Navigation – Plan du site

Shoreline change of the Mekong River delta along the southern part of the South China Sea coast using satellite image analysis (1973-2014)

Evolution du littoral du delta du fleuve Mékong le long de la partie sud de la Mer de Chine méridionale à partir d’une analyse d’images satellites (1973-2014)
Manon Besset, Edward J. Anthony, Guillaume Brunier et Philippe Dussouillez
p. 137-146

Résumés

Le delta du Mékong est le troisième plus grand delta du monde, conséquence d'un contexte morphosédimentaire favorable d'apport en sédiments élevé et d’une croissance rapide au cours de l'Holocène. L’analyse des images satellites Landsat sur la période 1973-2014 montre que près de 70 % des 160 km de linéaire côtier du delta de la Mer de Chine Méridionale ont été en forte érosion. Ces résultats indiquent qu’à une tendance progradante sur le long terme s’est désormais substituée une évolution régressive du trait de côte au cours des dernières décennies dans cette partie du delta. Cette érosion n’est pas liée à l’abandon de lobe deltaïque, mais plus probablement à la diminution des apports fluviaux en sédiments et à des variations dans le stockage sédimentaire dans le delta qui semblent être liées à des modifications induites par l'Homme. Ces modifications concernent notamment le piégeage de sédiments fluviaux par des barrages, l’exacerbation de la subsidence liée à des abstractions massives d’eau, et des extractions de granulats sur les lits des chenaux deltaïques. L'érosion est par ailleurs aggravée par le remplacement des mangroves protectrices par des élevages de crevettes. Cette érosion constitue un danger supplémentaire pour l'intégrité future d'un méga-delta déjà considéré comme particulièrement vulnérable à la subsidence, et à de futurs barrages de grande capacité.

Haut de page

Notes de la rédaction

Soumis le 29 juin 2015, accepté le 21 janvier 2016.

Texte intégral

We acknowledge funding from the Belmont Forum Project ‘BF-Deltas: Catalyzing Action Towards Sustainability of Deltaic Systems with an Integrated Modeling Framework for Risk Assessment’. The SPOT 5 images were provided by the CNES/ISIS program (© CNES 2012, distribution Spot Image S.A.).

1. Introduction

1The Mekong River delta (fig. 1) is considered as the third largest delta in the world, with an area estimated at nearly 100,000 km² (Coleman and Huh, 2004). With 18 million inhabitants, the Mekong delta is subjected to intensive agricultural exploitation with paddy fields, fruit trees, as well as shrimp and fish farms, representing respectively 60 %, 70 % and 60 % of the total Vietnamese production (Mekong River Commission, 2010). The Mekong delta is commonly described as Southeast Asia’s rice bowl. It is linked to a river with a length of 4,750 km and a drainage basin of about 832,000 km² (Milliman and Ren, 1995). The estimated mean liquid discharge of the Mekong is about 14,500 m³/s (Mekong River Commission, 2010). The annual hydrological regime is seasonal with an Indian Monsoon flood season (May-October) during which fluvial sediment is delivered to the delta and the coast. Estimates of the mean annual sediment load of the Mekong at Kratie, Cambodia, just upstream of the delta (fig. 1), vary from 50 to 160 Mt (Walling, 2008; Milliman and Farnsworth, 2011; Lu et al., 2014). The Indian Monsoon season also corresponds to one of low energy waves from the southwest (fig. 1) that generate weak longshore currents towards the northeast. During this season, the high mud output from the Mekong is essentially stored in the nearshore area of the delta distributary mouths (Wolanski et al., 1998; Unverricht et al., 2013). The pattern contrasts sharply with the shorter low-flow dry season which is characterized by strong waves generated by the northeast Pacific Monsoon winds (fig. 1). Active alongshore transfer of sediments southwestwards from the mouths is assured by these energetic trade-wind waves and by wind-forced and tidal currents. The tidal range decreases from about 3 m at spring tides along the South China Sea coast to less than 1 m in the Gulf of Thailand, which is also relatively sheltered from the higher-energy Pacific Monsoon waves.

2The Mekong delta grew rapidly to form a 700-km-long coastline in the South China Sea from 5.3 to 3.5 ka cal BP at rates of up to 16m/yr of seaward progradation (Ta et al., 2002). With increasing exposure to ocean waves this rate dropped to less than 10 m/yr at the mouths. This rate remained high, however, up to 26 m/yr, in the muddy Cà Mau sector in the southwest (Ta et al., 2002). This difference in rates is responsible for the skewed morphology of the delta towards the southwest (fig. 1). This difference is also reflected in a grain size gradient, from predominantly sandy at the mouths, where the growth pattern became increasingly dominated by sandy beach ridges (Tamura et al., 2012), to a predominantly muddy western sector characterized in the past by mangrove vegetation.

Fig. 1 – Study area.
Fig. 1 – Secteur d'étude.

Fig. 1 – Study area.  Fig. 1 – Secteur d'étude.

A. Catchment area of the Mekong River and inset showing the six river basin countries. B. The Mekong River delta in Vietnam. The delta and part of its network of canals and dykes. C. The wave rose for the South China Sea (Wavewatch III data from National Center for Environmental Prediction (NCEP): http://polar.ncep.noaa.gov/​waves/​download.shtml).
A. Bassin versant du Fleuve Mékong et une figure montrant les six pays du bassin versant. B. Le delta du Mékong au Vietnam. Le delta et une partie de son réseau de canaux et digues, C. Rose des houles pour la Mer de Chine Méridionale (données Wavewatch III du National Center for Environmental Prediction (NCEP): http://polar.ncep.noaa.gov/​waves/​download.shtml).

3The Mekong delta is increasingly subject to the adverse impacts of a number of human activities that have been commonly described in river deltas (Syvitski, 2008; Evans, 2012; Anthony, 2013), notably upstream river damming, riverbed aggregate extractions, and water abstraction. The potential political and ecological impacts of hydropower dams in the Mekong has been of the object of various studies (e.g., Grumbine and Xu, 2012; Yong and Grundy-Warr, 2012; Ziv et al., 2012). Loisel et al. (2014) have suggested that an annual decrease in suspended sediment concentrations off the delta of about 5 % a year from 2003 to 2012 was very likely related to dam impoundment of sediment. Loisel et al. (2014) showed that this 5 % annual decrease in suspended sediment concentrations exiting at the mouths of the Mekong was neither due to changes in wave conditions nor river liquid discharge, both of which remained relatively stable over the period 1997-2012.

4Riverbed mining has also been practised on an increasingly larger scale over the last decade, driven by strong development pressures, especially in Cambodia and Vietnam (Bravard et al., 2013). Brunier et al. (2014) documented a net cumulative deficit of 200.106 m3 for two delta distributary channels from a 10-year comparison of bed volume changes, and attributed these losses to massive channel-bedload mining. The elevation of the surface of the delta has also been significantly affected by water extraction, considered as responsible for generating accelerated subsidence. Forestry and aquaculture activities have also been extensively developed to the detriment of vegetation, composed of Melaleuca and mangrove forests, the two main species being Rhizophora apiculata rather inland, and Avicennia Alba near the sea (Phan and Hoang, 1993). Finally, erosion of the delta shoreline has been highlighted recently in academic studies (Boateng, 2012; Phan et al., 2015) and in newspaper reports (Viêt Nam News, 2014).

5Anthony et al. (2013) showed from an analysis of SPOT 5 satellite images spanning the period 2003-2011 that erosion is particularly perceptible southwest of the river mouths, while Besset et al. (2015) demonstrated that the erosion of the Mekong delta is largely explained by high retreat rates of the coast along the South China Sea sector downdrift of the mouths. This part of the shoreline is dominated by muddy sediments and intertidal seafront vegetation essentially consisting of mangroves. These conditions are different in the still prograding sector of the mouths of the river where the sandy sediments are reworked into beach ridges capped by aeolian dunes (Tamura et al., 2012). In this paper, we specifically focus on erosion of the muddy South China Sea shoreline of the delta (fig. 1). The erosion of this part of the delta represents a significant reversal of the massive progradational trend that accompanied much of the deltaic growth in the course of the Holocene. This erosion threatens the integrity of the delta, and will increasingly exacerbate its vulnerability. Using long-term (1973-2014) sets of satellite images, we analyse the pattern of erosion of this part of the delta’s shoreline and the possible causes of this erosion.

2. Methodology

6In order to track recent deltaic shoreline changes along the South China Sea coast south of the mouths of the Mekong, Landsat images of 1973 and 2014, acquired from the United States Geological Survey, were used. The image resolution is 60m for the earliest L1-5 MSS (1973) images and 30 m for the most recent L8 OLI-TIRS (2014) images (tab. 1). The images were selected in order to avoid as much as possible clouds and high tide. Images selected were only those of winter (dry season/lowest river discharge). All images were rectified from the result of the USGS referencing in the Global Geodesic system WGS 1984 to Universal Transverse Mercator (UTM zone 48N). The picture readjustment was made using the SPOT5 images of 2003 and 2011 (resolution 2.5 m) for base georeferencing (tab. 1). The shoreline was digitized for each image using as a reference the external limit of vegetation.

Tab. 1 – Technical characteristics of the 1973 and 2014 satellite images and the computed uncertainties.
Tab. 1 – Caractéristiques techniques des images satellites de 1973 et 2014 ainsi que les marges d’erreur calculées.

Tab. 1 – Technical characteristics of the 1973 and 2014 satellite images and the computed uncertainties.   Tab. 1 – Caractéristiques techniques des images satellites de 1973 et 2014 ainsi que les marges d’erreur calculées.

7Following this, cross-shore shoreline mobility was statistically analysed using the ArcMap extension module Digital Shoreline Analysis System (DSAS), version 4.3, coupled with ArcGIS® v10.2.2 (Thieler et al., 2009). The mangrove fringe was adopted as a good ‘shoreline’ marker, verified from extensive field reconnaissance in 2011 and 2012. We calculated every 100 m alongshore the shore-normal distance of the vegetation line to a base line for the two sets of dates. This distance, chosen as a compromise between quality of the interpretation and total length of analysed shoreline (160 km) was then divided by the time in years between two dates to generate a shoreline change rate, the End Point Rate (EPR) in DSAS 4.3, expressed in m/yr.

          

           [1]   (Dolan et al., 1991)

We then defined the annual error (E) of shoreline change rate from the following equation (Hapke et al., 2006):

          

           [2]

where d1 and d2 are the uncertainty estimates for the successive sets of images and T is the time in years between image sets. The obtained exact error band of 2.31 m/yr between 1973 and 2014 was further increased to ± 5 m/yr, which we consider as an extremely cautious error range (tab. 1).

8Coastal area variations (km²) representing land losses or gains associated with changes in shoreline position were calculated from 1 km-alongshore segments between two successive image dates by dividing area variation by the time in years between dates. The error (Ea) expressed in km²/yr was calculated using a method similar to that of shoreline change rate for each 1 km segment based on the following equation (Hapke et al., 2006):

          

                  [3]

where ShaE1973 and ShaE2014 (km²) are the mean shoreline area error estimates for the successive sets of images and T is the time in decimal years between image sets. ShaE1973 and ShaE2014 were obtained from the mean square computation of surface errors every 1-km-alongshore segment. The obtained area error band of ±0.0092 km²/yr.

9In order to determine changes in mangrove cover liable to be an exacerbating factor in shoreline erosion, the spectral signatures of vegetation in two high-resolution (2.5 m) SPOT 5 images (2003 and 2011) were analyzed using ENVI® software, version 4.7. These images were analysed by Anthony et al. (2015) in order to determine shoreline changes, and the results will therefore serve to calibrate the longer-term (1973-2014) shoreline changes. This will also enable a check as to the recent status of the shoreline, especially in order to determine whether long-term trends observed from the Landsat images are exacerbated over the period 2003-2011 which has been characterized by significantly stronger anthropogenic pressures in the delta. Colorimetric indices were determined on the basis of spectral bands from the multispectral satellite images. Segmentation and classification were conducted to extract only those bands that define the vegetation cover, before being merged to reduce the size of the file compatible with GIS ArcGIS®. The supervised classification of the SPOT 5 image with the Normalised Difference Vegetation Index (NDVI) method was standardized to the maximum between the images used in order to render them comparable. The near infrared reflectance (Band 3) and reflectance acquired in the red spectral band (Band 2) of SPOT-5 optical sensor were used. The normalized ratio to estimate the amount of vegetation is:

          

                  [4]

10The final and merged class was close to the [65-150] pixel range. The coverage of the final class of the image was converted into a layer of vectors that was subsequently converted into an importable layer file in ArcGIS®. This work is repeated for each image, for the years 2003 and 2011 in order to compare vegetation occupation in the coastal fringe.

3. Results

11The South China Sea coast shows marked differences in rates of area change (fig. 2). The coastal surface losses attained 4.40 km²/yr over the period 1973-2014, and even more between 2003 and 2011 with a loss of 6.98 km²/yr. In the same way, the surface annual gains for the overall study period were 3.75 km²/yr. As far as the recent period between 2003 and 2011 is concerned, the surface gain is only 0.37 km²/yr. The surface trend rate over the periods 1973-2014 and 2003-2011 are respectively -0.65 km²/yr and -6.61 km²/yr. The assessment shows a clear imbalance over the period 2003-2011 between surface losses and gains.

Fig. 2 – Graphs of coastal area (km²/yr) change rates (erosion, accretion) for the South China Sea coast of the Mekong River delta between 1973 and 2014 (this study) and between 2003 and 2011 (Anthony et al., 2013) analysed from Landsat and SPOT 5 satellite images respectively.
Fig. 2Graphique des taux d’évolution de surfaces côtières (érosion, accrétion, en km²/an) pour le littoral de la Mer de Chine Méridionale du delta du Mékong entre 1973 et 2014 (cette étude) et sur la période 2003-2011 (Anthony et al., 2013) analysés respectivement à partir d’images satellites Landsat et SPOT 5.

Fig. 2 – Graphs of coastal area (km²/yr) change rates (erosion, accretion) for the South China Sea coast of the Mekong River delta between 1973 and 2014 (this study) and between 2003 and 2011 (Anthony et al., 2013) analysed from Landsat and SPOT 5 satellite images respectively.Fig. 2 – Graphique des taux d’évolution de surfaces côtières (érosion, accrétion, en km²/an) pour le littoral de la Mer de Chine Méridionale du delta du Mékong entre 1973 et 2014 (cette étude) et sur la période 2003-2011 (Anthony et al., 2013) analysés respectivement à partir d’images satellites Landsat et SPOT 5.

12Nearly 70 % of this part of the shoreline of the Mekong delta is in erosion between 1973 and 2014, and this retreat affects notably Cà Mau Province (fig. 3). Retreat rates commonly exceeded 20 m/yr in places, and with peaks of over 50 m/yr. The erosion rates are similar to those calculated by Anthony et al. (2015) using the higher-resolution SPOT 5 images (fig. 2). The overall dominant retreat has entailed a significant loss of deltaic land along this muddy coast (fig. 3-4).

Fig. 3 – Evolution of the South China Sea coast of the Mekong River delta.
Fig. 3 – Evolution du littoral de la mer de Chine méridionale du delta du Mékong.

Fig. 3 – Evolution of the South China Sea coast of the Mekong River delta.  Fig. 3 – Evolution du littoral de la mer de Chine méridionale du delta du Mékong.

A. Graph of coastal area (km²/yr) change rates between 1973 and 2014 analysed from Landsat satellite images. B. Graph of shoreline change rates (m/yr) between 1973 and 2014 analysed from Landsat satellite images. C. Graph of coastal area (km²/yr) change rates between 2003 and 2011 analysed by SPOT 5 satellite images from Anthony et al. (2015). D. Map of shoreline changes between 1973 and 2014 analysed from Landsat satellite images.
A. Graphique des taux d’évolution de surface côtière (km²/an) entre 1973 et 2014 analysés à partir d’images satellites Landsat. B. Graphique des taux d’évolution du trait de côte (m/an) entre 1973 et 2014 analysés à partir d’images satellites Landsat. C. Graphique des taux d’évolution de surface côtière (km²/an) entre 2003 et 2011 analysés à partir d’images satellites SPOT5 par Anthony et al. (2015). D. Carte des évolutions du trait de côte entre 1973 et 2014 analysées à partir d’images satellites Landsat.

Fig. 4 – Photograph (2012) showing typical shoreline retreat in the southwestern part of the Mekong Delta, with waves causing scouring of the muddy foreshore.
Fig. 4 – Photographie (2012) montrant un recul côtier représentatif de la partie est du Delta du Mékong, les vagues causant le creusement de l’avant-côte vaseuse.

Fig. 4 – Photograph (2012) showing typical shoreline retreat in the southwestern part of the Mekong Delta, with waves causing scouring of the muddy foreshore.  Fig. 4 – Photographie (2012) montrant un recul côtier représentatif de la partie est du Delta du Mékong, les vagues causant le creusement de l’avant-côte vaseuse.

13The significant land loss along this muddy shoreline of the Mekong delta has gone hand in hand with notable changes in land-use. This started with heavy mangrove downcutting during the Vietnam War from the 1960s to the early 1970s, followed by the overexploitation of wood in the 1980s and 1990s to provide timber for the construction industry and for charcoal production, and then by the installation of shrimp farms in the 2000s (Phan and Huang, 1993; Christensen et al., 2008). Figures 5 and 6 show examples of land-use changes and vegetation loss between 2003 and 2011, mainly consisting of seafront fringe mangroves, along a sector of the eroding coast in Bac Lieu Province. Over the 1,700 ha area used as an example, 60 ha of vegetation were lost during the period 2003-2011. In these sectors of eroding coast, mangroves have been replaced by aquaculture, particularly shrimp farming (fig. 8).

Fig. 5 – Evolution of vegetation cover between 2003 and 2011.
Fig. 5 – Evolution de la couverture végétale entre 2003 et 2011.

Fig. 5 – Evolution of vegetation cover between 2003 and 2011.  Fig. 5 – Evolution de la couverture végétale entre 2003 et 2011.

A. SPOT 5 satellite image of 2003 in a coastal sector of Bac Lieu Province. B. SPOT 5 satellite image of 2011 in the same sector. C. An example showing a comparative analysis of losses and gains of vegetation in the coastal sector of Bac Lieu Province. These changes are mainly associated with the loss of fringe mangroves between 2003 and 2011 identified from two high-resolution SPOT 5 satellite images.
A. Image satellite SPOT 5 de 2003 sur un secteur côtier de la province de Bac Lieu. B. Image satellite SPOT 5 de 2011 sur le même secteur côtier. C. Un exemple montrant une analyse comparative de surfaces perdues et gagnées de végétation sur le secteur côtier de la province de Bac Lieu. Les évolutions sont principalement associées à la perte de mangroves côtières entre 2003 et 2011 identifiées depuis deux images satellites SPOT 5 de haute résolution.

Fig. 6 – Photograph (2012) showing residual mangrove stands and ongoing erosion along the South China Sea coast of the Mekong delta.
Fig. 6 – Photographie (2012) montrant des mangroves résiduelles dans une zone en érosion du littoral de la Mer de Chine.

Fig. 6 – Photograph (2012) showing residual mangrove stands and ongoing erosion along the South China Sea coast of the Mekong delta.Fig. 6 – Photographie (2012) montrant des mangroves résiduelles dans une zone en érosion du littoral de la Mer de Chine.

4. Discussion

14The long-term (41 years) trend yielded by this comparison of the Landsat images thus shows that erosion has been a chronic characteristic of this southernmost part of the deltaic coast, compared to relative stability of much of the delta coast along the Gulf of Thailand and clear progradation in the northern half of the South China Sea coast including the sector of the delta mouths (Anthony et al., 2013; Besset, 2015). The Mekong delta has been strongly modified by humans since the early 18th century (Thanh, 2014). Although the strong erosion of the South China Sea shoreline of the Mekong delta is multi-decadal, this erosion predates the strong increase in human pressures on the delta associated with the construction of river dams, increasing aggregate extraction and subsidence over the last decade or so.

15The chronic erosion of the South China Sea coast of the delta represents a significant change compared to the Holocene progradational pattern associated with this part of the delta. Marked changes in deltaic progradation are generally associated with the mechanisms of lobe creation and abandonment that have been abundantly described in the literature on river deltas (e.g., Elliott, 1986). This mechanism cannot, however, be invoked for the Mekong delta since progradation has been associated with the twin constants of: (1) relatively fixed multiple river mouths, and (2) southwestwards dispersal fine-grained sediment to feed the strongly prograded South China Sea sector (Ta et al., 2002). This chronic erosion is probably caused by a decreasing sediment supply that may be linked to more important sequestering of fluvial sediment over the subsiding deltaic wetlands as human occupation of the delta has increased over the last decades. Sediment modelling has shown that this eroding part of the delta presently receives less than 2 % of the terrestrial mud supply stored in the mouth sector (Xue et al., 2012). The drop in this sediment supply has been matched by erosion of the muddy shoreline because of a less dissipative wave regime. Phan et al. (2015) have recently shown the importance of mud-induced wave dissipation along the Mekong delta shoreline. As erosion proceeds and mud is dispersed, steeper coastal bluffs of eroding mud have a less dissipative role on wave energy.

16In addition to likely mud trapping behind dams that may explain the drop in fine-grained sediment supply to the sea highlighted by Loisel et al. (2014), there are no doubt other additional links between this erosion and an increasingly human-impacted Mekong delta. The high population densities and large-scale agricultural, fishing and aquaculture activities have engendered massive engineering and infrastructure development throughout the delta (Manh et al., 2014, 2015). Accelerated subsidence may be leading to enhanced overbank trapping of fluvial sediments, while anthropogenic changes in channel morphology and sediment budget may affect sediment delivery to the delta shoreline. Manh et al. (2014) have estimated that the sediment deposited in the Mekong delta floodplain ranges from 1 % in a low flood year to 6 % in a high flood year relative to the total sediment load at Kratie. This corresponds to an annual spatial average floodplain deposition of 0.3-1.8 mm. These rates are much lower than the current subsidence rates of 2-3 cm determined by Erban et al. (2014).

17The South China Sea coast shows marked differences in rates of erosion, with accretion prevailing locally (fig. 3). The critical erosion along the southernmost sector of the delta may also be due to a more normal shoreline orientation to Pacific Monsoon waves. By diminishing the vertical tidal excursion, the lower tidal range along this coast should also enhance wave reworking of the shore. Finally, erosion must also be higher here compared to the northeast portion of the coast closer to the mouth sector from where sediment is redistributed towards the southwest. In addition to these factors, this marked alongshore variability (fig. 3) may reflect differences generated by the presence or absence of mangroves and sea dyke construction, both of which an indirect reflection of population pressures along the coast (fig. 7) and corresponding human transformations of the coastal fringe. Sea dykes have been built extensively along much of the coast for protection from marine flooding and for shrimp farms. This has resulted in ‘mangrove squeeze’ and lowering of the wave-dissipating capacity of mangroves, which form ‘fringe mangroves’ occupying a narrow coastal band along the South China Sea (Phan et al., 2015). Aspects of dyke construction for protection against coastal erosion have been discussed by Schmitt and Albers (2014) who noted that the criteria used in positioning these dykes behind narrow fringe mangroves were not always efficiently analysed, and could potentially exacerbate coastal erosion. However, dykes have generally been efficient in slowing down erosion (Schmitt and Albers, 2014). It is interesting to note that dykes are commonly absent along the most critically eroding sector of Cà Mau province of the South China Sea coast (fig. 3). This is also the least populated of the three South China Sea provinces of the delta south of the mouths (fig. 7).

Fig. 7 – Population statistics for the three provinces along the eroding South China Sea coast of the Mekong delta south of the mouths.
Fig. 7 – Densité de population pour les trois provinces le long the la côte de Mer de Chine Méridionale en érosion du Delta du Mékong au sud des embouchures.

Fig. 7 – Population statistics for the three provinces along the eroding South China Sea coast of the Mekong delta south of the mouths. Fig. 7 – Densité de population pour les trois provinces le long the la côte de Mer de Chine Méridionale en érosion du Delta du Mékong au sud des embouchures.

Fig. 8 – Photograph (2012) showing shrimp farms on the left, separated from mangroves to the right by an earth dyke.
Fig. 8 – Photographie (2012) montrant des installations d’élevages de crevettes à gauche, séparées de la mangrove, à droite, par une digue en terre.

Fig. 8 – Photograph (2012) showing shrimp farms on the left, separated from mangroves to the right by an earth dyke. Fig. 8 – Photographie (2012) montrant des installations d’élevages de crevettes à gauche, séparées de la mangrove, à droite, par une digue en terre.

5. Conclusion

18The once strongly prograding South China Sea coast of the Mekong delta southwest of the mouths is now largely prone to erosion, with retreat affecting over 70 % of this 160 km-long muddy coast. This trend has been prevalent at least over the last forty years, as suggested by analysis of satellite images. We consider a human-induced decrease in river sediment supply as the prime cause of this erosion, although other factors such as patterns of mud storage and redistribution across an increasingly subsiding delta plain large-scale and mangrove removal to make space for aquaculture may exacerbate the process. The numerous dams planned for the future in the lower Mekong catchment will aggravate the present delta destabilisation. Given the already high vulnerability of the Mekong delta to subsidence, the sediment supply necessary to balance this process will deprive the coast more drastically in fresh sediment, and, therefore, erosion will likely be exacerbated.

Haut de page

Bibliographie

Anthony E.J. (2013) – Deltas. In Masselink G., Gehrels R. (Eds) Coastal Environments: Dynamics, Climate Change and Management, John Wiley & Sons, Oxford, 299-337.

Anthony E.J., Dussouillez P., Goichot M., Brunier G., Provansal M., Dolique F., Nguyen V.L., Loisel H., Mangin A., Vantrepotte V. (2013) – Large-scale erosion of the Mekong delta: the role of human activities, Abstract presented at 2013 Fall Meeting, AGU, San Francisco. California, 9-13 December.

Anthony E.J., Brunier G., Besset M., Goichot M., Dussouillez P., Nguyen V.L. (2015) – Linking rapid erosion of the Mekong River delta with human activities. Scientific Reports, 5:14745.
DOI : 10.1038/srep14745

Besset M., Brunier G., Anthony E.J. (2015) – Recent morphodynamic evolution of the coastline of Mekong river Delta: Towards an increased vulnerability. Geophysical Research Abstracts, 17, EGU2015-5427-1, EGU General Assembly 2015, Vienna.

Boateng I. (2012) – GIS assessment of coastal vulnerability to climate change and coastal adaption planning in Vietnam. Journal of Coastal Conservation, 16, 25-36.
DOI : 10.1007/s11852-011-0165-0

Bravard J.P., Goichot M., Gaillot S. (2013) – Geography of sand and gravel mining in the Lower Mekong River. First survey and impact assessment. EchoGéo. URL: http:// echogeo.revues.org/13659.
DOI : 10.4000/echogeo.13659

Brunier G., Anthony E.J., Goichot M., Provansal M., Dussouillez P. (2014) – Recent morphological changes in the Mekong and Bassac river channels, Mekong Delta: The marked impact of river-bed mining and implications for delta destabilisation. Geomorphology, 224, 177, 177-191.
DOI : 10.1016/j.geomorph.2014.07.009

Coleman M., Huh O.K. (2004) – Major Deltas of the World: A Perspective from Space. Coastal Studies Institute, Louisiana State University, Baton Rouge, LA. www.geol.lsu.edu/WDD/PUBLICATIONS/C&Hnasa04/C&Hfinal04.htm, 74 p.

Christensen S.M., Tarp P., Hjortso C.N. (2008) – Mangrove forest management planning in coastal buffer and conservation zones, Vietnam: a multimethodological approach incorporating multiple stakeholders. Ocean & Coastal Management, 51, 712-726.
DOI : 10.1016/j.ocecoaman.2008.06.014

Dolan R., Fenster M.S., Holme S.J. (1991) – Temporal analysis of shoreline recession and accretion. Journal of Costal Research, 7, 723-744.

Elliott T. (1986) - Chapter 6 Deltas. In Reading H.G. (Ed.) Sedimentary Environments and Facies, Blackwell Scientific Publications, Oxford, 113-154.

Erban L.E., Gorelick S.M., Zebker H.A. (2014) – Groundwater extraction, land subsidence, and sea-level rise in the Mekong Delta, Vietnam. Environmental Research Letters, 9, 084010 (6 pp).
DOI : 10.1088/1748-9326/9/8/084010

Evans G. (2012) – Deltas: the fertile dustbins of the world. Proceedings of the Geologists’ Association, 123, 397-418.
DOI : 10.1016/j.pgeola.2011.11.001

Grumbine R.E., Xu J. (2012) – Mekong hydropower development, Science, 332, 178-179.
DOI : 10.2307/29784012

Hapke C.J., Reid D., Richmond B.M., Ruggiero P., List J. (2006) – National Assessment of Shoreline Change Part 3: Historical Shoreline Change and Associated Coastal Land Loss Along Sandy Shorelines of the California Coast. USGS Report, 13-14.

Kondolf G.M., Rubin Z.K., Minear J.T. (2014) – Dams on the Mekong: Cumulative sediment starvation, Water Resources Research, 50, 5158-5169.
DOI : 10.1002/2013WR014651

Loisel H., Mangin A., Vantrepotte V., Dessailly D., Dinh D.N., Garnesson P., Ouillon S., Lefebvre J.P., Mériaux X., Phan T.M. (2014) – Variability of suspended particulate matter concentration in coastal waters under the Mekong's influence from ocean color (MERIS) remote sensing over the last decade. Remote Sensing of Environment, 150, 218-230.
DOI : 10.1016/j.rse.2014.05.006

Lu X., Kummu M., Oeurng C. (2014) – Reappraisal of sediment dynamics in the Lower Mekong River, Cambodia. Earth Surface Processes and Landforms, 39, 1855-1865.
DOI : 10.1002/esp.3573

Manh N.V., Dung N.V., Hung N.N., Merz B., Apel H. (2014) – Large-scale suspended sediment transport and sediment deposition in the Mekong Delta. Hydrology and Earth System Sciences, 18, 3033-3053.
DOI : 10.5194/hess-18-3033-2014

Manh N.V., Dung N.V., Hung N.G., Kummu M., Merz B., Appel H. (2015) – Future sediment dynamics in the Mekong Delta floodplains: Impacts of hydropower development, climate change and sea level rise. Global and Planetary Change, 127, 22-23.
DOI : 10.1016/j.gloplacha.2015.01.001

Mekong River Commission (2010) – Annual Mekong Flood Report 2009. Mekong River Commission (MRC), Office of the Secretariat in Phnom Penh, Cambodia, 80 p.

Milliman J.D., Ren M. (1995) – River flux to the sea: Impact of human intervention on river systems and adjacent coastal areas. In Eisma D. (Ed.): Climate Change: impact on coastal habilitation, Lewis Publication, Boca Raton, 57-83.

Milliman J.D., Farnsworth K.L. (2011) – River Discharge to the Coastal Ocean. Cambridge University Press. 392 p.
DOI : 10.5670/oceanog.2011.108

Phan N.H., Hoang T.S. (1993) – Mangroves of Vietnam, IUCN, Bangkok, 173 p.

Phan L.K., van Thiel de Vries J.S.M., Stive M.J.F. (2015) – Coastal Mangrove Squeeze in the Mekong Delta. Journal of Coastal Research, 31, 2, 233-243.
DOI : 10.2112/JCOASTRES-D-14-00049.1

Schmitt K., Albers T. (2014) – Area coastal protection and the use of bamboo breakwaters in the Mekong Delta. In Nguyen D.T., Takagi H., Esteban M. (Eds), Coastal Disasters and Climate Change in Vietnam: Engineering and Planning Perspectives, Elsevier, 175-198.
DOI : 10.1016/B978-0-12-800007-6.00005-8

Syvitski J.P.M. (2008) - Deltas at risk. Sustainability Science.
DOI : 10.1007/s11625-008-0043-3

Ta T.K.O., Nguyen V.L., Tateishib M., Kobayashib I., Tanabeb S., Saitoc Y. (2002) – Holocene delta evolution and sediment discharge of the Mekong River, southern Vietnam. Quaternary Science Reviews, 21 (16-17), 1807-1819.
DOI : 10.1016/S0277-3791(02)00007-0

Tamura T., Saito Y., Bateman M.D., Nguyen V.L., Ta T.K.O., Matsumoto D. (2012) – Luminescence dating of beach ridges for characterizing multi-decadal to centennial deltaic shoreline changes during Late Holocene, Mekong River delta. Marine Geology, 326-328, 140-153.
DOI : 10.1016/j.margeo.2012.08.004

Thanh N.D. (2014) – Climate change in the coastal regions of Vietnam. In Thao N.D., Takagi H. Esteban (Eds): Coastal Disasters and Climate Change in Vietnam: Engineering and Planning Perspectives, Elsevier, 175-198.
DOI : 10.1016/B978-0-12-800007-6.00008-3

Thieler E.R., Himmelstoss E.A., Zichichi J.L., Ergul A. (2009) – Digital Shoreline Analysis System (DSAS) version 4.0 - An ArcGIS extension for calculating shoreline change. U.S. Geological Survey Open-File Report, 2008-1278. http://woodshole.er.usgs.gov/project-pages/dsas/

Unverricht D., Szczucińskib W., Statteggera K., Jagodzińskib R., Lec X.T., Kwongd L.L.W. (2013) – Modern sedimentation and morphology of the subaqueous Mekong Delta, Southern Vietnam. Global and Planetary Change, 110, 223-235.
DOI : 10.1016/j.gloplacha.2012.12.009

Viêt Nam News (2014) – Erosion threatens valuable coastal forest. 03/17/2014 [Online] URL:
http://vietnamnews.vn/environment/252405/erosion-threatens-valuable-coastal-forest.html.

Walling D.E. (2008) – The changing sediment load of the Mekong River. AMBIO: A Journal of the Human Environment, 37, 150-157.
DOI : 10.1579/0044-7447(2008)37[150:TCSLOT]2.0.CO;2

Wolanski E., Nhan N.H., Spagnol S. (1998) – Sediment dynamics during low flow conditions in the Mekong River estuary, Vietnam. Journal of Coastal Research, 14, 472-482.

Xue Z., Liu J.P., Ge Q. (2010) – Changes in hydrology and sediment delivery of the Mekong River in the last 50 years: connection to damming, monsoon, and ENSO. Earth Surface Processes and Landforms, 36, 296-308.
DOI : 10.1002/esp.2036

Xue Z., He R., Liu J.P., Warner J.C. (2012) – Modeling transport and deposition of the Mekong River sediment. Continental Shelf Research, 37, 66-78.
DOI : 10.1016/j.csr.2012.02.010

Yong M.L., Grundy-Warr C. (2012) – Tangled nets of discourse and turbines of development: Lower Mekong mainstream dam debates. Third World Quarterly, 33, 1037-1058.
DOI : 10.1080/01436597.2012.681501

Ziv G., Baran E., Nam S., Rodríguez-Iturbe I., Levin S.A. (2012) – Trading-off fish biodiversity, food security, and hydropower in the Mekong River Basin. Proceedings of the National Academy of Science. 109, 5609-5614.
DOI : 10.1073/pnas.1201423109

Haut de page

Annexe

Version française abrégée

Le delta du Fleuve Mékong, troisième plus grand du monde, présente un fonctionnement particulièrement complexe en raison de son étendue spatiale, de son mode de progradation, de sa morphologie et de la diversité des environnements sédimentaires et de l’empreinte croissante des activités humaines sur ce mégadelta depuis deux siècles. Le delta du Mékong a connu une progradation vaseuse massive vers l’ouest en Mer de Chine Méridionale et dans le Golfe de Thaïlande (Ta et al., 2002 ; Xue et al., 2010). Cette progradation vaseuse jouxte un secteur d’embouchures multiples dominé par du sable et caractérisé par un taux de progradation moindre (Tamura et al., 2012)

A partir d’analyses statistiques basées sur des images satellites Landsat couvrant la période 1973-2014, l’étude de la mobilité du linéaire côtier principalement vaseux, sur la façade du delta correspondant au secteur de la Mer de Chine Méridionale, au sud-ouest des embouchures principales, montre une érosion globale (fig. 2). L’analyse des images montre que près de 70 % des 160 km de longueur côtière de cette partie du delta se sont fortement érodés. Cette érosion affecte notamment la moitié sud de cette côte avec un contraste assez net par rapport à la moitié nord plus proche des embouchures et qui, elle, a progradé. Ta et al. (2002) avait montré que la moitié sud de cette côte a connu une forte progradation tard dans l’Holocène, à partir de 6000 BP. Bien que la phase de commencement de cette érosion reste inconnue, cette tendance constitue une inversion nette par rapport à la dynamique de progradation massive de cette partie vaseuse du delta. L’analyse a porté sur les variations du trait de côte, calculées à partir de la méthode du Digital Shoreline Analysis System (Thieler et al., 2014), sous ArcGIS®, et des pertes surfaciques par le biais de l’outil Union également sous ArcGIS®. Sur cette période d’analyse de 41 ans, le littoral de ce secteur du delta recule globalement jusqu’à une cinquantaine de mètres par an. Ces taux sont confirmés par une analyse à plus courte échelle temporelle (2003-2011/12) effectuée par Anthony et al. (2013) à partir d’images SPOT 5 (fig. 2) à plus haute résolution que les images LANDSAT.

Trois faits principaux peuvent expliquer cette érosion du littoral avec, pour chacun, une composante anthropique majeure. Le premier est relatif à une baisse avérée des apports sédimentaires depuis la région des embouchures au nord-est du delta mise en évidence par l’analyse d’images satellites MERIS (Loisel et al., 2014). Les extractions de granulats dans le lit mineur des chenaux principaux (Bravard et al., 2013 ; Brunier et al., 2014), et la rétention sédimentaire par les nombreux barrages aménagés et en construction (Kondolf et al., 2014 ; Manh et al., 2014) sont autant de facteurs qui expliquent aussi la baisse de l’alimentation en sédiments à la côte et qui se répercute par l’érosion de ce secteur du delta. De plus, ajoutée à l’effet d’éloignement progressif des embouchures et, de ce fait, des apports sédimentaires fluviaux associés, la géométrie du linéaire côtier semble plus sensible, du fait de son orientation proche à la normale pour une large part de sa longueur (fig. 1), vis-à-vis des houles énergiques du nord-est qui constituent l’agent de forçage hydrodynamique marin principal s’exerçant sur ce delta. Un troisième facteur majeur d’aggravation de la fragilisation littorale relève des perturbations dans la répartition du couvert végétal côtier, constitué principalement de mangrove. Depuis le XVIIIe siècle, les provinces situées sur le delta du Mékong ont connu une très forte expansion démographique, reflétant l’attrait socioéconomique exercé par les zones deltaïques du fait de la richesse remarquable de leurs ressources naturelles (Than, 2014). La population deltaïque atteint aujourd’hui 18 millions d’habitants, se répartissant préférentiellement sur la bande côtière. Cette évolution démographique s’est accompagnée de besoins en eau croissants et d’extractions massives d’eau dans la plaine deltaïque, autant d’activités humaines aggravant considérablement la tendance naturelle à la subsidence (Erban et al., 2014).

Les exploitations aquacoles, depuis le début des années 1990, et les activités forestières, massives et extensives, progressent spatialement au détriment de la végétation en place sur la frange littorale, composée principalement de Rhizophora Apiculata et Avicennia Alba. Ces dégradations (fig. 3), combinées à la constitution vaseuse du littoral particulièrement mobile sous l’effet des vagues du nord-est, favorisent une libération et une dispersion des sédiments depuis la côte vers le large, par conséquent une érosion littorale. Les pertes d’espaces de la végétation, essentiellement composée de mangrove, sont bien mises en évidence par les images satellites (fig. 5). La mince frange de mangroves en front de mer se trouve souvent trop dégradée pour pouvoir former une protection efficace contre les houles. Il est important de noter que l’érosion montre une grande variabilité spatiale, du nord-est au sud-ouest (fig. 2), qui semble refléter non seulement des différences d’orientation vis-à-vis de la houle évoquées ci-dessus, mais aussi l’empreinte anthropique qui se traduit également par la mise en place de digues de protection destinées à ralentir l’érosion. A cet égard, la moitié sud du littoral de la Mer de Chine Méridionale paraît particulièrement fragile. L’éloignement de cette moitié sud par rapport aux embouchures par lesquelles transitent les sédiments fluviatiles est un autre facteur aggravant. Ce secteur explique l’essentiel du bilan global d’érosion (fig. 2).

Les pressions humaines évoquées rendent d’autant plus vulnérable le littoral du delta du Mékong au sud-ouest des embouchures, déjà extrêmement sensible aux forçages naturels fluviomarins (fig. 7). Cette érosion côtière, combinée à la subsidence exacerbée du delta, induit un risque non négligeable de submersion marine. Les perspectives de développement de grands projets de barrage hydro-électriques constitueront un facteur qui aggravera la vulnérabilité de ce delta fortement anthropisé dans les années à venir.

Haut de page

Table des illustrations

Titre Fig. 1 – Study area. Fig. 1 – Secteur d'étude.
Légende A. Catchment area of the Mekong River and inset showing the six river basin countries. B. The Mekong River delta in Vietnam. The delta and part of its network of canals and dykes. C. The wave rose for the South China Sea (Wavewatch III data from National Center for Environmental Prediction (NCEP): http://polar.ncep.noaa.gov/​waves/​download.shtml). A. Bassin versant du Fleuve Mékong et une figure montrant les six pays du bassin versant. B. Le delta du Mékong au Vietnam. Le delta et une partie de son réseau de canaux et digues, C. Rose des houles pour la Mer de Chine Méridionale (données Wavewatch III du National Center for Environmental Prediction (NCEP): http://polar.ncep.noaa.gov/​waves/​download.shtml).
URL http://geomorphologie.revues.org/docannexe/image/11336/img-1.png
Fichier image/png, 1010k
Titre Tab. 1 – Technical characteristics of the 1973 and 2014 satellite images and the computed uncertainties. Tab. 1 – Caractéristiques techniques des images satellites de 1973 et 2014 ainsi que les marges d’erreur calculées.
URL http://geomorphologie.revues.org/docannexe/image/11336/img-2.png
Fichier image/png, 135k
URL http://geomorphologie.revues.org/docannexe/image/11336/img-3.png
Fichier image/png, 821 octets
URL http://geomorphologie.revues.org/docannexe/image/11336/img-4.png
Fichier image/png, 738 octets
URL http://geomorphologie.revues.org/docannexe/image/11336/img-5.png
Fichier image/png, 1,3k
URL http://geomorphologie.revues.org/docannexe/image/11336/img-6.png
Fichier image/png, 1,4k
Titre Fig. 2 – Graphs of coastal area (km²/yr) change rates (erosion, accretion) for the South China Sea coast of the Mekong River delta between 1973 and 2014 (this study) and between 2003 and 2011 (Anthony et al., 2013) analysed from Landsat and SPOT 5 satellite images respectively.Fig. 2Graphique des taux d’évolution de surfaces côtières (érosion, accrétion, en km²/an) pour le littoral de la Mer de Chine Méridionale du delta du Mékong entre 1973 et 2014 (cette étude) et sur la période 2003-2011 (Anthony et al., 2013) analysés respectivement à partir d’images satellites Landsat et SPOT 5.
URL http://geomorphologie.revues.org/docannexe/image/11336/img-7.png
Fichier image/png, 22k
Titre Fig. 3 – Evolution of the South China Sea coast of the Mekong River delta. Fig. 3 – Evolution du littoral de la mer de Chine méridionale du delta du Mékong.
Légende A. Graph of coastal area (km²/yr) change rates between 1973 and 2014 analysed from Landsat satellite images. B. Graph of shoreline change rates (m/yr) between 1973 and 2014 analysed from Landsat satellite images. C. Graph of coastal area (km²/yr) change rates between 2003 and 2011 analysed by SPOT 5 satellite images from Anthony et al. (2015). D. Map of shoreline changes between 1973 and 2014 analysed from Landsat satellite images. A. Graphique des taux d’évolution de surface côtière (km²/an) entre 1973 et 2014 analysés à partir d’images satellites Landsat. B. Graphique des taux d’évolution du trait de côte (m/an) entre 1973 et 2014 analysés à partir d’images satellites Landsat. C. Graphique des taux d’évolution de surface côtière (km²/an) entre 2003 et 2011 analysés à partir d’images satellites SPOT5 par Anthony et al. (2015). D. Carte des évolutions du trait de côte entre 1973 et 2014 analysées à partir d’images satellites Landsat.
URL http://geomorphologie.revues.org/docannexe/image/11336/img-8.png
Fichier image/png, 151k
Titre Fig. 4 – Photograph (2012) showing typical shoreline retreat in the southwestern part of the Mekong Delta, with waves causing scouring of the muddy foreshore. Fig. 4 – Photographie (2012) montrant un recul côtier représentatif de la partie est du Delta du Mékong, les vagues causant le creusement de l’avant-côte vaseuse.
URL http://geomorphologie.revues.org/docannexe/image/11336/img-9.jpg
Fichier image/jpeg, 864k
Titre Fig. 5 – Evolution of vegetation cover between 2003 and 2011. Fig. 5 – Evolution de la couverture végétale entre 2003 et 2011.
Légende A. SPOT 5 satellite image of 2003 in a coastal sector of Bac Lieu Province. B. SPOT 5 satellite image of 2011 in the same sector. C. An example showing a comparative analysis of losses and gains of vegetation in the coastal sector of Bac Lieu Province. These changes are mainly associated with the loss of fringe mangroves between 2003 and 2011 identified from two high-resolution SPOT 5 satellite images. A. Image satellite SPOT 5 de 2003 sur un secteur côtier de la province de Bac Lieu. B. Image satellite SPOT 5 de 2011 sur le même secteur côtier. C. Un exemple montrant une analyse comparative de surfaces perdues et gagnées de végétation sur le secteur côtier de la province de Bac Lieu. Les évolutions sont principalement associées à la perte de mangroves côtières entre 2003 et 2011 identifiées depuis deux images satellites SPOT 5 de haute résolution.
URL http://geomorphologie.revues.org/docannexe/image/11336/img-10.png
Fichier image/png, 2,4M
Titre Fig. 6 – Photograph (2012) showing residual mangrove stands and ongoing erosion along the South China Sea coast of the Mekong delta.Fig. 6 – Photographie (2012) montrant des mangroves résiduelles dans une zone en érosion du littoral de la Mer de Chine.
URL http://geomorphologie.revues.org/docannexe/image/11336/img-11.jpg
Fichier image/jpeg, 1,2M
Titre Fig. 7 – Population statistics for the three provinces along the eroding South China Sea coast of the Mekong delta south of the mouths. Fig. 7 – Densité de population pour les trois provinces le long the la côte de Mer de Chine Méridionale en érosion du Delta du Mékong au sud des embouchures.
URL http://geomorphologie.revues.org/docannexe/image/11336/img-12.png
Fichier image/png, 12k
Titre Fig. 8 – Photograph (2012) showing shrimp farms on the left, separated from mangroves to the right by an earth dyke. Fig. 8 – Photographie (2012) montrant des installations d’élevages de crevettes à gauche, séparées de la mangrove, à droite, par une digue en terre.
URL http://geomorphologie.revues.org/docannexe/image/11336/img-13.jpg
Fichier image/jpeg, 892k
Haut de page

Pour citer cet article

Référence papier

Manon Besset, Edward J. Anthony, Guillaume Brunier et Philippe Dussouillez, « Shoreline change of the Mekong River delta along the southern part of the South China Sea coast using satellite image analysis (1973-2014) », Géomorphologie : relief, processus, environnement, vol.22 - n° 2 | 2016, 137-146.

Référence électronique

Manon Besset, Edward J. Anthony, Guillaume Brunier et Philippe Dussouillez, « Shoreline change of the Mekong River delta along the southern part of the South China Sea coast using satellite image analysis (1973-2014) », Géomorphologie : relief, processus, environnement [En ligne], vol.22 - n° 2 | 2016, mis en ligne le 04 mai 2016, consulté le 29 avril 2017. URL : http://geomorphologie.revues.org/11336 ; DOI : 10.4000/geomorphologie.11336

Haut de page

Auteurs

Manon Besset

CEREGE UM 34 – Europôle de l’Arbois – Aix-Marseille Université – Aix-en-Provence, France (besset@cerege.fr). Tél : +33 4 42 97 17 91.

Edward J. Anthony

CEREGE UM 34 – Europôle de l’Arbois – Aix-Marseille Université, IUF – Aix-en-Provence, France (anthony@cerege.fr).

Articles du même auteur

Guillaume Brunier

CEREGE UM 34 – Europôle de l’Arbois – Aix-Marseille Université – Aix-en-Provence, France (gbrunier@cerege.fr).

Articles du même auteur

Philippe Dussouillez

CEREGE UM 34 – Europôle de l’Arbois – Aix-Marseille Université – Aix-en-Provence, France (dussouillez@cerege.fr).

Articles du même auteur

Haut de page

Droits d’auteur

© Groupe français de géomorphologie

Haut de page
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Groupe français de géomorphologie
  • Revues.org