length of entire river network through city is nearly 280 kilometers, average density is 0.6-0.8km/1km 2 Slope is quite small,
Da Bac, Cam, Lach Tranh, Van Uc, Thai Binh, Bach Dang It provides abundant water source to serve human demand
Dai Hop and Bang La are located in the coast so it contains mainly alkaline soil, saline soil, sediment, feralit…
Forest resource: Dai Hop has nearly 450 hectares of mangroves along 4.2 kilometers of coastline Bang La has 730 hectares of mangroves in total Mangroves bring large benefits to aquaculture
Water resource: Dense water network (rivers, ponds) creates conditions to develop aquaculture
Sea resource: Coastline is so long and near estuary that it is advantageous for catching fish and anchoring.
Socioeconomic conditions
According to statistic in April 1 st 2009, population in Dai Hop (Kien Thuy) was
Dai Hop, a commune in Kien Thuy District, has a population of 9,491 people across 2,675 households, with an average population density exceeding 865 people per square kilometer Agriculture remains a dominant livelihood, with 69.5% of households engaged in farming and 8.1% involved in fishing The working-age population makes up about 46%, while approximately 600 residents are employed abroad Known for its high living standards, Dai Hop reported an average income of 13 million VND in 2008, reflecting a 53.8% increase compared to 2000.
According to the most recent statistic, population size in Bang La commune was
8765, average population density was 1339 people/km 2 , greater than Dai Hop commune
Dai Hop's economy primarily revolves around aquaculture and aquatic product exploitation, with aquatic product exploitation representing 43.5% and aquaculture 30.5% of economic activity in 2008 The remaining 26% comprises small-scale industry and service sectors Currently, Dai Hop is actively developing its economy by leveraging its rich fishing potential to promote sustainable growth and boost local livelihoods.
In Bang La, local authorities are focused on developing the economy through interdisciplinary approaches, emphasizing aquatic product exploitation The economic restructuring has shifted away from traditional salt-making, which was associated with low income, toward expanding animal husbandry, horticulture, and business services Additionally, numerous welfare projects—including improvements to roads, lighting, schools, clinics, markets, and irrigation systems—have been implemented, alongside the construction of hundreds of multi-storied villas, contributing to the area's overall development and modernization.
4.2.3 Ecological and economic values of mangroves
Dai Hop (Kien Thuy) and Bang La (Do Son) feature expansive mangrove forests along their sea dikes, vital for coastal protection and ecological health These mangroves play a crucial role in shielding local communities from extreme weather events while supporting economic development and enhancing biodiversity Their presence is essential for maintaining ecosystem stability, promoting sustainable livelihoods, and preserving the natural coastal environment.
Mangrove forests support a diverse ecosystem, hosting shrubs, birds, crabs, and other organisms that interact through a complex food web During tidal movements, these creatures forage, reproduce, and contribute to the ecological cycle, maintaining the health of the habitat Plants in mangroves provide organic material for crabs, fish, and shellfish, while also serving as a vital food source for various raptors When organisms die or excrete waste, their remains decompose and recycle nutrients, replenishing the organic substances essential for the ecosystem's sustainability This intricate process of energy transfer and biological interactions ensures the ecological balance and resilience of mangrove forests.
Mangrove forests play a crucial role in maintaining the ecological balance by supplying essential nutrients and oxygen to land and aquatic life through photosynthesis They serve as vital links between terrestrial and marine ecosystems, helping to stabilize and protect these peripheral environments The importance of mangroves in preserving ecological integrity and supporting biodiversity is undeniable, highlighting their value in ecosystem conservation.
Mangrove forests serve as vital foraging grounds and living habitats for diverse wildlife, providing abundant food resources and shelter from adverse weather Many species, including shrimps and various birds like egrets, depend on mangroves throughout their life cycles These ecosystems are especially important for waterfowls, which frequently forage for aquatic organisms such as fish, crabs, worms, and shrimps Consequently, mangroves are recognized as essential habitats that support biodiversity and serve as ideal environments for many wildlife species.
Mangrove root systems play a crucial role in reducing water pollution by absorbing inorganic substances from water They help lower levels of suspended particles, nitrogen, phosphorus, metals, and chemical oxygen demand through mechanisms such as microbial metabolism, soil surface absorption, chemical sedimentation, and plant digestion These natural processes make mangroves an effective solution for improving water quality and protecting aquatic ecosystems.
Flood prevention and protecting the coastlines: Mangroves can stabilize coastlines of the river shores and river mouths They also protect the coastline from wave erosion
Not only that, mangroves can stabilize water capacity of the substratum and soil surface Therefore, it helps steady and retain water to prevent flooding
Economic values: Each year, mangroves provide large amount of aquatic products
Mangroves significantly boost local incomes and enhance residents' quality of life by providing essential resources such as firewood to meet human needs Additionally, as a unique and vital ecosystem, mangroves offer valuable benefits for eco-tourism, attracting visitors and supporting sustainable development.
RESULTS AND DISCUSSION
Spatial distribution and structures of coastal mangroves in study sites
Spatial distribution of coastal mangroves
Hai Phong city, a coastal area with a 125 km dike system and approximately 152,000 hectares of natural land, has significantly expanded its coastal mangrove forests—from 293 hectares in 2012 to over 4,700 hectares—thanks to dedicated rehabilitation and restoration projects These efforts have been supported by both international and national funding sources, including the PAM 5325 program, the Red Cross initiative, and the ACMAMG organization's Japanese-led mangrove restoration programs.
The field survey also identified that spatially ecological distribution of mangroves in Dai Hop and Bang La (Fig 5.1)
Fig 5.1: Species distribution of coastal magroves in Bang La and Dai Hop, Hai Phong
The coastal mangroves in Hai Phong primarily consist of a simple species composition, including Kandelia obovata, Soneratia caseolaris, Bruguirea gymnorrhiza, Aegiceras corniculatum, and Acanthus Ebracteatus These mangrove habitats are typically characterized by muddy substrates, low slopes, and low oxygen levels, creating challenging conditions Additionally, they experience high water levels, low salinity, and slow tides, which influence the growth and distribution of mangrove species in the area.
Structural characteristics of coastal mangroves :
Based on these data as well as investigate the field, quality of mangroves in study area was assessed by following criterion:
Mangrove areas in the study region extend over 7 kilometers along the coastal dyke and reach more than 1 kilometer into the sea, covering a total area of over 600 hectares Specifically, the Dai Hop commune manages approximately 410 hectares of mangroves near the Van Uc estuary, while the remaining mangrove forests are situated in Bang La commune, near Do Son beach These mangrove ecosystems play a crucial role in coastal protection, biodiversity, and local livelihoods.
In the study area, the primary mangrove species identified are Kandelia obovata, Sonneratia caseolaris, Ancanthus ebracteatus, Avicennia marina, and Bruguiera gymnorhiza Among these, Kandelia obovata and Sonneratia caseolaris are the dominant species at the surveyed sites, highlighting their crucial role in the local mangrove ecosystem.
Mangrove density, defined by the number of plants per unit area, is a key characteristic of mangrove populations Higher density levels reflect stronger interactions among individual trees and contribute to the overall structure of the mangrove ecosystem Dense mangrove forests play a crucial role in reducing wave height, mitigating wind and ocean tidal effects, and providing vital protection to coastal areas Additionally, increased mangrove density supports the development of intertidal zones, enhancing biodiversity and ecosystem resilience.
Canopy cover, representing the percentage of canopy coverage within a forest, is a crucial factor for evaluating forest growth rates and understanding how much sunlight is absorbed or passes through the canopy The canopy covering ratio directly influences photosynthesis and overall forest health, making it an essential parameter for assessing forest quality During fieldwork, the covering percentage in each plot was recorded, ranging from 20% to 90%, providing valuable data for forest analysis.
Canopy cover in Dai Hop commune is relatively uniform, with a small variation between a maximum of 75% and a minimum of 60%, covering a wide area of 410 hectares as per 2006 local administration data, indicating good mangrove quality In contrast, canopy cover in Bang La varies significantly, with some plots exhibiting a high coverage of up to 90%, while others have only 20%, creating favorable conditions for the growth of the low tree layer, specifically Ancanthus ebracteatus.
Coastal mangroves have been planted since 1999 Our findings shown that mangrove regeneration is very good, occuring mostly closed to sea dikes and majority of
Mangrove structures in Bang La and Dai Hop can be categorized into three main zones from inland to seaward, as illustrated in Fig 5.1 The innermost area near the sea dyke features a mix of Kandelia obovata and Soneratia caseolaris species The middle zone comprises a combination of Kandelia obovata, Soneratia caseolaris, Bruguiera gymnorrhiza, and Aegiceras corniculatum, highlighting species diversity The seaward zone is dominated by pioneer mangroves, primarily consisting of Soneratia caseolaris alongside Acanthus Ebracteatus, indicating a natural transition toward colonization and stabilization of coastal areas.
Key mangrove structures are vital indicators of mangrove health, assessed through field surveys that focus on essential elements such as canopy height (H under), total tree height (H total), diameter at breast height (DBH), canopy diameter, and canopy cover A total of 47 standardized plots were established for the structural assessment, with the collected data summarized and averaged in Table 5.1, providing valuable insights into the condition of mangrove ecosystems.
Table 5.1: Synthesis of average mangrove structure characeteristics
Sonneratia caseolaris exhibits the greatest average height among mangrove species, with some trees exceeding 7 meters in height The canopy diameter varies among species, with Kandelia obovata averaging 1.4 meters and Bruguiera gymnorrhiza averaging 1.9 meters, indicating their significant canopy spread Conversely, Avicennia marina and Acanthus ebracteatus display the smallest canopy diameters, highlighting the diversity in growth patterns within mangrove ecosystems These measurements are essential for understanding mangrove biodiversity and ecosystem dynamics.
Bruguiera gymnorrhiza is the tallest species, reaching heights of up to 1.6 meters The total height is highest in Avicennia marina, followed by Acanthus ebracteatus and Kandelia obovata Conversely, Acanthus ebracteatus exhibits the smallest height among these species, measuring less than 1 meter.
Mangroves typically have a diameter greater than 30 cm, with Sonneratia caseolaris exhibiting the largest diameters among species studied The average diameter at breast height (DBH) for Kandelia obovata, Bruguiera gymnorrhiza, and Avicennia marina is approximately 10 cm, while Acanthus ebracteatus has the smallest DBH, measuring less than 6 cm Additionally, the canopy structure varies across species, influencing their ecological roles in mangrove ecosystems.
Comparison of different kinds of vegetation indces for mangrove classification
Calculation of vegetation indices in Dai Hop and Bang La commune
The VIs calculation results reveal the range of reflectance values for each habitat category (Table 5.2) The classification of mangrove classes based on these reflectance values, after applying different VI models, is illustrated in Figures 5.2a, 5.2b, and 5.2c These findings demonstrate how vegetation indices effectively distinguish mangrove habitats according to their reflectance characteristics.
Table.5.2.1 Values of vegetation Indices for mangrove classification
SVI (Simple Ratio Index); NDVI (Normalized Difference Vegetation Index); SAVI (Soil- Adjusted Vegetation Index)
Fig.5.2 Mangrove species in different vegetation indices: (a) SVI – Simple Vegetation Index; (b) NDVI- Normalized Difference Vegetation Index; (c) SAVI- Soil Adjusted Vegetation Index.
Accuracy Assessment of mangrove classification
Table 5.2.2 Accuracy assessment of image classified using SVI in 2014
Field work Kandelia and Sonneratia
Water Wetlands Agriculture Built-up Total Accuracy
Table.5.2.3 Accuracy assessment of image classified using NDVI in 2014
Field work Kandelia and Sonneratia
Water Wetlands Agriculture Built-up Total Accuracy
Table.5.2.4 Accuracy assessment of image classified using SAVI in 2014
Field work Kandelia and Sonneratia
Water Wetlands Agriculture Built-up Total Accuracy SAVI Image (%)
Table.5.2.5.Summarization of vegetation indices for mangrove classification
Table 5.2.5 indicates that SVI has the lowest accuracy at 68.3%, effectively distinguishing three classes According to Graetz (1990), canopy texture significantly influences the SVI model’s performance While SVI can effectively identify soil and vegetation, its accuracy diminishes in shady areas, emphasizing the importance of canopy texture in remote sensing analyses.
The study found that SAVI achieved the highest classification accuracy at 75.8%, making it the most effective vegetation index for classifying mangroves Adjusting the L parameter from 0.5 to 0.16, as recommended by Rondeaux et al (1996), improved the detection of mangroves by reducing soil noise across varying vegetation densities SAVI’s consistent sensitivity across the full range of vegetation cover also allows it to be effectively used for general vegetation classification Additionally, NDVI achieved an accuracy of 72.5%, closely comparable to SAVI, indicating that NDVI is also a reliable indicator for discriminating different vegetation classes.
Dynamics of coastal mangroves during 2010- 2014
5.3.1 Thematic maps and dynamics of coastal mangroves
Based on the findings in section 5.2, there is little difference between SAVI and NDVI for mangrove classification; therefore, this study employs NDVI (Normalized Difference Vegetation Index) to analyze coastal mangrove changes from 2010 to 2014 The results, presented in Table 5.3, indicate that the extent of coastal mangroves has generally increased over this period.
Classifier Total Accuracy Number of mangrove classes
2010 The extent of mangroves accounted for 381.8 ha in 2010, increased to 460.5 ha in
Between 2010 and 2013, mangrove areas expanded by 78.7 hectares, increasing from 393.2 hectares to 472.8 hectares, and continued to grow by an additional 12.5 hectares in 2014 In contrast, non-mangrove areas declined steadily during this period, decreasing by 12.2 hectares from 2010 to 2013 and a further 60.9 hectares from 2013 to 2014 The water-covered area experienced fluctuations between 2010 and 2014, decreasing by 102.9 hectares initially, then increasing again by 48.6 hectares in 2014 Overall, the data indicates a trend of mangrove expansion alongside a reduction in non-mangrove land and variable water coverage over the years.
Table 5.3.1: The extent of coastal mangroves in the study areas (ha)
Year Mangroves Non-mangroves Water
Fig 5.3.1: Distribution of mangrove extents during the period 2010 - 2014
Dynamics of coastal mangroves during period 2010 – 2014
In general, coastal mangroves have relatively small fluctuations during the period
2010 to 2014 (Table 5.3.2 and Table 5.3.3), but its extent increased over last 4 years The extent of mangroves increased by 20.6% from 2010 to 2013, greater than the period of
2013 - 2014 (2.7%) In contrast, the areas with non-mangroves in period 2010 - 2013 (6.3%) is less than period 2013 - 2014 (33.3%) It indicates that mangrove restoration and rehabititation programs are effective during this period (2010- 2014) these study sites
Table 5.3.2: Dynamic of mangroves during period 2010 -2013
Table 5.3.3: Dynamic of mangroves during period 2013 -2014
Fig 5.3.2: Spatial dynamics of coastal mangroves in study sites during two periods
Fig 5.3.1 and Fig 5.3.2 shows the extents of mangroves, non- mangroves and waters increased or decreased during the certain period as the following:
Period 2010 - 2013: In this period, mangroves area has increased significantly
Between 2010 and 2013, mangrove coverage increased from 381.8 hectares to 472.8 hectares, primarily due to active planting and conservation efforts aimed at disaster preparedness This significant growth was supported by the central government through the 327 Program (now the 661 Program) and initiatives by the Danish Red Cross, which collectively contributed to expanding and protecting mangrove forests.
Period 2013 – 2014 : In 2013, there were 219.87 ha of mangroves and forest area
Phong city focuses on mangroves protecting activities and propaganda activities, so mangroves area and quality increases continuously
Fig 5.3.3: Fluctuation of mangroves area in study area
5.3.2 Key drivers of coastal mangrove changes from 2010 to 2014
Between 2010 and 2014, the extent of mangroves showed a general increasing trend, growing from 381.8 hectares in 2010 to 472.8 hectares in 2014 The total increase in mangrove area over this period was 91 hectares, with the majority (78.7 hectares) added between 2010 and 2013 Key drivers contributing to the expansion of coastal mangroves include effective conservation efforts and reforestation initiatives.
Policies serve as essential guidelines for the conservation and sustainable development of forest ecosystems, with a particular focus on mangroves Since the 1990s, both regional and global awareness has heightened regarding the importance of implementing policies to protect coastal mangroves The development of effective policy frameworks has been crucial in promoting the preservation and sustainable management of these vital ecosystems.
Years Mangroves Non-mangroves Water government, which was performed in the area of coastal provinces in Northern Vietnam, were born in 1990s
Since 1998, detailed coastal mangrove management policies have been implemented at the study site The research area spans two adjacent communes, Bang La and Dai Hop, which are directly affected by these policies These mangrove regions extend over a 7 km strip outside the sea dyke, influenced by two integrated management systems and coordinated by dedicated teams from each commune.
Dai Hop and Bang La have both maintained their ranger departments annually to protect local ecosystems However, differences in community awareness have led to varying mangrove quality between the two communes Since 2005, local authorities and the Hai Phong Red Cross have continued implementing ongoing conservation projects to improve mangrove preservation.
In conclusion, policies and projects on coastal mangrove management in the study site have been shaped by the evolving stages of Vietnam’s natural resources policies, reflecting common features across different eras Since the primitive period (1990–1997), there have been positive policy changes that contributed to the recovery and development phase (1997–2005), followed by a focus on sustainable development from 2005 to 2014.
A result of interviewing local households in Dai Hop and Bang La communes and local administrations shown that there were the significant changes in relation to local livelihood.
GENERAL CONCLUSION, LIMITATION AND FURTHER STUDY
Conclusions
This study examines coastal mangroves in Kien Thuy and Do Son districts, Hai Phong, from 2010 to 2014 Key findings reveal significant changes in mangrove coverage and health over the period The research highlights the importance of conservation efforts to protect these vital ecosystems Additionally, the study emphasizes the need for sustainable management practices to ensure the long-term preservation of Hai Phong's mangrove forests.
Study sites are located in Kien Thuy and Do Son district, Hai Phong city with, nearly 8000 ha Coastal mangrove species are mainly Sonneratia caseolaris and Kandelia obovata
Three vegetation indices—SAVI, NDVI, and SVI—are effective for mangrove classification, with SAVI (75.8%) and NDVI (72.5%) demonstrating higher accuracy than SVI (68.3%) These results align with previous research, indicating that NDVI and SAVI are more reliable for coastal mangrove classification Therefore, it is recommended to prioritize NDVI and SAVI for accurate mangrove ecosystem mapping.
The extent of coastal mangroves generally increased from 2010 to 2014 Mangrove extents accounted for 381.8 ha in 2010, increased to 460.5 ha in 2013 and 472.8 ha in
2014 The key drivers significanly contribute to an increase of coastal managroves, namely good management policies on coastal management and improved local livelihoods.
Limitations and further study
Despite achieving some significant results, the study has notable shortcomings, including the large scope which may overlook the loss of certain coastal mangrove species To address these limitations and enhance accuracy, future research should utilize higher spatial resolution images such as SPOT and QuickBird to provide more precise and detailed insights into mangrove diversity and distribution Implementing advanced remote sensing techniques will help overcome current gaps and support better conservation strategies.
In addition, a number of sampling points should be increased for enhancing the accuracy of image classifications In addition, future study should use higher spatial resolution images for mangrove mapping
[1] Badhwar, G.D; Verhoef, W.; Bunnick,N.J.J., (1985) Comparative of suits and SAIL canopy reflectance models Remote Sensing of Environment ,17: 179-260
[2] Baret F, Guyot G and Major D (1989), TSAVI: A Vegetation Index Which Minimizes
Soil Brightness Effects on LAI and APAR Estimation, 12th Canadian Symposium on Remote Sensing and IGARSS'90, Vancouver, Canada, Vol 4
[3] Baret F Guyot G (1991), Potential and Limits of Vegetation Indices for LAI and AFAR
Assessment Remote Sensing and the Environment, Vol 35, pp 161-173
[4] Bannaii A, Huete A R, Morin D and Zagolski (1996), EfFets de la Couleur et de la
Brilliance du sol sur les Indices ds Vegetation, International Journal of Remote Sensing, Vol 17, No 10, pp 1885-1906
[5] Congalton , R.G , (1991) A review of assessing the accuracy of classification of remoted sensed data Remote Sensing Environment , 37(5): 35-46
[6] Elmore , A.J., Mustard , J.F., Minning , S.J 2000 Quantifying vegetation change in semiarid environment : Presicion and accuract of spectral mixture analysis and the normalized difference vegetation index Remote Sensing of Environment , 73: 83-102
[7 Tuan, L X., Y Munekage, Q T Q Dao, N H Tho and P T A Dao (2003)
Management in Mangrove Areas Environmental Informatics Archives 1: 38-52
[8] Gao, J., Chen, H.F., Zhang, Y., & Zha, Y (2004) Knowledge based approaches to accurate mapping of mangroves from satellite data Photogrammetric Engineering and Remote Sensing, 70, 1241-48
[9] Gong, P., Pu, R., and Miller, J.R 1995 Conifers forest leaf area index estimation along the Oregon transect using Airborne Spectrographic Imager data Photogrammetric Engineering and Remote Sensing , 61: 1107-1117
[10] Graet, R.D 1990 Remote sensing of terrestrial ecosystem structure : an ecologist’s pragmatic view Pp 5-30 InR.J Hobbs and H.A Mooney (Eds.), Remote sensing of Biospere Functioning New York : Springer – Vrlag
[11] Green , E.P.; Clark,C.D.; Mumby, P.J.; (1998) Remote sensing technique for mangrove mapping Int J of Remote Sensing , 19 (5):935-956
[12] Guyot, G., & Gu, x (1994) Effect of Radiometric Corrections on NDVI-Determined from SPOT-HRV and Landsat-TM Data Remote Sensing of Environment, 49,169-
[13] Huete A R (1988), A Soil Adjusted Vegetation Index (SAVI), Remote Sensing and the
[14] Jiang , Z., Huete, A.R., Chen, j., Chen, Y., 2006 Analysis of NDVI and scaled difference vegetation index retrievals of vegetation fraction Remote Sensing of Environment , 101: 366-378
[15] Lillesand, T.M., and R.W Kiefer and J.W Chipman Remote Sensing and Image
Interpretation 6 th ed New York: Wiley, 2007
[16] Ramsey, E.W and Jensen, J.L 1996 Remote sensing of mangrove wetlands: Relating canopy spactra to site specific data Photogrammetric Engineering and Remote Sensing , 62(8):939-948
[17] Richerdson A J, and Wiegand C L (1977), Distinguishing Vegetation from Soil
Background Information, Photogramm Eng Remote Sens., Vol 43, pp 1541-1552
[18] Rondeaux, G.; Steven , M.; Baret, F., (1996) Optimization of soil adjusted vegetation indices Remote Sensing of Environment , 51: 375-384
[19] Rose J W Jr., Mass R H, Schell J A, Deering D W and Harlan J C (1974),Monitoring the Vernal Advancement and Retrogradation (Green Wave Effect) of Natural
[20] Rose J W Jr, Mass R H, Schell J A, and Deering D W (1973), Monitoring Vegetation
Systems in The Great Plains with ERTS, Earth Resource Technology Satellite-1 Symposium, Goddard Space Flight Center, Washington DC, pp 309-317
[21] Qi J, Chehbouni A, Huete A R, Kerr Y H and Sorooshian S (1994), A Modified Soil
Adjusted Vegetation Index Remote Sensing and the Environment, Vol 48, pp 119-126
[22] Suffian , I., (2001) Udentification of Vegetation Species in the East Coast Mangrove of Peninsular Malaysia Using Vegetation Index M.Sc Thesis , KoLej Universiti Sains dan Teknologi Malaysia, Terengganu , Malaysia 190p
[23] Hai-Hoa, N., (2013) The relation of coastal mangrove changes and adjacent land-use:
A review in Southeast Asia and Kien Giang, Vietnam Ocean and Coastal
[24] Tuan, L D., T T K Oanh, C V Thanh and N D and Quy (2002) Can
Gio Mangrove Biosphere Reserve Ho Chi Minh city, Agricultural Publishing house
[25] Wolanski, E., M M Brinson, D R Cahoon and G M E Perillo (2009) Coastal Wetlands:
An Integrated Ecosystem Approach The Netherlands, Elsevier
[26] Campbell, J.B., (1996) Introduction to Remote Sensing Taylor and Francis, London, 622p
[27] Dat, P T and K Yoshino (2013) Comparing Mangrove Forest Management in Hai
Phong City, Vietnam towards Sustainable Aquaculture Procedia Environmental Sciences 17(0): 109-118
[28] Donato, D C., J B Kauffman, D Murdiyarso, S Kurnianto, M Stidham and M
Kanninen (2011) Mangroves among the most carbon-rich forests in the tropics Nature Geosci 4(5): 293-297.