Shoreline Change Analysis and Assessment

#### **Chapter 2**

## Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data and Digital Shoreline Analysis System

*Dibyendu Dutta,Tanumi Kumar, Chiranjivi Jayaram and Wasim Akram*

#### **Abstract**

Long-term (1973–2021) shoreline displacement, rate of change, and temporal pattern were examined using multi-date Landsat data and Digital Shoreline Analysis System (DSAS) along the 200 km coast of Hooghly estuary. Orthogonal transects of 100 m apart were casted for calculation of End Point Rate and Weighted Linear Regression rate on different temporal scales for seven analysis zones. The shoreline change pattern was established using Hierarchical agglomerative clustering. The study reveals that almost 43.45% of the beachfront has eroded and 56.55% has accreted during the past four decades. The average erosion rate varies between 0.01 and 13.71 m yr.<sup>1</sup> and accretion of 0.01 to 22.30 m yr.<sup>1</sup> . The littoral drift resulted in a maximum seaward aggression by 1096.89 m in the zone 1. Landward movement was maximum (602.96 m) in the zone 4. Although west bank is prograding @ 3.47 m yr.<sup>1</sup> (5.83), the east bank is eroding @ 1.30 m yr.<sup>1</sup> (4.08). Based on the cluster analysis about –1.87% of the shoreline exhibits consistent erosion over all the intervals, whereas trend was evident in 4.73% of the coastline. The portions of coastlines, which exhibit high erosion rate and consistent erosion need immediate attention and policy intervention.

**Keywords:** Hooghly estuary, shoreline change, erosion, accretion, Landsat, DSAS, change pattern

#### **1. Introduction**

The shoreline is the physical interface or intertidal margin between land and sea and constitutes one of the 27 global "Geo-indicators" referred by the International

Union of Geological Science [1] and International Geographic Data Committee (IGDC). Shoreline change is a dynamic natural process in the coastal areas induced by erosion/accretion that occurs over a range of temporal scales. The morphological evolution of the Hooghly estuary and its coastline is the result of two counteracting transport processes of sediment supply versus removal. When both the processes are balanced an equilibrium is reached. However, most often this balance is disturbed due to the influence of episodic and/or long-term natural forcing and anthropogenic interventions. As a consequence, the shoreline keeps changing its position [2–6] over a wide temporal scale, from geologic age to short-lived, extreme weather events such as storms and tsunamis. The long-term processes that shape the shoreline include sealevel rise (SLR), altered wind patterns [7], frequency and intensity of storms [7], offshore bathymetric changes [8], high energy swells [9] and supply of fluvial sediment input. In addition, anthropogenic activities *viz*., landcover changes in the river catchment, port and harbor and dam construction, dredging for maintaining navigation channels, aquaculture, protective embankments, beach nourishment, economic and tourist activities also exacerbate the coastline change on a short temporal scale. Engineering structures change the estuarine circulation patterns and may change the freshwater flow along with sediment and nutrient supply. In several instances, engineering modifications to the beach creates discontinuities in the historical shoreline position and mask underlying long-term behavior [10]. Another less reported phenomena are land subsidence which may occur naturally due to compaction of sediments or triggered by the excessive withdrawals of ground water. In general, the coastal landform establishes a morphodynamic equilibrium after episodic short-term perturbations. However, many times the combination of natural and manmade activities exacerbates the shoreline change and exhibits non-linear morphological responses to change [11].

According to Williams [12], the study of shoreline variation and forecast plays an important role in coastal zone management and it becomes more crucial in the context of anticipated climate change and sea-level rise [13]. In this context, one of the key requirements for effective coastal zone management is the availability of accurate position of the shorelines for analysis of changes in the past and future trends. Traditional methods of shoreline delineation include terrestrial surveys using landmarks, aerial photos [14, 15], Global Positioning Systems (GPS), terrestrial Light Detection and Ranging (LiDAR) or 3D scanners. But they are time-consuming, labour intensive and costly. In contrast the remote sensing data form space platform is more convenient, easy to process and above all freely available in the public domain. Remote sensing data has been extensively used in shoreline change studies because of their synoptic and repetitive coverage, multispectral capabilities enabling contrast between land and water in the infrared portion, and cost-effectiveness [14, 16]. Advanced image processing techniques can be employed on satellite data for precise extraction of the shoreline. Some of the methods used by different researchers include threshold level slicing and image classification technique [17], density slicing of TM band 5 [18], canny edge detection using DN threshold [19], mean shift segmentation [20], pixel-based segmentation using DN threshold [21], neural network [22], fuzzy logic [23, 24], texture analysis [25], machine learning [26] and incorporation of ancillary spatial data in the classification scheme [27–29]. Quantitative assessment of the spatio-temporal variation of shoreline at global scale has been carried out by several authors [30–32]. In this endeavor the twin technologies of Remote Sensing and Geographic Information System has been recognized as the most useful tools for quantifying the historic shoreline change [33, 34]. To avoid the discrepancy which might be introduced due to fluctuation of water level Yu et al. [35]

#### *Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

have used satellite images obtained at similar tidal heights. Chen and Chang [36] have done the tidal correction using high spatial resolution satellite images and real-time data of tidal level to reduce the impact of tidal level variability on the estimation of coastline change. In India also several studies have been carried out for shoreline change analysis using remote sensing data [37–39]. Most of the studies have used Digital Shoreline Analysis System [40], a software extension within the ArcGIS tool for measuring, quantifying, calculating and estimating of rate of change from multiple historic shoreline positions at different temporal scales [41–44]. The change metrics of DSAS are Net Shoreline Movement (NSM), Shoreline Change Envelope (SCE), End Point Rate (EPR), Linear Regression Rate (LRR) and Weighted Linear Regression Rate (WLR) among others. LRR and WLR enable multiple historic shorelines to be used to determine the rate of change by fitting a least-square regression line to all shoreline points for particular transects.

In the present study, Landsat satellite data of 8 temporal intervals between 1973 and 2021 were used for land-water discrimination, generation of shorelines and long-term change rate along with change pattern along the Hooghly estuary. The instantaneous land-water boundary was used as coastline which is relatively simple and can easily be identified using image transformation. The main objectives of the study are i) medium- and long-term changes in the shoreline at high spatial resolution using DSAS ii) to identify the erosion/accretion pattern and iii) to examine the role of change drivers.

The findings of the study will be useful for the managers and engineers to make scientific and rational policies for land use planning, to develop effective coastal protection strategies, predicting capacity for future coastal change due to climate and other drivers and improving impact and vulnerability assessments that include natural human sub-system interactions.

#### **2. Study area**

The Hooghly estuary is located in the southernmost part of Indo-Gangetic plain, flanked between East Midnapur (in the West) and South 24 Parganas district (in the East), extending between 21<sup>o</sup> 33<sup>0</sup> 1000N to 22<sup>o</sup> 13<sup>0</sup> 1600N latitude and 87<sup>o</sup> 45<sup>0</sup> 00<sup>00</sup> to 88<sup>o</sup> 18<sup>0</sup> 2200E longitude (**Figure 1**). The head Bay is a unique deltaic environment comprising a wide continental shelf, complex coastal geometry and high tidal range. Tide domination is indicated by exponentially tapering channels, with funnel-shaped mouths [45]. The region has formed, sculptured and modified due to continuous fluvial action of the Ganga and the Brahmaputra systems, intense tidal hydrodynamic behavior, climatic disturbances and anthropogenic activities [46]. The funnel-shaped estuary has a width of 6 km at its head and 25 km at the mouth, responsible for tidal asymmetry and flow variation leading to bank erosion [47]. The average depth of the water column is only 6 m [48]. The estuary receives 4 tributaries *viz*. Damodar and Rupnarayan rivers at its head, and Haldi and Rasulpur rivers at the middle at its west bank. In contrast, the east coast is punctuated by several closely spaced inlets. Based upon the tidal amplitude, the coastal region of West Bengal can be sub-divided into i) macro tidal (tidal range > 4 m) from Sagar to Bangladesh border and ii) meso-tidal (tidal range of 2–4 m) mostly Medinipur (Digha-Sankarpur-Junut) coastal plains to the west of the Hooghly estuary.

Geologically the basement of the Bengal basin is a part of the eastern edge of the Indian plate, which is being subducted beneath the China plate along the Sunda

**Figure 1.** *Index map of the Hooghly estuary.*

subduction zone and Naga-Lushai orogenic belt. The tectonic and depositional history of the Bengal basin has been controlled by several movements during Cretaceous-Tertiary periods. Due to the tectonic activity the Bengal basin has been tilted towards east resulting in successive changes in the course of the Ganga River towards east from the historical past. Due to this shifting, the deltaic region suffers from the paucity of fresh water discharge and sedimentation. Auto compaction of loosely attached sediments and gradual land subsidence is also another prominent geomorphic event occurring in this region [49–54] which mostly remains unnoticed. Morphometrically the Hooghly estuary is the product of continuous fluvial sedimentation in a series of para-deltaic lobe progradation systems developed on the western shelf margin areas and eastern troughs of the Bengal basin caused by the eustatic, isostatic and tectonic forces. The coastline presents various landforms such as tidal/mud flats, sandy beaches (located near Digha, Duttapur, Shyampur, Dadanpatra, Baguranjalpai, Dariapur and Nij Kasba), salt marshes (near Khejuri and at the mouth of Rasulpur river near Nij Kasba) and mangrove marsh (south of Patibunia). A vast extension of the muddy beach is found in South 24 Parganas, especially to the east of Bakkhali. The most striking feature is the development of successive rows of dunes (both Palaeo and Neo dunes) with intervening clayey tidal flats in the south of East Midnapur district between the stretches of Subarnarekha and Hooghly estuary is due to punctuations in the regression of the sea during Holocene [55]. Banerjee and Sen [56] opined that the

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

regression of sea along this coastal tract is around 6000-year BP which resulted in seaward shifting of shoreline and formation of Paleo-dunes. Accordingly, to Niyogi [57], six regular cycles of beach ridges alternating with a variable number of bars are visible in the area, which is indicative of the shifting of shorelines. According to Gaur and Vora [58], the shoreline position was 5–15 km inland from the present shoreline around 6000-year BP. The erosion and accretion patterns clearly show a continued geomorphic sculpturing of the Hooghly coast.

To capture the micro-level variability, alongshore is divided into 7 analysis zones (**Figure 2**) covering both the west and east bank. The zones in the west bank are delimited by the main inlets which are the freshwater sources, eventually draining into Bay of Bengal. The area delimitation of various zones, constituting transects and shoreline distances is given in **Table 1**. The west bank is divided into 3 zones whereas the east bank into 4 zones (**Table 1**). The total length of the coastline studied is 200 km of which 90 km on the western side and 110 km on the eastern side of the estuary. The studied coastline was divided into 1924 number of transects (Tn) separated by 100 m. The number of transects increases from west to east bank in the clockwise direction.

**Figure 2.** *Different analysis zones.*


**Table 1.**

*Salient description of different analysis zones.*

#### **3. Materials and methodology**

#### **3.1 Data used**

#### *3.1.1 Army and survey of India Topomaps*

The historic shorelines were digitized from Army Map Series (NSS&H, Edition-1, AMS) in 1:250,000 scale surveyed during 1942–1943) number NF-45: 7 (north of study area) and 11 (south of the study area) were used for the coastline change analysis. Besides Survey of India topomaps of 73 N-16, 73O -13,14; 79B - 4; 79C-1,2,6 surveyed during 1967 were also used for generation of high-water level (HWL) coastlines.

#### *3.1.2 Satellite data*

Landsat satellite data of 1973 to 2021 have been used for decadal and long-term trend analyses. The data has been selected based on clear sky condition, high tide date and time as well as season. For discrimination of land-water boundary shortwave infrared bands 5 (1.55–1.75 μm) and 7 (2.08–2.35 μm) of Landsat - 4, 5, 7 and bands 6 (1.566–1.651 μm) and 7 (2.107–2.294 μm) of Landsat – 8 (OLI) were used. The details of the satellite data used in the study are given in **Table 2**.

For the of satellite data tide and current prediction programme *viz*. WXTide32 was used which allows knowing the time of high-tide and low-tide, as well as the tide height (m). It is supported by more than 9500 stations worldwide with the capability to predict tides from 1970 through 2037. One of the nearest stations of the study area, e.g., Diamond Harbor was selected to know the high tide date, time and magnitude. The high tide timings were compared with the satellite overpass dates and time for the selection of images representative of coastlines on high-tide date. This enabled comparison of the shorelines under identical tide conditions by minimizing the variability due to the tidal cycle.

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*


**Table 2.** *Details of the satellite data used.*

#### **3.2 Methodology**

#### *3.2.1 Land-water discrimination and shoreline extraction*

There are seven types of coastline indicators *viz*. geomorphological reference lines, vegetation limits, instant tidal levels and wetting limits, tidal data, beach contours and storm lines. In the present study, the high-water levels (HWL) during spring tide have been considered for historical coastline change analysis. Several water indices have been used by many authors to extract the coastlines [59, 60]. Two of the most popular water indices are the Normalized Difference Water Index (NDWI) and Modified Normalized Difference Water Index (MNDWI). NDWI was primarily developed for Landsat MSS whereas MNDWI was developed for TM, ETM+ and OLI sensors [61]. The bands used to generate these indices mostly consist of green, near-infrared, middle-infra-red and shortwave infrared bands. The objective of all these indices is to enhance the contrast between land-water interfaces. Band 2 (0.52–0.60 μm) and band 5 (1.55–1.75 μm) of Landsat-7 and Band 3 (053–0.59 μm) and band 6 (1.57–1.65 μm) of Landsat-8 OLI were used for computing NDWI. The formulae of NDWI and MNDWI are given below:

$$\text{NDWI}\_{\text{MSS}} = \left(\mathbf{R}\_{\text{green}} - \mathbf{R}\_{\text{nir}}\right) / \left(\mathbf{R}\_{\text{green}} + \mathbf{R}\_{\text{nir}}\right) = \left(\mathbf{B}\_1 - \mathbf{B}\_4\right) / \left(\mathbf{B}\_1 + \mathbf{B}\_4\right) \tag{1}$$

$$\mathbf{M} \mathbf{N} \mathbf{D} \mathbf{W} \mathbf{I}\_{\text{TM}, \text{ETM}+} = \left( \mathbf{R}\_{\text{green}} - \mathbf{R}\_{\text{swir}} \right) / \left( \mathbf{R}\_{\text{green}} + \mathbf{R}\_{\text{swir}} \right) = \left( \mathbf{B}\_2 - \mathbf{B}\_5 \right) / \left( \mathbf{B}\_2 + \mathbf{B}\_5 \right) \tag{2}$$

$$\mathbf{M} \mathbf{N} \mathbf{D} \mathbf{W} \mathbf{I}\_{\text{OLI}} = \left( \mathbf{R}\_{\text{green}} - \mathbf{R}\_{\text{swir}} \right) / \left( \mathbf{R}\_{\text{green}} + \mathbf{R}\_{\text{swir}} \right) = \left( \mathbf{B}\_3 - \mathbf{B}\_6 \right) / \left( \mathbf{B}\_3 + \mathbf{B}\_6 \right) \tag{3}$$

Where, Rgreen = spectral reflectance of the green band, Rnir = spectral reflectance of near-infrared band and Rswir = spectral reflectance of the shortwave infrared band.

Before applying the water index on Landsat MSS data of 1973, the image was resampled to 30 m spatial resolution to make the resolution comparable with the rest of the datasets. A Boolean approach was used on the NDWI/MNDWI images to create two classes *viz*. land and water. The threshold for land-water boundary was kept >0.15 for OLI and > 0.1 for TM/ETM+. The resulting image had only two classes *viz*. land and water. However, some of the inter-tidal zones could not be demarcated due to very high sediment loading. Hence, a hybrid approach was followed wherein NDWI/ MNDWI, SWIR and topomaps were used for precise demarcation of the shorelines. Vectorization of the land-water boundary was done using the region growing tool of ERDAS/Imagine (ver. 9.1). The spectral Euclidean distance was set interactively to accurately capture the land water boundary. Some manual editing was also done on the shoreline vector.

#### *3.2.2 Shoreline change analyses*

Historical shoreline behavior was examined using Digital Shoreline Analysis System (DSAS, ver. 5.0), an extension tool of ArcGIS software (developed by the US Geological Survey) which calculates several change statistics *viz*. Net Shoreline Movement (NSM), Shoreline Change Envelope (SCE), End Point Rate (EPR), Linear Regression Rate (LRR) and Weighted Linear Regression Rate (WLR). It can analyze the time series of multiple shoreline positions [34] by using linear regression fit. Provision is there to include uncertainty of the input data in terms of assigned weights. Functionally DSAS performs 3 major activities *viz*., i) defining a baseline ii) generation of orthogonal transects and iii) computation of rates of changes. The tool enables the calculation of scales and rates of change statistics from multiple historic shoreline positions and sources. DSAS is a freely downloadable tool and is available at Woodshole [62] http://woodshole.er.usgs.gov/project-pages/dsas/. The details are available in Thieler and Danforth [63, 64] and Thieler et al. [34]. A brief description of the change statistics used in the study is given below.

*Shoreline Change Envelope (SCE):* A measure of the overall change in shoreline (m) at each transect considering the farthest and nearest position of the shoreline for the baseline location [65] irrespective of the dates.

*Net Shoreline Movement (NSM):* It is the distance (m) between the oldest and the youngest shorelines [66].

*End Point Rate (EPR):* It is derived by dividing the distance of shoreline change between two time periods by the time interval and expressed as m yr.<sup>1</sup> [65–69]. This method provides the net rate of change over the long term. It has both advantages and disadvantages; the advantage is that only two shorelines are required for computation of change rate but unable to use more than two date shoreline data.

*Linear Regression Rate (LRR):* It determines a rate-of-change statistic by fitting a least square regression line using all the intersection points all shorelines and individual transect [66, 68, 70]. The slope of the line is the rate of shoreline change. The advantages of linear regression include i) all the time-series data are used and ii) can reduce the impact of spurious values on the overall accuracy of change rate [71]. To ensure meaningful results from the regression model the temporal intervals of shorelines were kept well distributed over the analysis period.

*Weighted Linear Regression Rate (WLR):* This method takes into account the positional and measurement uncertainty of the shoreline positions [63, 72–74] especially when the shorelines were digitized from various sources and scales. The uncertainty values are incorporated in the DSAS as weights. The slope of this regression line is the shoreline change rate (m yr.<sup>1</sup> ). Using the weighing factor, WLR, standard error of the estimate (WSE), standard error of slope with user selected confidence interval (WCI) and R-squared value (WR2) are obtained [34]. The results of this method are controlled by the points with smaller positional uncertainty on the best-fit regression

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

line [65]. If no values are provided by the user, DSAS uses the default uncertainty value. The weight (w) is defined as a function of the variance in the uncertainty of the measurement (e) [34]:

$$\mathbf{w} = \mathbf{1}/(\mathbf{e}^2) \tag{4}$$

Where e = shoreline uncertainty value.

*Coefficient of Determination (R2 ):* It is the percentage of variance in the data that is explained by a regression [65]. It is used to verify the quality of the best-fit line regression.

#### *3.2.3 Calculation of data uncertainty*

The errors or uncertainties that arise due to different data sources, time of data acquisition, and the type of shoreline indicator were quantified based on several studies [73, 75]. According to Fletcher et al. [75] and Romine and Fletcher [76] there are two types of uncertainty: positional (seasonal and tidal fluctuations) and measurement (digitizing, pixel and rectification error). The uncertainty for each dataset was worked out considering the data product with due weightage of the quality of each data. The total uncertainty is used to calculate the weight and further working in the DSAS. Different uncertainties are explained below.

*Seasonal error (Es):* It is the error that arises due to seasonal changes of shoreline positions under the action of the waves and storms [75]. In the present study, all the scenes are of the winter season and hence this error was neglected.

*Tidal fluctuation error (Et):* It is the error associated with horizontal variability in shoreline position due to tides [75]. All the images in the present study correspond to high-tide values. Based upon the values the tidal range was considered as 2.12 m.

*Digitizing error (Ed)*: It is the error related to shoreline digitization [75]. The digitizing error was kept within half a pixel of Landsat data (15 m).

*Pixel error (Ep):* It relates to image precision (resolution). In the present study except for the Landsat MSS data all the TM, ETM+ and OLI data have a spatial resolution of 30 m. To make the pixel uniform the 60 m MSS pixels were resampled to 30 m. Thus, the average pixel error was neglected.

*Rectification error (Er)*: It is the square root of the mean error of the image rectification process [75, 76]. The rectification error in the level-2 Landsat products was found to be one fourth of a pixel, i.e., in the present case was decided as 7.5 m.

*Total Positional Uncertainty*: The total positional uncertainty (Ut) is the result of all errors that were previously estimated. It is defined as the square root of the sum of the squares of the sources of different errors [75, 76]. The formula of Ut is given as follows:

$$U\_t = \pm \sqrt{\left(E\_s^2 + E\_t^2 + E\_d^2 + E\_p^2 + E\_r^2\right)}\tag{5}$$

Where Es is the seasonal error, Et = tidal error, Ed = digitizing error, Ep = pixel error, and Er = rectification error. The annualized uncertainty (Ua) was calculated using the square root of the sum of the squares of total positional uncertainty for each shoreline divided by the analysis period [75] as is given below.

$$U\_a = \pm \frac{\sqrt{\sum\_1^n U\_{ii}^2}}{T} \tag{6}$$


#### **Table 3.**

*Uncertainties associated with shorelines obtained from different sources.*

Various uncertainties in the historical shoreline position between 1948 and 2021 is given in **Table 3**.

The weight (w) is defined as a function of the variance in the uncertainty of the measurement (e). Weighted Linear Regression Rate (WLR) was computed using the total positional uncertainty values.

#### *3.2.4 DSAS configuration*

It consists of four main steps as is given below.


*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

iv. *Calculation of change in shoreline position and other statistics*: Two types of statistics are generated by DSAS *viz*. distance (NSM) and rate (EPR, LRR, and WLR). The regional change rate is calculated by averaging the rates of changes from all the transects. The average coefficient of determination (R2 ) and uncertainties of the annual rate-of-change (m yr.�<sup>1</sup> ) are computed at a 95% confidence interval (LCI95 or WCI95).

#### *3.2.5 Prediction of tide condition*

The date, time and height of tide were calculated using WXTide32 package. The height of tide is governed by the following harmonic equation given in the Manual of Harmonic Analysis and Prediction of Tides, special publication no. 98, US Department of commerce [77].

$$h = H\_0 + \sum\_{n}^{N} f\_n H\_n \cos\left[a\_n t + (V\_0 + u)\_n - K\_{n'}\right] \tag{7}$$

Where, h is the height of tide at any time t.

H0 = the mean height of water level above datum used for prediction.

Hn = the mean amplitude of any constituent An.

fn = the factor for reducing mean amplitude to year of prediction.

an = the hourly speed of constituent An.

t = the time, in hours, reckoned from beginning of year or prediction.

(V0 + u)n = the Greenwich equilibrium argument of constituent An when t = 0.

Kn' = the modified epoch of constituent An.

N = the number or constituents used for the particular station.

In this equation except *h* and *t*, all other parameters are considered as the harmonic constant for any particular year and the place. Using these Harmonic constant, the successive value of tide height can be generated at any point of time. WXTide32 data pertaining to Diamond Harbor was considered to be representative of the present study. **Table 2** provides the satellite overpass time versus the low and high tide timings.

#### *3.2.6 Cluster analysis*

Cluster analysis is a technique used to classify cases into groups that are relatively homogeneous within themselves and heterogeneous between each other, based on a defined set of variables [78, 79]. Hierarchical agglomerative clustering using the Ward linkage method was followed in the present study. In this method, clusters are merged to reduce the variability within the cluster. At every stage the average similarity of the cluster is measured. A case is selected to enter the cluster if the inclusion in the cluster produces the least increase in the error. The number of the cluster centres was determined from 'Scree diagram' in which 'distance coefficients' are plotted against the 'stages'. The point at which there is a significant jump in the distance values was considered as the 'elbow' of the 'Scree plot'. The numbers of clusters were decided as the number of cases minus the step of the elbow. Once the clustering is done, K-mean classification is performed for all the transects using the number of cluster centres from 'Scree plot'. K-mean classification assign cluster membership and distance from the cluster centre to each case. Distance of the cluster centres are determined by using Euclidean distance as is given below:

$$d\_{\vec{\imath}} = \left[ \sum\_{l=1}^{q} \left( \varkappa\_{l\bar{l}} - \varkappa\_{\bar{l}l} \right)^{2} \right]^{\frac{1}{2}} \tag{8}$$

Where *dij* = ED for two individuals *i* and *j,* each measured on q variables, *Xij, Xji, i = 1, … q.* ED (dij) is calculated as the sum of squared differences between relative crossshore positions at each transect (T1, T2, T3, etc.) during each epoch (a) and for all epochs (N). Smaller ED values indicate the cases are more similar. To evaluate the robustness of the clusters Kruskal-Wallis one-way ANOVA test [80] was carried out. A detailed description of cluster analysis can be found in Everitt et al. [81] and Hennig et al. [78].

#### **4. Results and discussion**

#### **4.1 Shoreline configuration**

The 200 km stretch of the study region has varied beach types including wide sandy beaches to mudflat, the mixture of sand and mud, mangrove wetlands as well as open mixed jungle at the backdrop of sandy/muddy beaches. The considerable length of the shorelines has embankments (**Table 4**). The western bank consists mainly of sandy and muddy beaches whereas the east bank predominantly consists of a muddy and mangrove systems with intermittent gap areas where the beach is absent. Zone-wise brief description of the beach configuration is given below.

*Zone-1*: This zone constitutes the western bank of Hooghly estuary and falls between the outlet of Pichhaboni Khal and Rashulpur river, confined between the transects T25 and T211. The foreshore is mostly sandy. Several sandy beaches (near Haripur, Boral, Bankiput and Digene) and the large number of aquaculture ponds are present in this zone. A large difference between high and low water lines (maximum of about 850 m) is observed between Pichhaboni Khal to the south of Gopalpur based on Survey of India topomaps. From satellite imagery, significant accretion and expansion of forest land can be seen between the Pichhaboni outlet and the north-east of Junput.

*Zone-2:* This zone is represented by the transect number T218 to T524 and falls between the confluence of the Rashulpur river and Haldi river on the west bank of Hooghly. The foreshore is mostly muddy with small sandy beaches near Hijli and Khejuri. Thick forest vegetation is observed in the east of Thanaberia along the coast. Large numbers of aquaculture ponds are present between Kandlamari and Khejuri. There is no embankment present in the northern half of this zone. The difference


#### **Table 4.**

*Mean shoreline change (m) over different time intervals.*

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

between high and low water lines is comparatively less ( 80 m) between Talpati Khal and Haldi river.

*Zone-3:* This zone extends from the Haldi river to the Rupnarayan river confluence, demarcated by the transects T534 to T917. This zone is highly convex towards the bay and the foreshore is muddy. The difference between high and low water line increases from Gilabaria (T600) till Jhikarkhali (T685) which further increases near Horkhali. Series of brick kilns and few aquaculture ponds can be seen in this zone. There is no protective embankment present is this zone.

*Zone-4:* This zone falls in the east bank of Hooghly estuary between the transect T919 and T1174. Very narrow muddy foreshore in present here, however, in some places it is absent. Most of the river bank of this zone is embanked especially upto Kantabaria, however, from here till Kulpikata Khal there is no embankment. The narrow difference exists between high and low water lines except near Diamond Harbor. Series of brick kilns can be observed in the north (Simulbaria to Diamond Harbor) and south (Kantabaria to Kulpi) of this zone.

*Zone-5:* The zone covers the coastline between Kulpi and Kakdwip, defined by the transects T1175-T1454. The difference between high and low water lines is very small and sometimes absent. Almost the entire coastline has an embankment. Few brick kilns and aquaculture ponds are present in this zone.

*Zone-6:* This zone falls between Kakwip and Namkhana (T1455-T1574), totally embanked and shortest among all. The difference between high and low water lines is very small especially at south of Kakdwip and near Nadabhanga Khal. No brick kiln or aquaculture pond is present here.

*Zone-7*: This is the last and longest zone between Namkhana and Henry Island in the east bank of Hooghly, represented by the transect T1575-T1967. The foreshore consists mostly of a muddy area (Dakshin Durgapur) but a narrow sandy beach is present in the west of Namkhana and a wide sandy beach in the extreme south near Lakshmipur till the outlet of the Bakkhali river. In the south of Patibunia considerable area is under a mangrove swamp and open mixed jungle. Moderate to good forest exists in the extreme south of the region confined between T1912-T1967. Near Dakshin Durgapur, the difference between high and low water line is more (300 m). Most of the coastline is embanked. No brick kilns and aquaculture ponds are present in this zone.

#### **4.2 Spatiotemporal change in shoreline**

The large difference in the shoreline position was observed within each time interval and among different intervals. The dynamics of the shoreline are mainly due to disequilibrium in the morphological state and northward tapering nature of the estuary coupled with plausible subsidence due to auto-compaction of the Holocene sediments. One commonality among all the time intervals is the large variation in the seaward end of both the banks (**Figure 3**). During 1995–2000 and 2005–2010, the overall variation in the shoreline position is minimum. In comparison to the east bank west bank has more variation except for 1973–1980. Considering all the temporal intervals between 1973 and 2021 average recession is maximum in 1995–2000 ( 43 m 110.99) especially due to erosion in the southern part of the east bank of the estuary. In contrast, there is an increasing trend in the seaward extension of the shoreline since 2010. Between 2017 and 2021 the average accretion is 70.67 m ( 162.57). The maximum accretion length was 1386 m at T1486 (south-west of Kalinagar) in 1973–1980 whereas maximum erosion was 1062 m at T1865 (west of Fraserganj) during 1995–2000 (**Table 4**).

**Figure 3.**

*Shoreline changes recorded at different transects over different temporal intervals.*

The percentages of transects recorded aggradation or recession is given in **Figure 4**. From the figure, it is apparent that the proportion of aggradation and erosion does not match over the time intervals. The percentage of the transects exhibiting erosion was comparable during 2000–2005 (29%), 2010–2017 (30.20%) and 2017–2021 (27.81%). There was an abrupt increase in the erosion by 69.91% in 2010–2017. In general, there is a decreasing trend of erosion, especially after 2000 (**Figure 4**).

**Figure 5** depicts how each zone contributes to the total shoreline change. Between 1973 and 2021, zone 5 contributed maximum towards erosion. Other zones that contributed marginally to erosion include zone 7 and zone 6. Zone 6 showed consistent erosion in all the intervals except for 1973–1980. Very high annualized aggradation of 69.17 m and 29.93 m was recorded in zone 1 and 2 respectively over the entire period of 1973–2021.

It is interesting to note that while comparing the coastline of 2021 with respect to 1948 (not used in the DSAS), there is a significant recession (900 m) in the zone 2 (between Talpati Khal and Kaldalmari) and in the zone 3 (near Horkhali) by about 600 m. In the east bank, most significant erosion is noticeable in zone 5,

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

**Figure 4.** *Percentage of transcets showing erosion at different time intervals.*

**Figure 5.** *Contribution of each zone towards erosion / accretion at different time intervals (Z represents the zones).*

between Jadabnagar and Tilakmandal chak. The maximum landward retreat recorded was 2700 m near Uttar Chandannagar. On the other hand, accretion was observed in the south of zone 1 and 2 as well as in the north of zone 6. Quantitative analysis of the coastline change in this region has been carried out by Bandyopadhyay et al. [82], Raju et al. [83], Jana et al. [84], Rudra [85], Chakraborty [49] and Das et al. [86] along with their underlying mechanism. They have opined that beach erosion is attributed to various causes such as decrease of sediment supply from rivers, land subsidence, and interruption of longshore sediment transport by man-made structures. As the sea level rises, it causes waves to act on higher parts of the beach profile, resulting in enhanced erosion. If the sandy beaches disappear as a result sea-level rise, waves and storm surges, it will impact higher areas along the coastline [87].

Jana and Bhattacharya [88] used multi-resolution Landsat satellite imagery of 1972–2010 for shoreline change study along the 65 km long coastal stretch located between Rashulpur (Purba Medinipur) and Subarnarekha (Balasore) estuarine complex. The authors revealed that about 23 km of coastline recorded accretion, which was observed on several beaches such as at Talsari, Udaipur and Haripur, which were not affected by anthropogenic activities.

#### **4.3 Shoreline change rate**

The shoreline change rates were computed by linear regression and end point rate method at a lateral spatial interval of 100 m along the coast. The rates of changes of shoreline at different transect points estimated by EPR and LRR methods are given in **Figure 6**. Large variation in net shoreline movement and change rates were observed in the study region among various analysis zones (**Table 5**). Considering long term change between 1973 and 2021 four of the zones *viz*., 1, 2, 3 and 4 showed positive change (aggradation) by WRR method, the range of which varies between 0.24 m yr.<sup>1</sup> (zone 3) to as high as 9.45 m yr.<sup>1</sup> (zone 1). The very high recession was found in east bank at zone 6 ( 4.35 m yr.<sup>1</sup> ), followed by zone 5 (3.02 m yr.<sup>1</sup> ).

**Figure 6.** *The rate of change of shoreline by WRR and EPR method.*

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*


#### **Table 5.**

*Zone-wise average shoreline change envelope (SCE), net shoreline movement and change rate by EPR and WRR method.*

Overall, the rate of aggradation superseded the rate of erosion in the 48 years span. The zones which experienced maximum net seaward movement include zone 1 (553.34 m) whereas maximum net landward movement (erosion) of shoreline was found in zone 5 (137.22 m) (**Table 5**). There is a good agreement between both the methods (EPR and WRR) in respect of zone 2, 3, 4, 5 and 7. In zone 1 large difference in the shoreline change rate calculated by both the methods was recorded between Junput and Jagannathpur (T102 to T168), whereas in zone 6, the differences are significant in the region south of Kakdwip to Budhakhali along the Kakdwip river (T1455 to T1511).

Although, the net shoreline movement (NSM) values are less in zones 3, 6 and 7 but the shoreline change envelope records large variation which indicates that the inter-annual fluctuation is very high in these zones and morphodynamic processes are very active.

Based upon the rate of erosion/accretion by WRR method, the transects were grouped into 7 classes (**Table 6**). From the table, it is evident that most of the shoreline (more than 73.33% by WRR and 69.95% by EPR) exhibit erosion/accretion rate between 5 and + 5 m yr.<sup>1</sup> . Low erosion rate (< 1.0 m/yr) was exhibited by 13.46% and 11.43% of the shoreline in WRR and EPR method respectively (not presented in the table). The proportion of very high erosion (<sup>&</sup>lt;10 m yr.<sup>1</sup> ) and aggradation (>20 m yr.<sup>1</sup> ) is limited to less than 2% of the shoreline. The spatial


#### **Table 6.**

*Different classes of erosion/accretion rates and their contribution to the shoreline.*

distribution of different change classes by WRR method is given in **Figure 7a**. It can be seen from the figure that in the west bank only one segment exhibits high erosion (10 to 5 m yr.<sup>1</sup> ) whereas in the east bank at least 6 segments (east of Kharibaria) show high erosion. This area exhibits has a large difference between low and high tide lines. While comparing with the Army Series map of 1948, it was found that there is a significant landward movement of shoreline between 1948 and 1973. In the east bank, there is no area under high erosion in zone 4, however, in zone 5, 6, and 7 considerable area along the shoreline is under high to very high erosion state. There are 3 distinct stretches near Uttar Chandannagar, Ramganunagar, Madhusudanpur and Lakshimipur. Close observation with the Army toposheet of 1948 reveals that there is an extensive recession in this area. The Rangatala island which used to be an integral part of the east bank has almost reduced to half between Kulpi and Madhusudanpur. The southern half of zone 6 has a high to very high rate of erosion between Budhakhali and north or Namkhana. The zone 7 is punctuated by two major areas of high erosion i) in the west of Edward creek, dominated by mangrove swamp and open mixed jungle and ii) in the east of Henrys island. In contrast to erosion, high to very high aggradation (> 20 m yr.<sup>1</sup> ) is recorded between south of Gopalpur to Junput dominated by a wide sandy beach and inter-tidal difference. High aggradation is also observed in the south of the Rashulpur river confluence. In zone 2 high rates of accretion is observed in the north of Rashulpur river and east of Nij Kasba. In the east bank, there is no area of high accretion except in zone 7, near Lakshmipur dominated by mangrove swamps. This observation is in good agreement while comparing with the Army topo map of 1948.

#### **4.4 Temporal pattern of erosion/accretion**

To understand the temporal pattern of change direction, transects were grouped into two categories *viz*., eroding and aggrading type based upon displacement direction in each time interval. From 8 different temporal intervals 256 unique combinations were generated which were further grouped into 8 categories *viz*. i) consistent erosion (when in all the 8 temporal intervals the changes are negative) ii) mostly erosion (when erosion is recorded at least in 5-time intervals) iii) recent erosion (when last 3 or more consecutive intervals erosion is dominant) iv) mostly accretion (when accretion is recorded at least in 5-time intervals) v) recent accretion (when last 3 or more consecutive intervals accretion is dominant) vi) alternate (when erosion and accretion takes place alternatively) vii) trend reversal (changes from erosion to accretion over the years in a consecutive manner) and viii) others (when no definite trend is observed). **Table 7** provides change patterns and their contribution to the entire shoreline under study. At a long-temporal scale, 1.87% of the shoreline shows consistent erosion which is alarming and another 36.69% are mostly eroded. We could not find any transect recording accretion in all the time intervals. All the 3 accretion types (MA, RA and TREA) together account for 38.93% of the shorelines (**Table 7**). Reversal of trend towards aggradation during last 3 or more consecutive time intervals was found from 4.73% of the shoreline, mostly located in zone 1 and 2. Only a small proportion of the shoreline (0.42%) exhibits alternate erosion and accretion over the years and does not yield a definite trend. **Figure 7b** shows the changing pattern along the shorelines.

Some of the transects that recorded both high erosion rate (more than 5 m yr.<sup>1</sup> ) and consistent erosion are located in the north of Sibkalinagar (T1372-T1374), south of *Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*


#### **Table 7.**

*Temporal change pattern of shoreline behavior and their contribution.*

**Figure 7.** *Shoreline changes a) rate of erosion/accretion (m yr.<sup>1</sup> ) and b) change pattern.*

Budhakhali near Ghiya Khal (T1520-T1533), south of Nadabhanga Khal (T1552-T1557) and north of Duaragra Gang in zone 6 (T1569-T1574).

#### **4.5 Hierarchical agglomerative clustering**

Although, shoreline change analysis quantifies rates and directions of change, further analyses are needed to resolve distinct modes of coastal system behavior. Traditional shoreline changes analyses quantify the rate and direction of change by analyzing multi-date/historical data. However, there are some commonalities in terms of coastal system behavior. The Hierarchical agglomerative clustering was performed using the change matrix of all 8 temporal intervals to define the distinct coastal change behavior. Clustering was done using the Ward method which computes the sum of squared distance within the clusters and aggregates the clusters with the minimum increase in the overall sum of squares. The distance coefficients were plotted against the stage to generate a 'Scree diagram' (**Figure 8**). The number of clusters in the present study was 5 which was used for K-mean clustering. The cluster centres and the distances between cluster centres are given in **Tables 8** and **9** respectively.

The clusters captured a unique pattern of change at a temporal scale (**Table 8**). Among all the transects, 79.15% are represented by cluster 1 and only 0.94% by cluster 5. In clusters 1, 2 and 3 most of the transects show a balancing act of aggradation and erosion at different temporal intervals. The transects that recorded consistent erosion (**Figure 7**) were found in cluster 1 only. In cluster 4, erosion is dominant, while in cluster 5 accretion is dominant in most of the time span. The mean displacement of the

**Figure 8.** *Scree diagram defining the optimum number of clusters using elbow rule.*


#### **Table 8.**

*Various cluster centres.*

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*


#### **Table 9.**

*Distance between cluster centres.*

shoreline in cluster 1 is 3.25 m and the maximum is 99.25 m in cluster 3, constituting only 4.80% of the total transects. All the clusters show aggradation in terms of their mean displacement values except cluster 1.

#### **5. Discussion**

Beach profile morphology and coastline, change over a range of time and spatial scales. The short-term variability occurs over a period of days to a month as a result of i) episodic events (storms) ii) medium-term variability over several months (e.g., winter summer wave change) to several years (e.g., due to regional climate variability, engineering intervention and prevailing sedimentary processes) and iii) long term variability that occurs over a period of a decade to a century, associated mainly with climate change impact; and very long term millennial-scale evolution as a result of quaternary sea-level changes [89]. Broad-scale analysis of changes in shoreline position has the potential to highlight the role of regional forcing on large-scale coastal behavior, e.g., long-term tidal cycles [90] or sea level rise [4]. Shoreline change analysis is also useful to identify notable 'hotspots' of contrasting behavior [91, 92]. The Hooghly estuarine shoreline analyses studied here comprehend synthesis of historical shoreline change over 48 years supported by limited ground observations. The data has been analyzed at high spatial resolution (100 m, alongshore interval) along the entirety of a 200 km shoreline. In the area evidence for strong met-ocean forcing is ostensibly compelling. The phenomena of erosion and accretion are largely regulated by littoral current patterns and sediment influx from different rivers and the adjacent Bay of Bengal. The shoreline of this 200 km stretch has different configurations from the sandy beach to muddy swamp punctuated by anthropogenic footprints including brick kilns, aquaculture ponds, protective embankments and beach nourishment treatments. Beach nourishment projects and coastline protection structures can result in an artificial accretion of coastline in a short period [93]. Large variations exist in shoreline position within the same year and also among different years indicate the disequilibrium in the morphological state. There could be several external factors responsible for shoreline change including sea level rise, changes in the wave climatology and storm intensity as well as changes in the catchment characteristics due to deforestation and land degradation which results in higher sediment load in the terrestrial run-off. In contrast to surface runoff, engineering intervention through the construction of dams and barrages also makes the estuary sediment starved.

In long-term perspective, temporal data of PSMSL (Diamond Harbor and Haldia) reveals that the sea level is rising at the rate of 2.41 and 3.02 mm yr.<sup>1</sup> respectively. The sea surface temperature induced El Niño-Southern Oscillation (ENSO) has a significant role in global atmospheric circulation influencing the temperature and precipitation. The irregular pattern of El Nino and La Nina triggers rainfall variability over the Indian sub-continent. In recent years strong La Nina and very strong El Nino have been witnessed in 2010–2011 and 2015–2015 respectively. The monsoon rainfall variability has a direct relation with terrestrial run-off and estuarine water level. Since 1951 there were 8 strong to very strong El Nino and 7 strong types of La Nina years. The storm surges are another strong forcing factor in a short temporal scale that can change the shoreline configuration. Although, the frequency of cyclonic storms is declining over the Bay of Bengal but the intensity is increasing. Extremely severe cyclonic storms of 2019 and 2020 are the best examples causing extensive damage to the coastline embankments. Karunarathna et al. [89] found single storms or storm clusters predominantly change the supra tidal and inter-tidal part of the beach profile and that beach erosion volumes are strongly correlated to the power of the storm. Once the astronomical tides coincide with storms, extreme sea level occurs resulting in large-scale inundation and damage to the coastal structures. Besides warming of sea surface relative, sea level change can also happen due to vertical land motion that can result from glacial isostatic adjustment, tectonic processes, coastal subsidence and uplift caused by anthropogenic factors. High-frequency and short temporal scale sea level variability due to seiches, meteotsunamis are frequently under-represented in sea level studies and yet contribute to the extreme sea levels which are of great research interest and importance to coastal dwellers [94]. In general, coastal landforms affected due to short-term perturbations *viz*. cyclone generally attains a morphodynamic equilibrium often by adopting different 'states' in response to varying wave energy and sediment supply [95]. Nevertheless, elevated surge water levels are known to be important drivers of longer-term dynamics on sedimentary shorelines [96]. Besides sea level rise and storms alongshore sediment transport can also have an impact on the coastline change, in particular, it is likely to result in coastline accretion.

Most of the west bank of Hooghly estuary is prograding at the rate of 0.24 m yr.<sup>1</sup> in zone 3 to as high as 9.45 m yr.<sup>1</sup> in zone 1. Whereas recession is pre-dominant in the east bank, especially in zone 5, 6 and 7 accounting 0.38 to 4.35 m yr.<sup>1</sup> . In general, aggradation dominates over erosion. Large variation in the shoreline change envelope in zone 3, 6 and 7 reveals an active morphodynamic process. The different suite of behaviors in recent intra-decadal scale suggests that forcing of coastal change can be interpreted as a form of the time-dependent complex response of the kind envisaged by Schumm and Lichty [97] whereby changes over shorter time scale, are inherently associated with tighter cause-effect linkages at smaller spatial scales, and broader trends emerge over longer time-scales. Additionally, the phenomena of erosion and accretion are largely regulated by littoral current patterns and sediment influx from different rivers and the adjacent Bay of Bengal. The west bank of the estuary having sandy inter-tidal plain is aggrading over longer time scale whereas several areas in the east bank of muddy beaches record the high rate of erosion. The temporal pattern of erosion/accretion has been captured using the direction of change in each time interval. Some portions of the shorelines especially north of Kakdwip and Namkhana recorded a consistent high rate of erosion (<sup>&</sup>gt; 5 m yr.<sup>1</sup> ) over each interval. Although, only 1.87% of the area of the shoreline showed consistent erosion for all the time intervals but together with 'mostly erosion' type it constitutes 38.56% which is

#### *Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

alarming. These areas need to be protected from anthropogenic intervention and to be stabilized by rejuvenating protective embankments or vegetative barrier. Contrasting modes of prograding stretch adjacent to retreating stretch can be found in close proximity, particularly in zone 1 and 2, which suggests that local influences may be particularly important. Both these transitions in behavior suggest localized net littoral fluxes of sand and gravel from the north of estuary to the south-west. These localized instances of coupled behavior have led to a distinct net change in regional shoreline planform over a longer time scale. Some of the stretches of the shoreline exhibit distinct change of cuspate foreland from rounded to sharp apex especially north of Jhikarkhali and Madhusudanpur, north of Kakdwip. Erosion at the north and progradation at the west and south-west, illustrates south-west transport of sediments over the studied time scale whereas diffusive behavior dominated decadal-scale shoreline change.

The inter-temporal analysis using spatial smoothing windows of 1000 m showed that there is no consistent association between convexities/concavities and the erosion/accretion. Some concave stretches of shoreline exhibit erosional signatures, whilst others are accretional. The convexity of the shoreline near Horkhali (in the west bank) increased over time but decreased near Madhusudanpur on the east coast, however near Kakdwip and Patibunia the convexity remained almost unchanged over the years. Some of the concave stretches of the shoreline showed seaward accretion in the west bank, e.g., at Nij Kasba. The eroding sediments move parallel to the coast by alongshore currents from north to south direction and are expected to deposit around the concave coast owing to the lower current velocity [93]. As a result, the coastline can advance to the ocean around these regions. Several studies claim that a concaveshaped coastline tends to exhibit accretion while a convex-shaped coastline tends to exhibit erosion [93]. However, in the present study, several concave stretches of the east coast exhibited landward retreat of coastline typically along the Rangafala channel near Lakshmipur, between Ghiya Khal and Duraragra Gang (north of Namkhana) and small patches in Patibunia island. Presumably, both diffusive and anti-diffusive (unstable) behavior is operational [98] which are likely to change as the shoreline planform adjusts in response to the consequent patterns of erosion and deposition.

With the anticipated increase in global mean temperature by about 0.5°C, the thermal expansion and melting of ice caps and glaciers are inevitable [13] but this effect may be masked by inlet dynamics and coastal engineering projects even over extended time periods. However, the implication is that sea level rise is a secondary but inexorable cause of beach erosion in such areas which may lead to high-energy swells to reach further up the beach and redistribute sand offshore. Apart from the external and natural forces there are alarming uncontrolled anthropogenic activities which have imposed excessive pressure on the coastal landuse and exacerbating beach erosion problems along the Hooghly estuary. This will have ominous implications for ever-increasing coastal population and associated livelihood [99]. There is a need for decoupling the long-term forces from the anthropogenic effects and projecting the future scenario of coastal changes for effect coastal planning and enforcement.

#### **6. Conclusion**

The study of historical evolution and sculpturing of the coastal areas of Hooghly estuary in terms of short and longer time scale has significant importance in evaluating the criticality in shoreline change. The findings of the present study revealed that geospatial techniques are very useful for analyzing and predicting shoreline dynamics. The short- and long-term changes have been estimated using the DSAS extension tool of ArcGIS. The tool enables the calculation of several change metrics and also the rate of changes from time-series shoreline positions and helps in determining the zones of erosion and accretion. The variation is higher in the west bank than east bank except for 1973–1980. Considering the entire study period average recession is maximum in 1995–2000 (43 m 110.99) especially due to erosion in the southern part of the east bank of the estuary. Zone 5 contributed maximum towards erosion, however, in general, there is a decreasing trend of erosion, especially after 2000. While comparing with 1948-topomaps there is a significant recession (900 m) in zone 2 (between Talpati Khal and Kaldalmari) and in zone 3 (near Horkhali) by about 600 m. On the east bank, the most significant erosion is noticeable in zone 5, between Jadabnagar and Tilakmandal chak. The maximum landward retreat recorded was 2700 m near Uttar Chandannagar. The shoreline erosion is attributed to various causes such as decrease of sediment supply from rivers after construction of barrages in the upstream, land subsidence due to natural compaction or extraction of ground water, interruption of longshore sediment transport by man-made structures and dredging operation to maintain the navigation channel. In contrast, there is an increasing trend in the seaward extension of the shoreline since 2010. Between 2017 and 2021 the average accretion is 70.67 m (162.57). Very high annualized aggradation of 69.17 m and 29.93 m was recorded in zone 1 and 2 respectively over the study period. The shoreline change rate computed using WLR method reveals that zone 1, 2, 3 and 4 show the positive change (aggradation) which varies between 0.24 m yr.<sup>1</sup> (zone 3) to as high as 9.45 m yr.<sup>1</sup> (zone 1). The very high recession was found in the east bank in zone 6 ( 4.35 m yr.<sup>1</sup> ), followed by zone 5 (3.02 m yr.<sup>1</sup> ). More than 73% area of the shoreline exhibits erosion/accretion between 5 and 5 m yr.<sup>1</sup> . The proportion of very high erosion (<sup>&</sup>lt; 10 m yr.<sup>1</sup> ) and aggradation (> 20 m yr.<sup>1</sup> ) is limited to less than 2% of the shoreline. The temporal change pattern was examined using change direction in each time interval. About 1.87% of the shoreline shows consistent erosion in which at all the time-interval the direction of change was negative and an additional 36.69% constitutes of mostly eroded, characterized by erosion in at least 5 epochs. There is no area where consistent accretion was observed. In about 4.37% of the shoreline trend reversal from erosion to accretion has been observed. The change rate and pattern maps generated in the study will be helpful for policy makers to prepare a strategic coastal management plan and for future policy intervention. It is suggested that there should have a regular monitoring mechanism of this estuarine region to keep watch on the shoreline change and triggering factors and regulatory purpose.

#### **Acknowledgements**

The authors are thankful to the Chief General Manager, RRSCs (NRSC) for his keep interest and sustained support to carry out this study. Thanks, are also due to ea rthexplorer.usgs.gov for providing satellite data freely to the user community.

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

#### **Author details**

Dibyendu Dutta<sup>1</sup> \*, Tanumi Kumar<sup>1</sup> , Chiranjivi Jayaram<sup>1</sup> and Wasim Akram<sup>2</sup>

1 Regional Remote Sensing Centre – East (NRSC), Kolkata, India

2 Vidyasagar University, Midnapur, West Bengal, United States

\*Address all correspondence to: ddisro@gmail.com

© 2022 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

### **References**

[1] Lockwood M. NSDI Shoreline Briefing to the the FGDC Coordination Group. Washington, DC: Project report to NOAA/NOS; 1997

[2] Boak, EH, Turner IL. Shoreline definition and detection: A review. Journal of Coastal Research. 2005;**21**(4): 688-703

[3] Richmond MD. A Guide to the Seashores of Eastern Africa and the Western Indian Ocean Islands. Stockholm, Sweden: Swedish International Development Cooperation Agency; 1997

[4] Zhang K, Douglas BC, Leatherman SP. Global warming and coastal erosion. Climate Change. 2004;**64**(1):41-58. DOI: 10.1023/B: CLIM.0000024690.32682.48

[5] Hanslow DJ. Beach erosion trend measurement: A comparison of trend indicators. Journal of Coastal Research. 2007;**50**:588-593

[6] Paterson S, Donnell AO, Loomis DK, Hom P. The Social and Economic Effects of Shoreline Change: North Atlantic, South Atlantic, Gulf of Mexico, and Great Lakes Regional Overview. Lexington, MA: Eastern Research Group, Inc; 2010. p. 02421

[7] Regnauld H, Pirazzoli PA, Morvan G, Ruz M. Impacts of storms and evolution of the coastline in western France. Marine Geology. 2004;**210**:325-337. DOI: 10.1016/j.margeo.2004.05.014

[8] Cooper JAG, Navas F. Natural bathymetric change as a control on century-scale shoreline behaviour. Geology. 2004;**32**:513-516

[9] Vassie JM, Woodworth PL, Holt MW. An example of North Atlantic deep-ocean swell impacting Ascension and St Helena islands in the central South Atlantic. Journal of Atmospheric and Oceanic Technology. 2004;**21**:1095-1103

[10] Galgano FA, Douglas BC, Leatherman SP. Trends and variability of shoreline position. Journal of Coastal Research. 1998;**26**:282-291

[11] Dronkers J. Dynamics of Coastal Systems. Advanced Series on Ocean Engineering. Vol. 25. Hackensack, NJ: World Scientific Publishing Company; 2005

[12] Williams SJ. Sea-level rise implications for coastal regions. Journal of Coastal Research. 2013;**63**:184-196. DOI: 10.2112/SI63-015.1

[13] Intergovernmental Panel on Climate Change (IPCC). Climate Changes: The Science of Climate Change, Summary for Policymakers and Technical Summary of the Working Group 1 Report. New York: Cambridge University Press; 1996

[14] Cendrero A. Mapping and evaluation of coastal areas for planning. Ocean and Shoreline Management. 1989;**12**:427-462. DOI: 10.1016/0951-8312(89)90023-4

[15] Ryan TW, Sementilli PJ, Yuen P, Hunt BR. Extraction of shoreline features by neural nets and image processing. Photogrammetric Engineering and Remote Sensing. 1991; **57**:947-955

[16] Lillesand TM, Kiefer RW, Chipman JW. Remote Sensing and Image Interpretation. Hoboken, N.J: John Wiley; 2015

[17] Braud DH, Feng W. Semi-automated construction of the Louisiana coastline

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

digital land/water boundary using Landsat Thematic Mapper satellite imagery. Department of Geography & Anthropology, Louisiana State University, Baton Rouge, Louisiana, USA. Louisiana Applied Oil Spill Research and Development Program, OSRAPD Technical Report Series, 97 002; 1998

[18] Frazier PS, Page KJ. Water body detection and delineation with Landsat TM data. Photogrammetric Engineering and Remote Sensing. 2000;**66**:1461-1467

[19] Liu H, Jezek KC. Automated extraction of coastline from satellite imagery by integrating canny edge detection and locally adaptive thresholding methods. International Journal of Remote Sensing. 2004;**25**: 937-958

[20] Di K, Ma R, Li R. Automatic shoreline extraction from highresolution IKONOS satellite imagery. In: Proceedings of the ASPRS Annual Conference (May 2003). Anchorage: Alaska; 2003

[21] Chalabi A, Mohd-Lokman H, Mohd-Suffian I, Karamali K, Karthigeyan V, Masita M. Monitoring shoreline using IKONOS image and aerial photographs: A case study of Kuala Terengganu area, Malaysia. In: ISPRS Commission VII mid-term symposium "Remote Sensing: From Pixels to Processes". Enschede: The Netherlands; 2006. pp. 8-11

[22] Berberoglu S, Lloyd CD, Atkinson PM, Curran PJ. The integration of spectral and textural information using neural networks for land cover mapping in the Mediterranean. Computers & Geosciences. 2000;**26**:385-396

[23] Bastin L. Comparison of fuzzy c-means classification, linear mixture modeling and MLC probabilities as tools for unmixing coarse pixels. International Journal of Remote Sensing. 1997;**18**:3629-3648

[24] Zang J, Foody GM. A fuzzy classification of sub-urban land cover from remotely sensed imagery. International Journal of Remote Sensing. 1998;**19**:2721-2738

[25] Stuckens J, Coppin PR, Bauer ME. Integrating contextual information with per-pixel classification for improved land cover classification. Remote Sensing of Environment. 2000;**71**:282-296

[26] VLS. Feature Analyst Version 4.1 for Imagine. Reference Manual. Missoula, USA: Visual Learning Systems Inc.; 2007

[27] Harris PM, Ventura SJ. The integration of geographic data with remotely sensed imagery to improve classification in an urban area. Photogrammetric Engineering & Remote Sensing. 1995;**51**:993-998

[28] Stefanov W, Ramsey MS, Christensen PR. Monitoring urban land cover change: An expert system approach to land cover classification of semi-arid to arid urban centres. Remote Sensing of Environment. 2001;**77**: 173-185

[29] Vogelmann JE, Sohl T, Howard SM. Regional characterization of land cover using multiple sources of data. Photogrammetric Engineering and Remote Sensing. 1998;**64**:45-57

[30] Addo KA, Jayson-Quashigah PN, Kufogbe KS. Quantitative analysis of shoreline change using medium resolution satellite imagery in Keta, Ghana. Marine Sciences. 2011;**1**:1-9. DOI: 10.5923/j.ms.20110101.01

[31] Maiti S, Bhattacharya AK. Shoreline change analysis and its application to

prediction: A remote sensing and statistics-based approach. Marine Geology. 2009;**257**:11-23

[32] Nassar K, Mahmod WE, Fath H, Masria A, Nadaoka K, Negm A. Shoreline change detection using DSAStechnique: Case of North Sinai coast, Egypt. Marine Georesources & Geotechnology. 2019;**37**:81-95. DOI: 10.1080/1064119X.2018.1448912

[33] Nayak S. Use of satellite data in coastal mapping. Indian Cartographer. 2002;**22**:1

[34] Thieler ER et al. The Digital Shoreline Analysis System (DSAS) Version 4.0 an ArcGIS Extension for Calculating Shoreline Change. Reston, VA: U.S. Geological Survey; 2009

[35] Yu K, Hu C, Muller-Karger FE, Lu D, Soto I. Shoreline changes in west-Central Florida between 1987 and 2008 from Landsat observations. International Journal of Remote Sensing. 2011;**32**(23): 8299-8313

[36] Chen WW, Chang HK. Estimation of shoreline position and change from satellite images considering tidal variation. Estuarine and Coastal Shelf Science. 2009;**84**:54-60. DOI: 10.1016/j. ecss.2009.06.002

[37] Alesheikh A, Ghorbanali A, Nouri N. Coastline change detection using remote sensing. International Journal of Environmental Science & Technology. 2007;**4**:61-66. DOI: 10.1007/BF03325962

[38] Chandrasekar N, Viviek VJ, Saravanan S. Coastal vulnerability and shoreline changes for southern tip of India-remote sensing GIS approach. Journal of Earth Science and Climate Change. 2013;**04**:10. DOI: 10.4172/2157- 7617.1000144

[39] Kumar A, Allu NC, Jayappa KS. Shoreline changes and morphology of spits along southern Karnataka, west coast of India: A remote sensing and statistics-based approach. Geomorphology. 2010;**120**:133-152. DOI: 10.1016/j.geomorph.2010.02.023

[40] Digital Shoreline Analysis System (DSAS). ver. 5.0 ver. 5.0 User Guide, Open-File Report 2018-1179, US Department of the Interior, US Geological Survey, 2018, 110 p

[41] Brooks SM, Spencer T. Temporal and spatial variations in recession rates and sediment release from soft rock cliffs, Suffolk coast, UK. Geomorphology. 2010;**124**:26-41

[42] Burningham H, French J. Understanding coastal change using shoreline trend analysis supported by cluster-based segmentation. Geomorphology. 2017;**282**:131-149

[43] González-Villanueva R, Costas S, Pérez-Arlucea M, Jerez S, Trigo RM. Impact of atmospheric circulation patterns on coastal dune dynamics, NW Spain. Geomorphology. 2013;**185**:96-109. DOI: 10.1016/j.geomorph.2012.12.019

[44] Sheik M. A shoreline change analysis along the coast between Kanyakumari and Tuticorin, India, using digital shoreline analysis system. Geo-spatial Information Science. 2011;**14**(4):282-293

[45] Patra A, Bhaskaran PK. Trends in wind-wave climate over the head bay of Bengal region. International Journal of Climatology. 2016;**36**:4222-4240. DOI: 10.1002/joc.4627

[46] Ghosh A, Mukhopadhyay S. Quantitative study on shoreline changes and erosion hazard assessment: Case study in Muriganga–Saptamukhi interfluve, Sundarban, India. Modelling

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

Earth Systems and Environment. 2016;**2**: 1-14. DOI: 10.1007/s40808-016-0130-x

[47] Bandyopadhyay S, Nandy S. Trends of sea level rise in Hugli estuary, India. Indian Journal of Geo-marine Science. 2011;**40**:802-812

[48] Biswas AN. Geohydro-morphometry of Hooghly estuary. Journal of Institute of Engineers (India). 1985;**66**:61-73

[49] Chakraborty S. Delineation of morpho-structural changes of some selected islands in the ganga delta region, West Bengal, India: A spatio-temporal change detection analysis using GIS and remote sensing. International Journal of Science and Nature. 2013;**4**:499-507

[50] Islam MR, Begum SF, Yamaguchi Y, Ogawa K. The Ganges and Brahmaputra rivers in Bangladesh: Basin denudation and sedimentation. Hydrological Processes. 1999;**13**:2907-2923

[51] Stanley DJ, Hait AK. Holocene depositional patterns, neotectonics and Sundarban mangroves in the Western Ganges–Brahmaputra delta. Journal of Coastal Research. 2000;**16**(1):26-39

[52] Goodbred SL Jr, Kuehl SA, Steckler MS, Sarker MH. Controls on facies distribution and stratigraphic preservation in the Ganges– Brahmaputra delta sequence. Sedimentary Geology. 2003;**155**:301-316

[53] Alam M, Alam MM, Curray JR, Chowdhury MLR, Gani MR. An overview of the sedimentary geology of the Bengal Basin in relation to the regional tectonic framework and basinfill history. Sedimentary Geology. 2003; **155**:179-208

[54] Islam SN, Gnauck A. Mangrove wetland ecosystems in Ganges– Brahmaputra delta in Bangladesh.

Frontiers in Earth Sciences in China. 2008;**2**:439-448. DOI: 10.1007/s11707- 008-0049-2

[55] Chakrabarti P. Evolutionary history of the coastal quaternaries of the Bengal plain, India. Proceedings of Indian National Science Academy. 1995;**61**: 343-354

[56] Banerjee M, Sen PK. Palaeobiology in understanding the change of sea level and coast line in Bengal basin during Holocene period. Indian Journal of Earth Science. 1987;**14**:307-320

[57] Niyogi D. Geological background of beach erosion at Digha, West Bengal. Bulletin of Geology Mining and Metallurgical Society of India. 1970; **43**:1-36

[58] Gaur AS, Vora KH. Ancient shorelines of Gujarat, India, during the Indus civilization (Late mid-Holocene): A study based on archaeological evidences. Current Science. 1999;**77**: 180-185

[59] Mcfeeters SK. The use of the normalize difference water index (NDWI) in the delineation of open water features. International Journal of Remote Sensing. 1996;**17**:1425-1432. DOI: 10.1080/0143/1169608948714

[60] Xu H. Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery. International Journal of Remote Sensing. 2006;**27**(14):3025- 3033. DOI: 10.1080/01431160600589179

[61] Li W, Du Z, Ling F, Zhou D, Wang H, Gui Y, et al. A comparison of land surface water mapping using the normalized difference water index from TM, ETM plus and ALI. Remote Sensing. 2013;**5**:5530-5549. DOI: 10.3390/ rs5115530

[62] Woodshole. (1990) Available from: http://woodshole.er.usgs.gov/projectpages/dsas/

[63] Thieler ER, Danforth WW. Historical shoreline mapping (II) application of the digital shoreline mapping and analysis systems (DSMS/ DSAS) to shoreline change mapping in Puerto Rico. Journal of Coastal Research. 1994;**10**(3):600-620

[64] Theiler ER, Dadforth WW. Historical shoreline mapping (I): Improving techniques and reducing positioning errors. Journal of Coastal Research. 1994;**10**(3):549-563

[65] Himmelstoss EA. DSAS 4.0 installation instructions and user guide. In: Thieler ER, Himmelstoss EA, Zichichi JL, Ergul A, editors. 2009 Digital Shoreline Analysis System (DSAS) Version 4.0 — An ArcGIS Extension for Calculating Shoreline Change. Reston, Virginia, USA: U.S. Geological Survey Open-File Report 2008–1278; 2009

[66] Oyedotun TDT. Shoreline geometry: DSAS as a tool for historical trend analysis. In: Geomorphological Techniques. Vol. 3. London, UK: British Society for Geomorphology; 2014. pp. 1-12

[67] Chand P, Acharya P. Shoreline change and sea level rise along coast of Bhitarkanika wildlife sanctuary, Orissa: An analytical approach of remote sensing and statistical techniques. International Journal of Geomatics and Geosciences. 2010;**1**:436-455

[68] Faye I, Giraudet E, Gourmelon F, Henaff A. Cartographienormalisée de l'évolution du trait de côte. Mappemonde. 2011;**104**:12

[69] Jamont MF. Etude des aléas naturels sur le «Sud Vendée et marais Poitevin» - Rapport de phase 2 -Caractérisation des aléas de référence. France: Vendée; 2014. p. 67

[70] Prukpitikul S, Buakaew V, Keshdet W, Kongprom A, Kaewpoo N. Shoreline change prediction model for coastal zone management in Thailand. Journal of Shipping and Ocean Engineering. 2012;**2**: 238-243

[71] Honeycutt MG, Crowell M, Douglas BC. Shoreline-position forecasting: Impact of storms, rate-calculation methodologies, and temporal scales. Journal of Coastal Research. 2001;**17**: 721-730

[72] Anders FJ, Byrnes MR. Accuracy of shoreline change rates as determined from maps and aerial photographs. Shore and Beach. 1991;**59**:17-26

[73] Crowell M, Leatherman SP, Buckley MK. Historical shoreline change: Error analysis and mapping accuracy. Journal of Coastal Research. 1991;**7**:839-852

[74] Moore LJ. Shoreline mapping techniques. Journal of Coastal Research. 2000;**16**:111-124

[75] Fletcher CH, Romine BM, Genz AS, Barbee MM, Dyer M, Anderson TR, et al. National Assessment of Shoreline Change: Historical Shoreline Change in the Hawaiian Islands. Reston, Virginia: US Department of the Interiors, US Geological Survey; 2012. p. 55

[76] Romine BM, Fletcher CH. Armouring on eroding coasts leads to beach narrowing and loss on Oahu, Hawaii. In: Cooper JAG, Pilkey OH, editors. Pitfalls of Shoreline Stabilization: Selected Case Studies. Coastal Research Library. Dordrecht, Netherlands: Springer Science and Business Media; 2012

*Shoreline Change Analysis of Hooghly Estuary Using Multi-Temporal Landsat Data… DOI: http://dx.doi.org/10.5772/intechopen.103030*

[77] Schureman P. Manual of Harmonic Analysis and Prediction of Tides, Special Publication no. 98. Washington: US Department of commerce; 1994

[78] Hennig, C. (2015). Clustering strategy and method selection. Available from: https://arxiv.org/abs/1503.02059

[79] Legendre P, Legendre L. Numerical Ecology, in Developments in Environmental Modelling. 3rd ed. Vol. 24. Amsterdam, The Netherlands: Elsevier; 1998

[80] Kruskal WH, Wallis WA. Use of ranks in one-criterion variance analysis. Journal of the American Statistical Association. 1952;**47**:583-621. DOI: 10.2307/2280779

[81] Everitt BS, Landau S, Leese M, Stahl D. Cluster Analysis. 5th ed. Chichester, UK: Wiley; 2011. p. 330

[82] Bandyopadhyay S, Mukherjee D, Bag S, Pal DK, Rudra K. 20th century evolution of banks and islands of the Hugli estuary. In: Singh S, Sharma HS, De SK, editors. Geomorphology and Environment. Kolkata, India: ACP Publishers; 2004. pp. 235-263

[83] Raju DK, Santosh K, Chandrasekar J, Tiong-Sa T. Coastline change measurement and generating risk map for the coast using geographic information system. The International Archives of Photogrammetry, Remote Sensing and Spatial Information Science. 2010;**38**(part II):492-497

[84] Jana A, Sheena S, Biswas A. Morphological change study of Ghoramara Island, eastern India using multi temporal satellite data. Research Journal Recent Science. 2012;**1**:72-81

[85] Rudra K. The Atlas of the Changing River Courses in West Bengal. Kolkata: Sea Explorers' Institute (SEI); 2012

[86] Das S, Choudhury MR, Das S, Khan S. Monitoring shore line and inland changes by using multi-temporal satellite data and risk assessment: A case study of Ghoramara Island, West Bengal. International Journal of Geoscience and Technology. 2013;**1**:1-20

[87] Nimura C. Rock art and coastal change in bronze age Scandinavia. In: Encountering Imagery: Materialities, Perceptions, Relations. Dept. of Archaeology and Classical Studies, Stockholm University; 2013. pp. 117-132

[88] Jana A, Bhattacharya AK. Assessment of coastal erosion vulnerability around Midnapur-Balasore coast, eastern India using integrated remote sensing and GIS technique. Journal of the Indian Society of Remote Sensing. 2013;**41**:12. DOI: 10.1007/ s12524-012-0251-2

[89] Karunarathna H, Horrillo-Caraballo J, Kuriyama Y, Mase H, Ranasinghe R, Reeve DE. Linkages between sediment composition, wave climate and beach profile variability at multiple timescales. Marine Geology. 2016;**381**:194-208

[90] Gratiot N, Anthony EJ, Gardel AA, Gaucherel C, Proisy C, Wells JT. Significant contribution of the 18.6-year tidal cycle to regional coastal changes. Nature Geoscience. 2008;**1**:169-172

[91] Hapke CJ, Lentz EE, Gayes PT, McCoy CA, Hehre RE, Schwab WC, et al. A review of sediment budget imbalances along Fire Island, New York: Can nearshore geologic framework and patterns of shoreline change explain the deficit? Journal of Coastal Research. 2010;**26**:510-522

[92] McNinch JE. Geologic control in the near-shore: Shore-oblique sandbars and shoreline erosional hotspots, Mid

Atlantic Bight, USA. Marine Geology. 2004;**211**:121-141

[93] Xu N. Detecting coastline change with all available Landsat data over 1986–2015: A case study for the state of Texas, USA. Atmosphere. 2018;**9**(3):107. DOI: 10.3390/atmos9030107

[94] Woodworth PL, Melet A, Marcos M, Ray RD, Wöppelmann G, Sasaki YN, et al. Forcing factors affecting sea level changes at the coast. Surveys in Geophysics. 2019;**40**:1351-1397

[95] Woodroffe CD. Coasts: Form, Process and Evolution. Cambridge, UK: Cambridge University Press; 2003. p. 623

[96] Chaverot S, Héquette A, Cohen O. Changes in storminess and shoreline evolution along the northern coast of France during the second half of the 20th century. Zeitschrift für Geomorphologie, Supplementary. 2008;**52**:1-20. DOI: 10.1127/0372-8854/2008/0052S3-0001

[97] Schumm SA, Lichty RW. Time, space, and causality in geomorphology. American Journal of Science. 1965;**263**: 110-119

[98] Ashton AD, Murray AB, Littlewood R. The response of spit shapes to wave-angle climates. In: Sixth International Symposium on Coastal Engineering and Science of Coastal Sediment Process. Reston, VA: American Society of Civil Engineers; 2007. DOI: 10.1061/40926 (239)27

[99] Bird ECF. Coastline Changes: A Global Review. Chichester: John Wiley and Sons; 1985. p. 219

#### **Chapter 3**

## Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS Technique

*Ali Masria, Karim Nassar and Mohamed Galal Eltarabily*

#### **Abstract**

This study employs a digital shoreline analysis system (DSAS) to identify and evaluate historical changes in the coastline along the North Sinai coast of Egypt. Using multi-temporal satellite images, change detection is explored along coastline over 27 years (1989–2016). The annualized uncertainty of shoreline changes was calculated. Erosion and accretion patterns were automatically quantified via four statistical parameters in the DSAS model namely net shoreline movement (NSM), rate of 8.17 m year<sup>1</sup> was recorded at the west seaside of El-Tinah plain throughout the 27 years. This recession of the shoreline is attributed to the joint effect of the stormy climate of the western seaside and the sediments transport from the Nile Delta. shoreline has progressed west of El-Bardawil inlet towards El-Arish harbor, where wave-induced littoral transport is ceased by the construction of jetties. The shoreline at the downdrift side of the jetties to the east has adversely retreated where the subsequent beaches are reverted at rates of 4.5 and 2.9 m year<sup>1</sup> . Lastly, the EPR model was utilized for quantifying shoreline changes in the near future of years 2025, 2035, and 2050.

**Keywords:** satellite imagery, remote sensing technique, morphodynamic detection, DSAS model, North Sinai Coast

#### **1. Introduction**

Coastal zones are now experiencing increased natural and human disruptions, such as sea-level rise, coastal erosion, and resource overexploitation, to name a few. Coastal erosion affects almost 80% of the world's beaches, with rates ranging from 1.0 cm year<sup>1</sup> to 30 m year<sup>1</sup> , posing a major threat to several coastal regions [1].

According to [2], increased knowledge of many driving forces is affecting the health of global coastal ecosystems has expedited efforts to evaluate, monitor, and reduce coastal stressors to understand the spatial distribution of erosion risks, predict their growth tendency, and support mechanism research on erosion and its solutions.

Shoreline extraction and change detection rates at different times are critical for coastal zone monitoring. The coastline, defined by [3] as the position of the landwater interface at a single point in time, is a highly dynamic characteristic that serves as a predictor of coastal erosion and accretion. Shoreline changes occur on a variety of time scales, ranging from geological to short-term catastrophic events. Waves, winds, tides, sea-level rise, frequent storms, geomorphic processes of erosion and accretion, and human activities are all factors that affect these changes [4].

Several international studies looked at quantitative and qualitative analysis of shoreline spatiotemporal fluctuations [5–13].

Alternatively, efforts were made to estimate the potential position of the shoreline to reduce the impact of the upcoming erosion activity. Moreover, for future predictions of shoreline spatial change, extensive and reliable information regarding historical and present coastline position is required. In a GIS framework, shoreline prediction models are simple to implement. With the help of historical data, several statistical models for example the Average of Rates (AOR), Least Median of Squares (LMS), Linear Regression Rate (LRR), End Point Rate (EPR) model, and Jackknife model (JK) were used to evaluate shoreline prediction [12, 14–19].

Few encouraging investigations have been conducted along Egypt's North Sinai shore [20] used an aerial picture taken in 1955 and a topographic map analyzed in 1992 to describe the shoreline alteration along Sinai's northern coast. Despite this, the magnitude of shoreline variations was not quantified in their analysis due to the inability of the analyzed maps'surveying methodologies to calculate it. Moreover [21] used a hydrographic survey to investigate the impact of the El Arish power station (located west of the El Arish valley coast) on the surrounding area on Sinai's Mediterranean coast. The authors discovered a 5.5 m/year coastline retreat east of the El Arish power plant breakwater [22] used topographic maps from 1973 with satellite pictures from 1984 and 1996 to tutor the coastal changes over the western half of the North Sinai coast (i.e., from El Tinah Bay to El Bardawil Lake). They also calculated how the area of El Bardawil Lake changed throughout time. They discovered that the extent of El Bardawil Lake changed dramatically from 1973 to 1984, losing an average of 128 km2 , then slowing to a loss of 5 km<sup>2</sup> from 1984 to 1996. El Banna et al. (2009) used the same method to track changes in the shoreline along the North Sinai coast for 15 years (from El Bardawil Lake to Rafah) by studying TM and ETM true color Landsat pictures from 1986 to 2001. The accretion and erosion rates were calculated to be +0.076 km<sup>2</sup> year<sup>1</sup> and 0.123 km<sup>2</sup> year<sup>1</sup> , respectively.

To sum up, the extensive literature survey undertaken in the North Sinai coast area demonstrated that the current published data related to the area needs to be improved and renovated. Furthermore, it cannot determine coastline change rates with high-precision approaches. The current study uses GIS and DSAS geospatial approaches to examine shoreline changes along the North Sinai coastline from 1989 to 2016. Furthermore, the current research aims to: (1) apply three different

*Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*

semi-automated shoreline extraction methods, including Histogram threshold of band ratio, Histogram threshold of band 5, and Tasseled Cap Transformation (TCT); (2) plot and measure shoreline accretion/erosion rates using several statistical methods functionalized in DSAS, including NSM, LRR, EPR, and LMS; and (3) develop a decision-support algorithm that can vigorously support in elaborating shoreline accretion/erosion rates; (4) using the EPR model, outline a futuristic decision plotting based on the North Sinai shoreline forecast in the years 2025, 2035, and 2050.

#### **2. Study area**

Sinai's coastal area is considered an essential part of Egypt's Mediterranean Coast [22]. It is a geographical connecting point between Asia and Africa, with the Gulf of Suez and the Suez Canal on the west, the Gulf of Aqaba and the Egyptian-Israeli border on the east, and the Mediterranean Sea on the north (**Figure 1**). The latitudes and longitudes are (28°–31°N) and (32° 30/ –34° 30/ E) respectively. The northern Sinai coast stretches for about 220 km along the Mediterranean Sea, extending from Port Said in the west to Rafah in the east, from the Egyptian border [23]. The current

#### **Figure 1.**

*False-color composite images of the study area and shoreline digitization in different periods from (1989–2016) for the three zones (a, b, c) respectively, after [4].*

study area is split into three subzones based on the vulnerability of coastal areas as well as the availability of data from the field and remotely sensed data. Zone I contain El Tinah Plain Bay, which stretches 38.5 km from Port Said in the west to El Bardawil Lake in the east, (**Figure 1a**). Zone I is characterized by some features such as Lagoons, vegetation cover, and fish ponds. Zone II includes El Bardawil Lake (**Figure 1b**). This lake covers approximately 60% of Sinai's northern coast. It has a total area of over 700 km<sup>2</sup> and is approximately 72.5 km long, 22 km wide-ranging, and 2 m deep. The Mediterranean Sea is isolated from the Lake by shallow sand barriers that range in width from 300 to 1000 m, and are overtopped by storm waves in the winter. It has three inlets joining it to the Mediterranean Sea, two of which are manmade (no. 1 and 2) and one of which is natural (El Zaranek inlet), [24]. Zone III includes the El Arish Valley shore, which almost forms a 37 km west-to-east intersection between the El Arish power plant and Rafah, (**Figure 1c**).

#### **2.1 Wave climate at study area**

The intensity and direction of wave action along Egypt's Mediterranean coast are inextricably linked to significant pressure systems over the Mediterranean and North Atlantic [25]. Wave heights reach 1.16 m and average 0.4 m during the spring and summer, with the prevailing wave direction being NW. Prevailing wave direction is come mainly from N, NNW, and NW in Winter. The maximum wave height is 4.25 m, with an average wave height of 0.51 m and a period of 6.5 sec.

Wave data was analyzed previously by [26–28] along Egypt's Mediterranean coast show that waves from the northwest predominate (81%), with small components from the northeast (14%) and southwest (5%). The prevailing wave direction is the main key of the eastward-flowing alongshore current. Reversed alongshore currents are generated by waves incoming from the N, NNE, and NE (**Figure 2**). The main alongshore current path (62–65%) on the North Sinai coast is from west to east, stimulated by waves from the NNW and NW, according to preceding measured data. However, west trending alongshore currents (24–29%) result from the remaining wave components from the N, NNE, and NE, particularly during March and April due to easterly winds. Furthermore, with a range of 31 cm, the tide along Sinai's

#### **Figure 2.**

*Wave rose for study area, and wave induced currents' directions (modified from El Banna et al., 2009), after [29].*

Mediterranean coast is micro-tidal and semi-diurnal. The average high water level is 20 cm, and the average low water level is 11 cm [30].

#### **3. Materials and methods**

#### **3.1 Data source**

This study used multi-temporal satellite data from Landsat TM, ETM, and OLI/TIRS that cover our coast from 1989 to 2016. Even so, thanks to the shortage of cloud-free imagery during the selected period, satellite images could not be obtained at regular intervals. The polynomial geo-rectification method is used to ortho-rectify the selected satellite images, as it is afterward used to track changes in the shoreline along the Sinai Peninsula's northern coast Satellite images' data are described in detail in **Table 1**. Data acquired for North Sinai coastline surveying from El Tinah bay to El Arish valley was supplied by the Egyptian Institute of Oceanography and Fisheries in 2010.

#### **3.2 Image processing**

Image processing carried out in this study were strip filling, georeferencing, and radiometric correction. Firstly, gap filling was applied to image 2010 for all its bands using modeling done by [31] in Arc GIS 10.2.2 using the python algorithm, see **Figure 3**.

Ground Control Points (GPCs) are used to implement the geometric correction process (i.e. more than 40 GCPs are identified on the images), [10, 32]. The geometric correction is accomplished utilizing ENVI 5.3 software to reduce distortions caused by scale variation, angle, and lens distortion.


**Table 1.**

*Details of satellite dataset for North Sinai (acquired via https://earthexplorer.usgs.gov/).*

#### **Figure 3.**

*Landsat image for zone I in 2010; (a) before gap filling; (b) after gap filling.*

The image is projected to actual coordinate Universal Transverse Mercator (UTM), WGS-84 datum. After georeferencing, (RMSE) was found to be less than 0.5 pixels, indicating that the images were geometrically well-matched. After that, a radiometric correction is applied using the ENVI software's radiometric, which combines the sun and view angle effects, as well as sensor calibration and atmospheric correction. Eventually, all georeferenced images are processed in ArcGIS to get the coastline digitized.

#### **3.3 Shoreline delineation and uncertainties**

Shorelines are the high water line as surveyed by GPS units in kinematic mode [33]. Meanwhile, automatic coastline demarcation from low resolution satellite images is a complicated job due to the unclear boundary between water and land in saturated zone [34]. Three semi-automatic delineation approaches are first tried for Landsat images ETM 2010 in this study to identify the best digitization methodology that gives the least error with the related field data in 2010 (**Figure 4**).

Since water absorbs the majority of radiation in the near-infrared and mid-infrared regions of the spectrum, its reflectance in these wavelengths is nearly zero; nevertheless, the reflectance is higher in these areas for land cover than water bodies. As a result, the coastline can be derived from a single band image. As a consequence, getting the binary image is becoming simple by estimating the histogram threshold for one of the infrared bands (i.e. Band 5) of the TM or ETM imagery [35]. Another method is to use the histogram threshold of band ratio technique, which produces a binary image by combining the two conditions of Band (2)/Band (4) ≥ 1.0 and Band (2)/Band (5) ≥ 1.0 for producing a binary image [36].

Moreover, the Tasseled Cap Transformation technique (TCT) is also used to extract shorelines. The coefficients for TCT of Landsat data are determined from [37]. TCT reconstitutes the spectral information of the six ETM bands into three primary perspective elements using coefficients deduced from sampling known land cover spectral features. The moistness component is used to distinguish land from water among the three main view elements (brightness, greenness, and wetness). In this

*Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*

#### **Figure 4.** *Methodology framework to extract shoreline.*

wetness index band, the land-water configuration is clearly visible, and a binary image could be easily acquired. Finally, in 2010, each technique's raster binary image is transformed into a vector image, which can then be used to extract the coastline border.

To identify the best technique, a comparison is made in Arc GIS 10.2.2 for the three regions between the derived shorelines (e.g., by TCT) and the observed shorelines in 2010 (**Figure 5**). This comparison is assessed using DSAS tools, which developed by the United States Geological Survey (USGS). The original purpose of this extension is to compute change rate in coastline' positions. Using DSAS is summarized in the following steps: After extraction of shoreline, the baseline is created; creating transects; calculating the distances between coastline and baseline for transects; finally, the shoreline change rate is calculated [38]. Accordingly, the deviation between the derived and observed shoreline is computed using 1800 perpendicular to the baseline transects. These transects are accurately cast at intervals of 20 m (**Figure 5a**–**c**).

**Figure 6** depicts a validation process between data from a field investigation and extracted shoreline from satellite image obtained in 2010. It is based on the coupling of DSAS software and Arc GIS 10.2.2. The residuals between the measured and computerized shorelines in 2010 at each transect line from 1 to 1800 were estimated using both the histogram threshold of band 5, histogram threshold of band ratio, and TCT, as shown in **Figure 6a**–**c**. It is noticed that data are reasonably correlated (**Figure 6a1**, **b1**, and **c1**).

The normalized root means square error (NRMSE) is considered to find the best method that precisely extract the coastline. TCT technique is proved to be better in shoreline delimitation using low resolution satellite imagery (medium resolution). It achieved the least NRMSE for all zones, **Figure 6a1**, **b1**, and **c1**. As a result, the TCT technique was used to demarcate shorelines in 1989, 1998, 2003, and 2016 (see sectors a1, b1, b2, c1, and c2 in **Figure 1a**–**c**).

#### **3.4 Shoreline 'change rate**

Changes in the shoreline locations are calculated using different four analysis methods (i.e., EPR, LRR, LMS, and NSM). The End Point Rate (EPR) is easily

#### **Figure 5.**

*Comparison of digitized shorelines based on field data from 2010 and the corresponding Landsat imagery (e.g., using TCT) for (a) Zone (I); (b) Zone (II); (c) Zone (III).*

determined by dividing the length (in m.) between two coastlines by number of the years (Eq. (1) and **Figure 7**). This method is widely used by different coastal researchers and is widely used in shoreline movement rate calculations [39–42].

$$EPR = \frac{L\_1 - L\_2}{t\_1 - t\_2} \tag{1}$$

where:

L1 and L2 are the distances between the baseline (benchmark) and the shoreline, while t1 and t2 are the dates of the two shoreline locations.

Linear Rate Regression (LRR) is the second method for calculating change rates. For a specific transect, this method entails fitting a least-squares regression line to

*Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*

**Figure 6.**

*Validation process between the shoreline monitored in the field and the shoreline detected by imagery 2010 for the different zones based on NRMSE of ; (a,a1) histogram threshold of the band (5); (b,b1) histogram threshold of band ratio; (c,c1) TCT.*

multiple shoreline location points, (**Figure 8**). R-squared (Eq. (2)), R2 > 0.87 has been held as the threshold of certainty in our research, considering a confidence interval (LCI) of 95%. R<sup>2</sup> at each transect line are calculated as follows:

$$R^2 = 1 - \sqrt{\frac{\sum\_{i=1}^{N} \left(L - L\_p\right)^2}{\sum\_{i=1}^{N} \left(L - L^-\right)^2}}\tag{2}$$

where:

L: observed distance between the reference line(baseline) for a coastline' data point; Lp: forecast value based on the best-fit linear regression equation;

L�: Average of the observed shoreline data points; and.

N: number of dates.

The sample data are used to calculate an average offset in the linear regression method, and the formula for the line is deduced by reducing this value so that the source points are as near to the regression line as possible. In the least median of

#### **Figure 7.**

*Detecting changes of zone I; (a, b) Satellite images (TM) and OLI/TIRS of band (5) for the year1989 and 2016; (c, d) TCT's equivalent binary images from 1989 and 2016; (e) The Change detection image for the period 1989 to 2016; (f) Vector map showing the erosion/accretion pattern showed in vector map for the period 1989 to 2016.*

#### **Figure 8.**

*Explanatory example of NSM, EPR, LRR, and LMS computation; (a) Map of multi-temporal shoreline locations west and east El Bardawil inlet (1); (b) transect line' details (x) and coastline intersection; (c) Time series of shoreline distances from the baseline along the transect line (x).*

*Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*


#### **Table 2.**

*The classification of Shoreline according to EPR, LRR, and LMS.*

squares method (LMS), instead of using the average, the median value of the squared residuals is used to identify the optimal equation for the line (**Figure 8**).

The net spacing (in meters) between the past and present shoreline locations for each transect is recognized as the Net Shoreline Movement (NSM) (i.e., 1989 and 2016). It represents a distance rather than a rate (**Figure 8b**).

EPR, LRR, LMS, and NSM have negative values, implying landward decline of the shoreline, whereas positive values indicate landward advancement. The erosion/ accretion rates measured along the North Sinai coast are divided into seven categories (**Table 2**) [43].

#### **4. Results and discussion**

#### **4.1 Historical shoreline change detection (1989–2016)**

A long-term process of two-dimensional shoreline change detection has been extensively investigated along the coastal line of different three zones over a 27-year period (1989–2016). This procedure is conducted through different steps. Firstly, binary images from 1989 and 2016 are derived for each zone using TCT techniques to separate land and water. This step masks the land cover with all of its categories. Second, the binary images are converted from raster to vector (feature class) using ArcGIS10.2.2 software, with two main polygon attributes: water and land. Finally, the two polygon layers are superimposed to assess shoreline erosion/accretion trend from 1989 to 2016, (**Figure 7f**).

As a result, **Figure 7a** and **b** show satellite TM and OLI/TIRS images of the band (5) for zone I in 1989 and 2016, respectively, while **Figure 7c** and **d** show their classified binary images The post-classification change detection image (**Figure 7e**) on the other hand, shows severe erosion in El Tinah Bay's western part. This erosion is the result of the combined effects of the coast's stormy climate and the restriction of sediment movement from the Nile Delta as a result of the construction of both the jetties at Suez Canal entrance and seawalls at eastern canal. Besides that, a portion of the incident wave's energy is shifted into the adjacent beach due to the construction of this seawall. Consequently, the shifted energy, soil disconnection has occurred in the western part of El Tinah Bay. Based on the hydrodynamic processes on the North Sinai coast, alongshore currents induced sediments to move from west to east, resulting a highly sensitive eroded area (**Figure 7f**, **f1**).

In 2013, a natural opening nearly in the middle of El Mallaha Lagoon was formed as a result of this erosion. Furthermore, as a result of hindering the sediment path by inlet (2) jetties and groins at east of the inlet, part of the transported sediments has settled nearly in the middle of El Tinah Bay shoreline. In 2015, the artificial inlet (2) was completely blocked due to sediment restrictions.

The eastern part of the Bay, on the other hand, appears to be relatively stable. This part's shoreline is almost straight, with no major merged parts to erode or major embayments to receive sediments, hence, the shore zone is nearly stable; Only a few small pockets of accumulation have been noticed. To quantify the dynamical changes in the zone I coastline from 1989 to 2016, an asymmetrical difference vector map is created from binary images in ArcGIS10.2.2 and is then classified into two categories: erosion and accretion pattern, (**Figure 7f**). The change detection clearly shows a cumulative accretion of +3.442 km<sup>2</sup> and a rate of +0.127 km<sup>2</sup> /year, while the cumulative erosion is �5.409 km<sup>2</sup> and a rate of �0.2 km<sup>2</sup> /year over a period of 27 years (**Table 3**).

The defined trend of the two-dimensional shoreline change rate (km<sup>2</sup> /year) along the coastline, extracted using TCT technique is noticed to be rather coherent with other earlier studies when compared to the other two remote sensing techniques. This is evident when the current results have been compared with the previous results in researches of [22, 23] as shown in **Table 3** and **Figure 9**. Moreover, owing to variation


**Table 3.**

*Calculated area of erosion(-ve), accretion(+ve), and net balance surfaces along the North Sinai coast.*

*Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*

**Figure 9.** *Mean change rates (km2 /year) along the North Sinai shoreline.*

in both the picked elapsed times for each research and the accuracy of each methodology, there is a quiet discrepancy between the results.

#### **4.2 Analysis of shoreline kinematics by DSAS**

The digitized shorelines have been used in the ArcGIS extension Digital Shoreline Analysis System (DSAS) to calculate the rate of shoreline change in vector format over a specific period of time [44]. DSAS is a statistical software applied in coastal research to compute rate of change from historical GIS-based shoreline positions [45].

In this study, DSAS is utilized to calculate shoreline change rate from different historical shoreline locations along the North Sinai coast in 1989, 1998, 2003, 2010, and 2016. The method for determining shoreline change rates begins with the creation of a personal geodatabase for the extracted shoreline positions in ArcCatalog 10.2.2. Each shoreline has attributes that include date, length, ID, shape, and uncertainty. Each image's acquisition date is entered in the date column, whereas the length, ID, and shape are easily obtained. Uncertainties are also measured (**Table 2**) and recorded in the uncertainty column as integers. The five shoreline positions are then appended to one shapefile. Thereafter, speculative baseline is formed from the shoreline. Three different methods are available in DSAS to delineate baseline: (1) constructing a baseline along the shoreline at a particular distance; (2) utilizing a previously established baseline; (3) buffering method. The last method is the most consistent and accurate technique for baseline demarcation because it uses the same sinuosity shape as the nearby shoreline, so it was selected for the current study [12].

The baseline is then created at a buffering distance of 1000 meters offshore from the nearest shoreline.

These attributes provide DSAS with information about the sequence of transects as well as the baseline's position in relation to the shoreline (onshore or offshore). Transects have been set orthogonally from the benchmark (baseline) along the coastline of various years in 100 m intervals for the three different zones. Finally, the shoreline change rates are statistically computed using the various techniques (i.e., EPR, LRR, LMS, and NSM). As shown in (**Figure 10a**–**c**), a qualitative analysis is performed to determine the related erosion/accretion transects using the NSM model.

**Figure 10.**

*Qualitative analysis of erosion/accretion transects using NSM, conducted in DSAS, in the years of 1989, 1998, 2003, 2010, and 2016 for (a) zone (I); (b) zone (II); (c) zone (III).*

The field that connects the table of NSM statistical results to the transect feature class is the field that they have in common. Where the values in the transect-ID field of the NSM results table are equal to the object identifier field (Object ID) in the transect feature class. After completing the joining process, the symbology of the transect feature class can be adjusted to classify transects into two categories: erosion (green transects) and accretion (orange transects), (**Figure 10**). Most beaches in zones I, II, and III are susceptible to accretion and retreat (1989–2016), according to the delineation of erosion and accretion transects.

Furthermore, the results reveal that, 49.61% (191 transects), 73.52% (533 transects), and 72.24% (255 transects) of the coastline corresponding to 19.1, 53.3, and 25.65 km are experiencing erosion for zone I, II, and III respectively. On ht. eother hand, 50.39%, 26.48%, and 27.76% of the coastline with lengths of 19.4, 19.2, and 9.85 km are suffering accretion (**Table 3**).

The rates of shoreline change are computed annually for each zone during the period from 1989 to 2016 using the statistical outcomes of EPR, LRR, and LMS (**Figure 11**). The findings of this study are summarized in **Table 3**, which shows the average and maximum rates of coastline advancement and decline, as well as the percentage of degradation and deposition transects, both locally and globally. In **Figure 11** and **Table 3**, the positive and negative values of the EPR, LRR, and LMS reveal accumulation and recession areas, respectively. Apart from the last quarter of the eastern part of the El Tinah plain Bay coastline, which appears to be markedly stable over a quarter-centennial period (27 years), almost the entire coastal area of the North Sinai is sensitive to either accretion or retreat, as shown in **Figure 11a**–**c**. Meantime, when comparing the outputs of EPR and LRR to the outputs of LMS, the overall measurable change rates of the North Sinai coast show a significant affinity. Whereas, the achieved R-squared values show a strong correlation

*Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*

**Figure 11.**

*Shoreline change rates by EPR, LRR, and LMS (m/year) for Zones, (a) I; (b) II; and (c) III during the period from 1989 to 2016. Shoreline change rates by EPR, LRR, and LMS (m/year) for Zones, (a) I; (b) II; and (c) III during the period from 1989 to 2016.*

**Figure 12.**

*Comparison of shoreline change rates (m/year) calculated by, (a) EPR vs LRR; (b) EPR vs LMS; (c) LRR vs LMS for the overall North Sinai coast.*

between EPR and LRR, with a value of 0.978. The situation is different regarding EPR vs. LMS and LRR vs. LMS, as the correlation values are 0.8 and 0.836 respectively (**Figure 12a**–**c**).

#### **5. Conclusions**

Geospatial techniques and DSAS models were utilized to assess the shoreline morphodynamic changes along the North Sinai shoreline between 1989 and 2016 via multi-temporal satellite images. The semi-automatic shoreline extraction method (Tasseled Cap Transformation technique, TCT) was accustomed to digitalize the shoreline positions in 1989, 1998, 2003, and 2016. Extreme variance in the spatial scale characterizes the study area where the highest obtained coastal erosion/accretion kinematics for El Arish valley coast, El Bardawil Lake, and El Tinah Bay are �1.61/+1.95 km<sup>2</sup> , �6.95/+4.37 km<sup>2</sup> , and � 5.41/+3.44 km<sup>2</sup> , respectively.

Moreover, the construction of eastern jetty of the Suez Canal extremely lowered sediments inputs to El Tinah Bay, which highlighted the erosion of the western segment by wave hydrodynamics and the eastwards alongshore currents. Contrary, the eastern part of El Tinah Bay has demonstrated a nearly constant shoreline throughout the study period. Instantaneously, protection jetties of El Bardawil inlet 1, El Bardawil inlet 2, and El Arish Harbor have intermittent long-shore sand movement resulting in a continuous erosion at the downdrift side and an accretion at their updrift side. The institution of El Arish power plant has substantially decreased the sedimentary routine and created destructive impacts on coastal dynamics in the west of El Arish harbor.

In the meantime, the forthcoming speculation of the North Sinai coastline variations is predicted using the End Point Rate (EPR) model for the near future of years 2025, 2035, and 2050 after model validation based on 2010 data. A well-matching between the historical and futuristic trends of shoreline is obtained which means the calculation is almost succeeding the same accretion and erosion patterns. Study results in this chapter prove that medium-resolution satellite images, geospatial features of the GIS, and digital shoreline analysis system (DSAS) successfully assessed the coastal morphodynamic changes and shoreline detection of the North Sinai coast and could be used for other coastal areas based on the data quality and availability. Additionally, the results of this study deliver a high-reliable tool to the decision-makers and the coastal managers to support their decision when developing sustainable coastal management plans for North Sinai coast.

*Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*

#### **Acknowledgements**

The author would like to thank the editor and the reviewers for their constructive comments for enhancing the chapter quality.

#### **Author details**

Ali Masria1,2\*, Karim Nassar2 and Mohamed Galal Eltarabily3,4

1 Civil Engineering Department, College of Engineering, Jouf University, Saudi Arabia

2 Irrigation and Hydraulics Engineering Department, College of Engineering, Mansoura University, Mansoura, Egypt

3 Civil Engineering Department, Faculty of Engineering, Port Said University, Port Said, Egypt

4 UC Kearney Agricultural Research and Extension Center, University of California, Parlier, CA, USA

\*Address all correspondence to: aatef@ju.edu.sa; ali\_maaasria@mans.edu.eg

© 2022 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

### **References**

[1] Addo KA, Walkden M, Mills JPT. Detection, measurement, and prediction of shoreline recession in Accra, Ghana. ISPRS Journal of Photogrammetry and Remote Sensing. 2008;**63**(5):543-558

[2] Zhang Y, Lu D, Yang B, Sun C, Sun M. Coastal wetland vegetation classification with a Landsat Thematic Mapper image. International Journal of Remote Sensing. 2011;**32**(2):545-561

[3] Genz AS, Fletcher CH, Dunn RA, Frazer LN, Rooney JJ. The predictive accuracy of shoreline change rate methods and alongshore beach variation on Maui, Hawaii. Journal of Coastal Research. 2007;**23**(1):87-105

[4] Van TT, Binh TT. Shoreline change detection to serve sustainable management of coastal zone in Cuu Long Estuary. In: International Symposium on Geoinformatics for Spatial Infrastructure Development in Earth and Allied Sciences. Vol. 1. 2008

[5] Appeaning Addo K, Jayson-Quashigah PN, Kufogbe KS. Quantitative analysis of shoreline change using medium resolution satellite imagery in Keta, Ghana. Marine Sciences. 2012;**1**(1): 1-9. DOI: 10.5923/j.ms.20110101.01

[6] Kroon A, Kabuth AK, Westh S. Morphologic evolution of a storm surge barrier system. Journal of Coastal Research. 2013;**65**:529-534. DOI: 10.2112/ SI65-090

[7] Tran Thi V, Xuan ATT, Nguyen HP, Dahdouh-Guebas F, Koedam N. Application of remote sensing and GIS for detection of long-term mangrove shoreline changes in Mui Ca Mau, Vietnam. Biogeosciences. 2014;**11**(14): 3781-3795. DOI: 10.5194/bg-11-3781- 2014

[8] Murali RM, Dhiman R, Choudhary R, Seelam JK, Ilangovan D, Vethamony P. Decadal shoreline assessment using remote sensing along the central Odisha coast, India. Environment and Earth Science. 2015;**74**(10):7201-7213. DOI: 10.1007/s12665-015-4698-7

[9] El Sharnouby BA, El Alfy KS. Coastal changes along Gamasa Beach, Egypt. Journal of Coastal Zone Management. 2015;**18**(1). DOI: 10.4172/ 2473-3350.1000393

[10] Masria A, Nadaoka K, Negm A, Iskander M. Detection of shoreline and land cover changes around Rosetta Promontory, Egypt, based on remote sensing analysis. Landscape. 2015;**4**(1): 216-230. DOI: 10.3390/land4010216

[11] Bheeroo RA, Chandrasekar N, Kaliraj S, Magesh NS. Shoreline change rate and erosion risk assessment along the Trou Aux Biches–Mont Choisy beach on the northwest coast of Mauritius using GIS-DSAS technique. Environment and Earth Science. 2016;**75**(5):1-12. DOI: 10.1007/s12665-016-5311-4

[12] Nandi S, Ghosh M, Kundu A, Dutta D, Baksi M. Shoreline shifting and its prediction using remote sensing and GIS techniques: A case study of Sagar Island, West Bengal (India). Journal of Coastal Conservation. 2016;**20**(1):61-80

[13] Kermani S, Boutiba M, Guendouz M, Guettouche MS, Khelfani D. Detection and analysis of shoreline changes using geospatial tools and automatic computation: Case of jijelian sandy coast (East Algeria). Ocean and Coastal Management. 2016;**132**:46-58. DOI: 10.1016/j.ocecoaman.2016.08.010

[14] Burgess K, Jay H, Hosking A. Futurecoast: Predicting the future *Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*

coastal evolution of England and Wales. Journal of Coastal Conservation. 2004;**10** (1–2):65-71. DOI: 10.1652/1400-0350 (2004)010[0065:FPTFCE]2.0.CO;2

[15] Kuleli T. Quantitative analysis of shoreline changes at the Mediterranean coast in Turkey. Environmental Monitoring and Assessment. 2010;**167** (1–4):387-397. DOI: 10.1007/ s10661-009-1057-8

[16] Chand P, Acharya P. Shoreline change and sea-level rise along coast of Bhitarkanika wildlife sanctuary, Orissa : An analytical approach of remote sensing and statistical techniques. International Journal of Geomatics Geoscience. 2010;**1**(3):436-455

[17] Mukhopadhyay A, Mukherjee S, Mukherjee S, Ghosh S, Hazra S, Mitra D. Automatic shoreline detection and future prediction: A case study on Puri coast, Bay of Bengal, India. European Journal of Remote Sensing. 2012;**45**(1): 201-213. DOI: 10.5721/EuJRS20124519

[18] Mondal I, Bandyopadhyay J, Dhara S. Detecting shoreline changing trends using principle component analysis in Sagar Island, West Bengal, India. Spatial Information Research. 2017;**25**(1):67-73. DOI: 10.1007/ s41324-016-0076-0

[19] Bruno DE, Barca E, Goncalves RM, de Araujo Queiroz HA, Berardi L, Passarella G. Linear and evolutionary polynomial regression models to forecast coastal dynamics: Comparison and reliability assessment. Geomorphology. 2018;**300**:128-140. DOI: 10.1016/j. geomorph.2017.10.012

[20] Frihy OE, Lotfy MF. Shoreline changes and beach-sand sorting along the northern Sinai coast of Egypt. Geo-Marine Letters. 1997;**17**(2):140-146. DOI: 10.1007/s003670050019

[21] Frihy OE, Badr AA, Selim MA, El Sayed WR. Environmental impacts of El Arish power plant on the Mediterranean coast of Sinai, Egypt. Environmental Geology. 2002;**42**(6):604-611. DOI: 10.1007/s00254-002-0563-6

[22] Azab MA, Noor AM. Change detection of the North Sinai Coast by using remote sensing and geographic information system. In: The 4th International Conference and Exhibition for Environmental Technologies Environment. 2003

[23] El Banna MM, Hereher ME. Detecting temporal shoreline changes and erosion/accretion rates, using remote sensing, and their associated sediment characteristics along the coast of North Sinai, Egypt. Environmental Geology. 2009;**58**(7): 1419-1427

[24] El-Asmar HM, Hereher ME. Change detection of the coastal zone east of the Nile Delta using remote sensing. Environment and Earth Science. May 2010;**62**(4):769-777. DOI: 10.1007/ s12665-010-0564-9

[25] Hamed AAAB. Atmospheric circulation features over the southeasterly part of the Mediterranean Sea in relation with weather conditions and wind waves at Alex. 1979

[26] Nafaa MG, Fanos AM, Elganainy MA. Characteristics of waves off the Mediterranean Coast of Egypt. Journal of Coastal Research. 1991;**7**(3): 665-676

[27] Frihy O, Dewidar K. Patterns of erosion/sedimentation, heavy mineral concentration, and grain size to interpret boundaries of littoral sub-cells of the Nile Delta, Egypt. Marine Geology. 2003;**199**(1–2):27-43. DOI: 10.1016/ S0025-3227(03)00145-2

[28] Abo Zed AI, Gewilli YM. Wave climate and coastal response at Rosetta coast, Nile Delta, Egypt. In: Proceedings of 3rd Regional Conference on Arab Water. 2006

[29] Nassar K, Fath H, Mahmod WE, Masria A, Nadaoka K, Negm A. Automatic detection of shoreline change: Case of North Sinai coast, Egypt. Journal of Coastal Conservation. 2018;**22**(6): 1057-1083. DOI: 10.1007/s11852-018- 0613-1

[30] Frihy O, Badr A, Selim M, Sayed EW. Environmental impacts of El Arish power plant on the Mediterranean coast of Sinai, Egypt. Environmental Geology. 2002;**42**(6):604-611

[31] Bustillos LV. Gap fill for Landsat 7 images—A correction of SLC-off. 2012

[32] Maanan M, Ruiz-Fernandez AC, Maanan M, Fattal P, Zourarah B, Sahabi M. A long-term record of landuse change impacts on sediments in Oualidia lagoon, Morocco. International Journal of Sediment Research. 2014; **29**(1):1-10

[33] Moore LJ. Shoreline mapping techniques. Journal of Coastal Research. 2000:111-124

[34] Maiti S, Bhattacharya A. Shoreline change analysis and its application to prediction: A remote sensing and statistics-based approach. Marine Geology. 2009;**257**(1):11-23

[35] Alesheikh AA, Ghorbanali A, Nouri N. Coastline change detection using remote sensing. International journal of Environmental Science and Technology. 2007;**4**(1):61-66. DOI: 10.1007/BF03325962

[36] Niya AK, Alesheikh AA, Soltanpor M, Kheirkhahzarkesh MM. Shoreline change mapping using remote sensing and GIS-Case Study: Bushehr Province. International Journal of Remote Sensing Applications. 2013;**3**(3): 102-107

[37] Huang C, Wylie B, Yang L, Homer C, Zylstra G. Derivation of a tasselled cap transformation based on Landsat 7 at-satellite reflectance. International Journal of Remote Sensing. 2002;**23**(8):1741-1748

[38] Thieler ER, Himmelstoss EA, Zichichi JL, Miller TL. Digital shoreline analysis system (DSAS) Version 3.0: An ArcGIS extension for calculating shoreline change. United States Geological Survey, Open-File Report. 2005

[39] Dolan R, Fenster MS, Holme SJ. Temporal analysis of shoreline recession and accretion. Journal of Coastal Research. 1991:723-744

[40] Thieler ER, Himmelstoss EA, Zichichi JL, Miller TL. The Digital Shoreline Analysis System (DSAS) version 3.0, an ArcGIS extension for calculating historic shoreline change. US Geological Survey. 2005

[41] Hwang CS, Choi CU, Choi JS. Shoreline changes interpreted from multi-temporal aerial photographs and high-resolution satellite images. A case study in Jinha beach. Korean Journal of Remote Sensing. 2014;**30**(5)

[42] Ayadi K, Boutiba M, Sabatier F, Guettouche MS. Detection and analysis of historical variations in the shoreline, using digital aerial photos, satellite images, and topographic surveys DGPS: Case of the Bejaia bay (East Algeria). Arabian Journal of Geosciences. 2016; **9**(1):26

[43] Natesan U, Parthasarathy A, Vishnunath R, Kumar GEJ, Ferrer VA. *Assessment of North Sinai Shoreline Morphodynamics Using Geospatial Tools and DSAS… DOI: http://dx.doi.org/10.5772/intechopen.103031*

Monitoring longterm shoreline changes along Tamil Nadu, India using geospatial techniques. Aquatic Procedia. 2015;**4**: 325-332

[44] Hegde AV, Akshaya BJ. Shoreline transformation study of Karnataka coast: Geospatial approach. Aquatic Procedia. 2015;**4**(2015):151-156

[45] Thieler ER, Himmelstoss EA, Zichichi JL, Ergul A. The Digital Shoreline Analysis System (DSAS) version 4.0-an ArcGIS extension for calculating shoreline change. 2009

Section 3
