mapping of coastal landforms and volumetric change analysis in the south west coast of kanyakumari south india using remote sensing and gis techniques

Research PaperMapping of coastal landforms and volumetric change analysis in the south west coast of Kanyakumari, South India using remote sensing and GIS techniques S.. Change detection

Research Paper

Mapping of coastal landforms and volumetric change analysis in the

south west coast of Kanyakumari, South India using remote sensing and

GIS techniques

S Kaliraja,⇑, N Chandrasekarb, K.K Ramachandrana

a

Central Geomatics Laboratory (CGL), ESSO – National Centre for Earth Science Studies (NCESS), Akkulam, Thiruvananthapuram 695011, Kerala State, India

b

Centre for GeoTechnology, Manonmaniam Sundaranar University, Tirunelveli 627012, Tamil Nadu, India

a r t i c l e i n f o

Article history:

Received 29 March 2016

Revised 26 October 2016

Accepted 26 December 2016

Available online xxxx

Keywords:

Geomorphic Change Detection

DEM of Differencing

GIS and remote sensing

South-west coast of Kanyakumari

a b s t r a c t

The coastal landforms along the south west coast of Kanyakumari have undergone remarkable change in terms of shape and disposition due to both natural and anthropogenic interference An attempt is made here to map the coastal landforms along the coast using remote sensing and GIS techniques Spatial data sources, such as, topographical map published by Survey of India, Landsat ETM+ (30 m) image, IKONOS image (0.82 m), SRTM and ASTER DEM datasets have been comprehensively analyzed for extracting coastal landforms Change detection methods, such as, (i) topographical change detection, (ii) cross-shore profile analysis, (iii) Geomorphic Change Detection (GCD) using DEM of Difference (DoD) were adopted for assessment of volumetric changes of coastal landforms for the period between 2000 and

2011 The GCD analysis uses ASTER and SRTM DEM datasets by resampling them into common scale (pixel size) using pixel-by-pixel based Wavelet Transform and Pan-Sharpening techniques in ERDAS Imagine software Volumetric changes of coastal landforms were validated with data derived from GPS-based field survey Coastal landform units were mapped based on process of their evolution such

as beach landforms including sandy beach, cusp, berm, scarp, beach terrace, upland, rockyshore, cliffs, wave-cut notches and wave-cut platforms; and the fluvial landforms Comprising of alluvial plain, flood plains, and other shallow marshes in estuaries The topographical change analysis reveals that the beach landforms have reduced their elevation ranging from 1 to 3 m probably due to sediment removal or flat-tening Analysis of cross-shore profiles for twelve locations indicate varying degrees of loss or gain of coastal landforms For example, the K3-K30profile across the Kovalam coast has shown significant erosion (0.26 to 0.76 m) of the sandy beaches resulting in the formation of beach cusps and beach scarps within a distance of 300 m from the shoreline The volumetric change of sediment load estimated based

on DoD model depict a loss of 241.69 m3/km2for 62.82 km2of the area and land gain of 6.96 m3/km2for 202.80 km2of the area during 2000–2011 However, an area of 26.38 km2unchanged by maintaining equilibrium in sediment budgeting along the coastal stretch The study apart from providing insight into the decadal change of coastal settings also supplements a database on the vulnerability of the coast, which would help the coastal managers in future

Ó 2016 National Authority for Remote Sensing and Space Sciences Production and hosting by Elsevier B

V This is an open access article under the CC BY-NC-ND license (

http://creativecommons.org/licenses/by-nc-nd/4.0/)

1 Introduction

Geomorphic landforms of a coast is an expression of the

charac-teristics of prevailing coastal processes over long-term scale The

landforms of the coastal transition zone are sensitive to erosional and depositional processes due to actions of waves, littoral current, wind, sediment transport and certain anthropogenic activities

2004; Pavlopoulos et al., 2009; Chandrasekar et al., 2012) Coastal landform configurations are dependent on the pre-existing coastal settings, geological structures and a variety of coastal processes Therefore, mapping of landforms provides Insight into such morpho-hydrodynamic milieu (Davies, 1972; Nordstrom, 2000;

http://dx.doi.org/10.1016/j.ejrs.2016.12.006

1110-9823/Ó 2016 National Authority for Remote Sensing and Space Sciences Production and hosting by Elsevier B.V.

This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).

Peer review under responsibility of National Authority for Remote Sensing and

Space Sciences.

⇑ Corresponding author.

E-mail addresses: s.kaliraj@ncess.gov.in (S Kaliraj), profncsekar@gmail.com

(N Chandrasekar), raman.kk@ncess.gov.in (K.K Ramachandran).

Contents lists available atScienceDirect

The Egyptian Journal of Remote Sensing and Space Sciences

j o u r n a l h o m e p a g e : w w w s c i e n c e d i r e c t c o m

Along the Indian coast too, the tectonic and structural formations

and continental shelves primarily responsible for shaping the

land-forms which are acted upon subsequently by the prevailing

hydro-dynamic settings characteristics (Nayak and Sahai, 1985;

Chandrasekar and Rajamanickam, 1995; Sajeev et al., 1997; Sanil

Kumar et al., 2006; Magesh et al., 2014)

Most of the landforms along southern coast of Tamil Nadu

par-ticularly on the south west coast of Kanyakumari district have

undergone morphological deformation due to the effect of

Tsu-namic occurred on December 26, 2004 (Chandrasekar et al.,

2012) Artificial structures like groins, revetments, seawall and

jet-ties those came up in the recent years have modified the coastal

processes causing severe erosion on down-drift side in the coastal

area (Kaliraj et al., 2013) Assessment of coastal landform changes

can help in the analysis of coastal vulnerability (Nicholls et al.,

2007; Kaliraj and Chandrasekar, 2012; James et al., 2012;

Joevivek et al., 2013) Conventionally, mapping of coastal

land-forms is performed using pre-existing maps, field observation

and other collateral data sources compiled for different times

and different scales which can lead to inaccurate information due

to dynamic nature of coastal landforms (Desai et al., 1991;

using multi-temporal satellite images can provide robust

informa-tion on shape, distribuinforma-tion, and morphological status during past

and present (Butler and Walsh, 1998; Bocco et al., 2001; Smith

et al., 2006; Bubenzer and Bolten, 2008; Abermann et al., 2010)

Recent technological advancement in remote sensing and

survey-ing techniques provides adequate information on spatial

distribu-tion of coastal landforms in GIS environment enabling us to

prepare coastal geomorphologic map with higher granularity on

a larger scalability (Chockalingam, 1993; Chandrasekar et al.,

2000; Slaymaker, 2001; Nayak, 2002; Jayappa et al., 2006; Smith

and Pain, 2009; Kaliraj and Chandrasekar, 2012) Coastal landforms

of an area can be extracted using the Landsat ETM+ image with or

without slope and topographical measurements onto a GIS based

complementary platform (Mujabar and Chandrasekar, 2011)

Moreover, recent advances in remote sensing and GIS play an

important role on the development of numerical modelling of

sur-face processes for quantitative assessment of morphological

changes of landforms (Blanchard et al., 2010) GIS technique is an

effective platform for mapping thematic features with

correspond-ing attributes Geo-computational algorithms facilitates automatic

extraction of geomorphic landforms from the combination of

data-sets such as, satellite image, DEM and topographical map using

numerical modelling, pixel-based classification and cellular

auto-mated techniques in GIS environment (Dawson and Smithers,

2010) High spatial resolution images of IKONOS, Quick Bird, and

GeoEye incorporated with DEMs are progressively used for

assess-ment of volumetric changes of coastal landforms (Bubenzer and

Bolten, 2009; James et al., 2012) The mapping of geomorphic

land-forms using remotely sensed images requires knowledge of basic

interpretation elements such as tone, texture, shape, size and

pat-tern, for unambiguous delineation of landforms For example, the

beaches and associated landforms have been identified based on

linear shape and fine to medium coarse pattern (Rao, 2002)

Land-forms are interpreted using multispectral images based on

inter-pretation element keys to extract the information relatively

accurate up to the post-field verification process (Maksud Kamal

2012, the coastal landforms are classified on the basis of

topo-graphical variations resulting from differential erosion and

accre-tion processes, for example, the geomorphic units of alluvial

plain, pediplain, structural hills and residual hills are mapped using

DEM incorporated multispectral IRS-ID LISS-III images using visual

interpretation technique along with field check While digital

anal-ysis of landform extraction is faster, appropriate based on spectral

signature, and pattern recognition of image properties using math-ematical that would be able to detect, cluster and classify the fea-tures to represent the real world Previous investigations have underlined advantages of using DEM and Lidar datasets for geo-morphic detection and volumetric change of sediment load along the coastal area (Shaikh et al., 1989; Anbarasu, 1994; Lillysand and Kiefer, 2000; Wright et al., 2006; Waldhoff et al., 2008; Smith and Pain, 2009; Blanchard et al., 2010) Assessment of topo-graphical changes using DEMs provide insight on changes of sedi-ment load due to erosion or deposition processes signifying past and present morphological structural response to coastal processes over time (Lane et al., 2003; Zhang et al., 2005; Wheaton et al., 2010; Schwendel et al., 2012) The DEM datasets acquired on two different times can preferably be used to measure vertical differ-ence in sediment loads of the coastal landforms based on topolog-ical and morphometric rules (James et al., 2012) The DEM datasets such as SRTM and ASTER are being used for Geomorphic Change Detection analysis because of its mission specified accuracy, i.e high vertical accuracies over terrain surface and bare soils and medium accuracies in terms of spatial resolutions (Cuartero

et al., 2004) The topographical changes of the sediment load in the coastal landforms has been estimated from the temporal DEMs using the extracted cross-shore profile analysis that provide ade-quate information on geomorphic change of the various landforms

in vertical scale (Gyasi-Agyei et al., 1995; Zandbergen, 2008;

Geomor-phic Change Detection (GCD) analysis provides volumetric change

of coastal landforms from the DEMs acquired for different periods

of interval (Lee, 1991; Wheaton et al., 2007; Siart et al., 2009; James et al., 2012) The GCD analysis is concerned with DEM of Dif-ference (DoD) algorithm to estimate quantitative changes of land-forms, in a diverse set of environments, and at ranges of spatial scales and temporal frequencies (Wheaton et al., 2010; Hicks,

2012) The volumetric change of geomorphic features is estimated using two DEM data sets acquired for two different periods can result in estimating of land loss and land gain for a vast area appro-priately validated through field surveys and measurements

Midor-ikawa (2004) have obtained the area and volume of geomorphic features that closely matched with field measurements Stereo-pair of images are able to provide three-dimensional representa-tions of the features through accurate derivation of digital elevation models (DEMs) The topographical changes of landforms estimated from these datasets have positive correlation with field measurements and hence useful for monitoring how landforms change over time due to subsidence or uplift of the coastal surface (Cuartero et al., 2004; Mith and Clark, 2005).Knight et al., 2011

have incorporated images and DEMs for rapid assessment of land-form changes for large areas and have demonstrated that the remote sensing provides complete requirements if synergized with ground validation and measurements which can even be extended

to geomorphological studies across all spatial scale Mapping of landforms through field observation allows the most direct way

to capture the landform characteristics and enable as a basis for terrain assessment and geomorphological analysis The accuracy

of field mapping is subjective and affected by the skills and expe-rience of one who maps The volumetric change of sediment load estimated using DEMs are capable of generating superior results

on land loss and land gain that are relatively closer to the field-based measurements apart from providing spatial ensemble of coastal landforms with exceptional details (Smith and Pain,

2006) Many researchers have confirmed that the DEM derived results along with field data can produce relatively high accuracy

in geomorphic change measurement for coastal area (Aniello, 2003; Nikolakopoulos et al., 2006; Zandbergen, 2008; Potts et al., 2008; Toutin, 2008; Blanchard et al., 2010) The primary aim of

the present study is to map coastal landforms and assess the

volu-metric change of sediment load over a decade along the south-west

coast of Kanyakumari using integrated remote sensing and GIS

techniques The present study therefore used different change

detection techniques such as (i) topographical change analysis,

(ii) cross-shoreprofile change analysis, (iii) DEM of Difference

(DoD) algorithm based Geomorphic Change Detection (GCD)

anal-ysis for estimating the volumetric changes (land loss or land gain)

along the coastal stretch using the ArcGIS platform This

studya-part from assimilating decadal changes in landforms would also

delineate various influencing factors that would form primary

information source for coastal vulnerability management and

would help in the preparation of developmental plans against

any possible natural disasters that likely to affect the coastal

region

2 Study area

The study area is located along the south-west coast of

Kanyakumari district, Tamil Nadu, India The geographical

coordi-nates extend from 77°9049.2000E to 77°34015.0000E longitude and

8°6032.6000N to 8°14015.3000N latitude The coastal stretch is

extended for a length of 58 km from Kanyakumari to

Thengapatti-nam in southeast to northwest direction (Fig 1) There are three

major drainage networks such as Pazhayar, Valliyar and

Thamira-barani along with their tributaries flowing in southerly direction

from the Western Ghats These are primary sources contributing

discharge to maintain coastal landforms debouching their

sedi-ment load during both southwest and northeast monsoons (Chandrasekar and Mujabar, 2010) The coastal area is character-ized by various landforms such as sandy beaches, coastal plains, beach terraces, sand dunes, rocky shore, estuaries and other fluvio-marine landforms (Kaliraj et al., 2013) The coastal upland

in the Kanyakumari, Muttam and Colachel area are mainly associ-ated with rocky-shores and offshore outcrops acting as natural bar-rier to wave actions and storm surges Sandy beaches are formed

on the Sanguthurai, Chothavilai, Pillaithoppu, Ganapathipuram, Rajakkamangalam, Colachel and Simonkudiyiruppu coastal stretches due to swashing of large amount of sediments resulting from waves (Hentry et al., 2010) However, the major parts of the coastal areas namely Kovalam, Pallam, Manavalakurichi, Mandai-kadu and Inayamputhenthurai are noticed with severe erosional activities due to backwashing of sediments by destructive wave actions Onshore margin of the study area comprises Late Quater-nary deposits composed of complex settings of granite-biotite-illuminate underlain sandstone interlined with sand, silt and clay partings and overlaid by sandy materials (Loveson, 1993) The coastal surface is generally sloping towards sea interspersed with settlements, coconut plantation, shallow water bodies like back-water and creeks (Jena et al., 2001; Magesh et al., 2014) Along the near shore area, the sand dunes are roughly parallel to shore though discontinuously distributed along the coast The coastline along the Kanyakumari coast have experienced erosion due to high-energy wave action The Teri sand dunes (reddish brown) are located along the coastal stretch from Kovalam to Manakudi with thickness increasing from 1.5 m in coastal headlands to a maximum of 7.0 m in the interior terrestrial area The crystalline

Fig 1 Geographical location of the study area.

rock types such as quaternary rocks, clay sand and sandy materials

are predominantly found along the coast The rocky boulders and

sea cliffs are found in the Muttam, Kanyakumari and Cape Comorin

coasts and sandstones are found along the study area are made up

of igneous rock, and silt clay materials (Loveson, 1993) The alluvial

sediments admixture with clay are found deposited at the mouth

of the Thamirabarani estuary in Thengapattinam and Pazhayar

estuary in Manakudi Sandstone with clay intercalation structures

is present along the eastern part of Thengapattinam coast The

study area is prevailing with a sub-tropical climate with the

nor-mal annual rainfall varying from 826 mm to 1456 mm and the

annual mean minimum and maximum temperatures are

23.78°C–33.95 °C respectively The landforms along the coast

fre-quently alter in morphological distribution due to both natural and

anthropogenic factors and hence the present study is performed to

understand coastal landforms and their changes

2.1 Coastal and oceanographic characteristics of the study area

Evolution of coastal landforms in various locations along the

study area is mainly subject to ocean and coastal processes Wave

height is one of the main factors considered in setting up of hazard

management system along the coastal region Mean significant

wave height along the coast is estimated around 1.4 m rendering

higher energy along the coast causing erosion or subsidence of

bea-ches (Hentry et al., 2012) The low-energy waves (with wave

height <1 m) leads accretional processes due to entrainment of

sediments carried by the swash (Murty and Varadachari, 1980;

Jena et al., 2001; Mishra et al., 2011) Waves and currents

prevail-ing in an area influences erosion and accretion of the beaches and

headland features It has been reported that along this coastal

stretch, the mean annual wave energy ranges from 0.5 to 8.5 kJ/

km2 Nevertheless, the Kanyakumari coast is reportedly

experi-enced with very high wave energy (6.5–8.5 kJ/km2) owing to the

peculiar nature of the coastal configuration Due to the unique

geo-graphical location of the coast at the southernmost tip of the Indian

sub-continent, the waves and currents waves approaching from

various directions while breaking at varying angle with the coast

generate longshore currents in different directions and intensity

The coastal zones having lower energy (0.5–2.5 kJ/km2 and 1.5–

3.5 kJ/km2) wave setup result in sandy beaches due to deposition

of littoral sediments, consequent to which young beaches and

other depositional landforms are formed The southwest coast

experiences three types of littoral current systems based on the

wave direction and wind blow namely southwest monsoon current

(June–September), north-east monsoon current

(October–Decem-ber) and summer current (March–May) The movement of seasonal

currents varies in different parts causing shoreline changes due to

deposition or scouring away of sediments by seasonal movement

of longshore currents

The average longshore current velocity along the coast is

mea-sured as 0.14 m/s, whereas the fastest flow of its velocity observed

is 0.32 and 0.28 m/s in the Kanyakumari and Kovalam coasts

dur-ing both SW and NE monsoons This coastal stretch faces erosion

due to littoral sediment movement towards north during the NE

monsoon, while it is reversed towards southern direction during

SW monsoon The coast between Rajakkamangalam and Manakudi

has been notified as low current velocity zones (0.14–0.22 m/s)

and these areas have experienced accretion due to the presence

of sea cliffs, eroded Teri sand dunes and very narrow sandy pocket

beaches The coastal configuration from Manavalakurichi to

Then-gapattinam is towards southeast and north-west, where the

sum-mer and SW monsoon high velocity (0.28–0.30 m/s) currents

flowing to south act on the headlands This results in severe

ero-sion along the coast scouring and removal of sediments by

back-wash and longshore current permanent and episodic change in

the low-lying areas of the coast The tidal fluctuation is from 4.0

to 6.0 m along the estuaries causing significant changes in land-forms Whereas, areas experiencing lower tidal range from 1.0 to 2.0 m show tendency of releasing suspended sediments to the coast

The aeolian (wind) process is controlling formations and shape

of the beaches and backshore landforms such as sand dune, barrier dune (foredune) depending on the seasonal wind velocity and directions Beaches along the open coast are eroded due to abra-sion or scouring through sandblasting of wind-borne action by trapping sediments on the backshore Sand dunes in the area con-sists of dry sands that got piled or heaped-up by continuous eolian action over a long period of time For example, the parabolic dune complexes have been evolved to the present elevation of 2–4 m due to accumulation of wind transported sediments from blowouts

or open beaches along the various parts of the study area The fore-dune complexes along the coastal stretches have also been formed due to sand blowing out from the incipient beaches and steadily growing to the seaward side The recent development of coastal structures like groins, seawalls, revetments and jetties are inter-vening with the natural rhythm of the coastal process causing sev-ere erosion on the down-drift side complementing accretion on the up-drift side It is observed that the shape, size and distribution of landforms are frequently influenced by coastal and oceanographic factors and along certain stretches due to the influence of anthro-pogenic activities

3 Materials and methods Mapping of coastal landforms is primary to understanding of evolution of any coastal area The south-west coast of Tamil Nadu comprises of various landforms that are experiencing morphody-namic changes in shape, size and distribution due to various coastal hydrodynamic factors including human interferences (Hentry et al., 2012) The depositional landforms like beaches and foredunes are maintaining stable morphological structures along coastal stretches prevailing with constructive wave action coastal stretches prevailing with constructive wave action (Kaliraj and Chandrasekar, 2012) Quantifying volumetric change

of sediment load in a particular area provides insight into the ero-sional or depositional processes taking place over a period of time (Schwendel et al., 2012) The coastal landforms and vegetation cover along the coast has been significantly altered in terms of morphological settings has been significantly altered in terms of morphological settings after the Tsunami occurred on 26th, December 2004 (Chandrasekar et al., 2012) However, the coastal landforms have gradually disappeared in the down-drift side of the coastal structures such as groins, revetments, seawall and jet-ties due to interference to the littoral sediment flow along the coast (Kaliraj et al., 2013)

In some parts of the open coast, the high-energy waves and sea-sonal movement of littoral currents directly influence the sediment transport causing frequent changes of landforms and their mor-phological characteristics (Chandrasekar et al., 1996; Kaliraj

beach profile analysis reveals that severe erosion in the Kovalam, Murungavilai, Mandaikadu, and Inayamputhenthurai coastal zones due to destructive wave actions and seasonal movement of littoral currents Meanwhile, the constructive waves lead to processes of

on the beaches of Sanguthurai, Chothavilai, and Midalam and up-drift side of Muttam coast (Cherian et al., 2012) The landforms

of the coast are highly sensitive to marine and terrestrial forces to maintaining equilibrium and stability to the morphological struc-tures, and hence analysis of the changes in coastal landforms using Remote sensing and GIS techniques indispensable inputs for

coastal zone management Recent developments of geospatial

technologies enable synergizing and analyzing spatial data sources

such as maps, images and DEMs for extracting coastal landforms of

an area in higher scalability (Siart et al., 2009)

The volumetric changes of sediment load can be estimated

using DEMs acquired on temporal frequencies to derive vertical

difference in elevation at a point of observation to assimilate

changes with time (Farrell et al., 2012) The geomorphic change

assessment using DEM of Difference (DoD) provide insight into

volumetric change of landforms based on the numerical

morpho-dynamics models (Wheaton et al., 2010) Earlier studies have

demonstrated that the remote sensing and GIS approaches are

effective tools for analysis especially by incorporating time-line data to inquire into morphological change of the coastal landforms 3.1 Mapping of coastal landforms

In the present study, the integrated remote sensing and GIS technique is employed for extracting the coastal geomorphological landforms at high resolution The various types of spatial data sources such as topographical map (scale 1:25,000) published by Survey of India in the year 2000, Landsat ETM+ image (30 m) acquired for 2011, IKONOS multi-spectral high resolution image (3.2 m), ASTER and SRTM DEM datasets are used for mapping the

Fig 2 Methodology flow-chart of coastal landform mapping and volumetric change analysis.

coastal landforms through a series of systematic geo-processing

operations with ArcGIS 10.2 software The Garmin ETREX 30 GPS

was used for ground truth verification, pre and post field

verifica-tion, training sets selection and checking the landform boundaries

The location point accuracy of Garmin ETREX 30 GPS has

limita-tions in field survey using, which horizontal positioning devialimita-tions

can be up to 4 m in open coastal region with proper sky view and in

the worst scenario up to 10 m in areas with canopy cover, and

con-siderable built-up where satellite visibility is much affected

Stan-dard procedures were adopted for geomorphological mapping

from multiple spatial data sources employing visual or digital

interpretation/classification techniques (Zandbergen, 2008)

Fig 2shows the geo-processing functional flow of coastal

land-form mapping and volumetric change analysis Wherein, the

mul-tiple data sets were georeferenced to a common datum (WGS84)

with projection onto UTM and systematic operations using

geopro-cessing tools were executed for pre-progeopro-cessing and feature

extrac-tion in a GIS environment The multispectral images are spectrally

enhanced using majority filtering and mean filtering algorithm to

obtain smooth spectral pattern of features in the image (Wood,

1996; Wright et al., 2006; Mujabar and Chandrasekar, 2011)

Even-tually, the DEM is processed using hydrological analysis tool in

ArcGIS to remove sink pixels Sub-pixel classification has been

exe-cuted using the hybrid model composed of parallelepiped and

maximum likelihood algorithms based on the concept of best

pos-sibilities for operator controlled pixel allocation method (Kaliraj

beach berms, beach cusps, beach ridge and beach terrace from

the satellite image with medium resolution (30 m) is a complex

process owing to their shape and dimension on the earth surface

In order to overcome this complexity, the sub-pixel classification analysis is executed on 5 5 km grids (for nearshore area) for inte-grated images of high and medium resolution images This analysis extracts one feature (i.e beach berms) at a time and the repeated analysis produces small dimensional features Hence, the multiple layers of different landforms are overlaid together and graphically represented them in 1: 10,000 scale Encouragingly, the result of classified image shows an overall accuracy is 89.61% with a kappa coefficient statistics of 0.89 for 100 control points indicating the acceptable accuracy of the classified image Based on the individual feature classes, the producer’s accuracy is recorded as 74–100% and the user’s accuracy is estimated as 60–100%, and it is mostly exceeded 90% in the final output Additionally, the immanent pix-els within a feature class are eliminated through majority-filtering technique (kernel window size 5 5) to achieve a stronger gener-alization for mapping the thematic classes in vector format (Potts

et al., 2008) Furthermore, the resultant map is classified into var-ious landforms by comparing the morphometric rule using visual image interpretation techniques The SOI topographical map has been used to derive the basic geomorphic information and tectonic elements of the coastal area such as elevation (contour line), spot heights, benchmarks, high water line (HWL), and other natural and manmade landmarks and landscapes Landsat ETM+ and IKO-NOS images were used as primary data source for mapping the coastal landforms such as coastal plain, beach landforms, flood plain, swale, water bodies, swamp or marsh lands, tidal plats, back-water creeks, and estuaries (Magesh et al., 2014) The combined datasets of DEM and image provide a vital clue for mapping the

Fig 3 Coastal landforms map of the study area in 1: 10,000 scale extracted using Landsat ETM+ image, IKONOS image, ASTER DEM and Topographical map.

coastal landforms (Blanchard et al., 2010) The ASTER DEM (30 m)

data has been incorporated with Landsat ETM+ image and IKONOS

image for detecting the various geomorphologic units such as sand

dunes, barrier dunes, coastal uplands, beach slope, cliffs, rocky

shore, offshore rocky outcrops, mining area, sand spits and

man-made structures like groins, seawalls, revetments, jetties Finally,

the combined datasets of ASTER DEM and topographical map have

been to demarcate the topographical (relief) characteristics

land-forms in the study area

3.2 DEM of Differencing of volumetric change analysis

The GIS-based Geomorphic Change Detection (GCD) analysis

provides volumetric change of sediment load in the landforms

using DEM datasets acquired over periods of interval (Lee, 1991;

Wheaton et al., 2010; James et al., 2012) The GCD analysis is

con-cerned with DEM of Difference (DoD) algorithm for estimate the

quantitative changes of landforms of the earth surface, in a diverse

set of environments, and at a range of spatial scales and temporal

frequencies (Hicks, 2012) In the present study, Geomorphic

Change Detection of coastal landforms is estimated from SRTM

and ASTER DEM datasets acquired for the years 2000 and 2011

respectively using DEM of Difference (DoD) method The DoD is a

mathematical algorithm for quantifying the volumetric change of

the landforms using DEM datasets acquired on two different

peri-ods (Wheaton et al., 2010) The DEM datasets register the absolute

ground elevation model of an area in the form of a raster data in

which each grid cell (pixels) contains an elevation (height) values

Therefore, the DEM datasets are used worldwide for terrain

visual-ization, hydrological modelling, orthorectification, and geomorphic

change assessment (Fabris and Pesci, 2005; Nikolakopoulos et al.,

2006; Siart et al., 2009; Blanchard et al., 2010; Marghany et al.,

2010) Conversely, the DEM datasets such as SRTM (90 m) and

ASTER (30 m) are constrained due to varying spatial and temporal

resolutions whilst they are used directly for Geomorphic Change

Detection analysis and they have limited experiences within the

user community (James et al., 2012) Alternatively, to overcome

this, DEM datasets were analyzed based on the most popular Data

Fusion techniques such as pixel-by-pixel based Wavelet Transform

and Pan-Sharpening techniques using ERDAS Imagine software to

synergize DEM onto a common scale, i.e the output DEMs are

resampled to 10 10 m pixel (Wechsler, 2000; Wheaton, 2008;

Blanchard et al., 2010; Magesh et al., 2014) The coastal landform

features such as beach berms, beach terrace, beach cusps, beach

ridges were demarcated from IKONOS images (3.42 m) due to their

smaller spatial dimensions The layer consisting of the coastal

geo-morphological features are overlaid on the DEM for estimating

vol-umetric changes over time The resultant of DEMs have preserved their pixel attributes (elevation), as it is an original image and the RMSE (root mean square error) values of DEMs have noted as 3.46 m and 6.05 m respectively The DEM derived through the data fusion technique is found superior (with RMSE of 3.46 m) com-pared to the SRTM DEM Global scale (RMSE 16 m) or Local scale (6 m) Similarly, the RMSE of the reworked ASTER DEM is 6.05 compared to 25 m for Local scale (United States Geological Survey, 1997; Milan et al., 2011) Therefore, RMSE of both DEMs are reasonably good for Geomorphic Change Detection if not for the expected precision, a clear differentiation is possible which is indicative of geomorphic changes The GCD analysis incorporates DoD algorithm processed using ArcGIS AddIn Tool namely ‘‘Geo-morphic Change Detection Tool” (GCD v6.0) developed by

changes of coastal landforms during the year of 2000 and 2011 The DoD algorithm computes the differences by subtracting pixel values of two DEMs using the equation dE= Z2 Z1, where dEis a output DEM showing changes in volumetric scale (m3); Z1 is a DEM of earlier period (i.e SRTM DEM acquired on February 2000, and Z2is a DEM of later period (i.e ASTER DEM acquired on Octo-ber 2011) Thus, the output DEM provides volumetric change of sediment load (dE) on various landforms due to erosion and depo-sition with time In which, the negative and positive values repre-sent the land lost (erosion) and land gain (deposition) However, the product of DoD may propagate and amplify certain uncertain-ties and therefore, it is essential to identify and minimize errors (Wood, 1996; Blanchard et al., 2010) These complexities are elim-inated from the output DEM the equations ZActual= ZDEM± dZ, where

ZActualis the true elevation value obtained from the topographical map (scale 1:25,000) published by Survey of India in 2000 and dZ

is the vertical error component of input datasets Thus, the result

of error analysis ensures the quality of output DoD map on the rel-ative accuracy of volumetric change assessment (Siart et al., 2009; Marghany et al., 2010; Wheaton et al., 2010; James et al., 2012) Finally, the output map is converted into vector layer for prepara-tion of geo-database of landform features with attributes including name, areal extent, and volumetric change rate using ESRI-ArcGIS 10.2 software

4 Results and discussion The various types of coastal landforms are extracted from a combination of datasets using remote sensing and GIS techniques for the study area The landforms have undergone remarkable changes due to marine and terrestrial factors, which are responsi-ble for the formation of erosional and depositional landforms

Table 1

Classification of the coastal landforms in the south-west coast of Kanyakumari.

Sl no Origin process

of landforms

Erosional features Depositional features

cliffs, Wave-cut platforms, Wave-cut notches, Cliff terraces, and Headlands

Sandy beach, Sand bar, Sand spits, Beach ridge, Beach berm, Tidal flat, Mud flat, and Coastal plains

Erosion – is due to backwashing of sediments by waves, currents, placer mining and man-made structures

Accretion – is due to swashing of sediments by low wave energy and sediment deposition by longshore drift

2 Fluvio-Marine Estuaries, Shoal and Swale Deltaic plain, Sand bar (at

estuary)

Modification of landforms – due to tidal regime, divergent wave action Formation of landforms – due to accumulation of river discharged sediments by tidal and wave divergent action

terraces, Pediplain, Bajada, and Structural hill

Alluvial plain, Deltaic plain, Flood plain, Leeves

Erosion of landform – due to runoff and overland flow Deposition of landform – due to discharge of sediments by the river and channels

the shore cliffs

Sand dune (older) and Barrier dune

Erosion – due to deflation of dune sand by wind from land and sea and high energy wave actions

Accretion - due to accumulation of sediments by sand blown towards inland from the beaches in front of them by onshore winds

(Saravanan and Chandrasekar, 2010) The volumetric change of

landforms is quantified using DEM datasets in GIS software

envi-ronment, and the results are presented detailing the location,

dis-tribution and volumetric change of sediment load for various

landforms along the coastal area

4.1 Mapping of coastal geomorphic features and its spatial

distribution

The spatial distribution of the coastal landforms along the study

area is depicted in theFig 3.Table 1numerates the classification of

coastal landforms of the study area based on their evaluation

pro-cesses Evolution of marine landforms is the result of cyclic coastal

processes of ocean waves, winds, tides, and currents resulting in

the formation of erosional and depositional landforms along the

coast (Samsuddin et al., 1991) The coastal plain is a flat or gently

sloping surface distributed along the backshore region It

com-prises of sand, silt and clay particles manifested as

geomorpholog-ical entities, such as, sand dunes, plantation, shrub vegetation,

saltwater ponds, and backwater creeks resulting from the

deposi-tion of sediments over long periods The total area of coastal plain

covers 51.91 km2resulting from the 17.78% of the total study area

(Table 2) Out of these, the younger formation (28.59 km2) is

dis-tributed in the northeastern part of Kanyakumari coast and

Manakudi-Pillaithoppu coastal tracks, while the older formation

(23.32 km2) is in the northern part of Muttam coast and

middle-western part of Chinnavilai, Mandaikadu, Keezhkulam and

Mida-lam coastal areas Beach landforms such as sandy beaches, cusps,

ridges, berms, terraces, scarps, sandbars and sand spits are

dis-tributed along the nearshore region

This is result of swashing or backwashing of littoral sediments

due to action of waves, wind and littoral currents (Ahmad, 1972;

Sanil Kumar et al., 2006; Kaliraj et al., 2014) The total expanse of

this category is estimated around 6.38 km2 which is equal to

2.19% of the total area Among them, the sandy beaches

(1.16 km2) are extensively developed in the different parts of

coastal stretches including Manakudi, Chothavilai, Sanguthurai,

Periyakadu, Ganapathipuram, Muttam east, Kottilpadu, Midalam

and Inayamputhenthurai coast Beach terraces are gently sloping

features developed due to sediment deposition, typically bounded

by ridges and scarps on landward and seaward sides respectively

They are formed either from a pre-existing shoreline through

mar-ine abrasion or erosion or due to accumulations of sediments in the

low wave energy zones by emerging slightly as marine-built

ter-races Patches of beach terraces are distributed in various parts

of the study area especially between Muttam and

Rajakkaman-galam coastal stretches and the coverage of landform is restricted

to 0.43 km2(0.15%) However, enormous areas of the beaches were

deformed and modified due to the Tsunami waves on 26th

Decem-ber 2004 (Chandrasekar et al., 2006) Beach ridges are formed as

narrow and curve shaped features parallel to the shoreline

between Manakudi and Pillaithoppu stretches due to

swash-over-wash of sediments by the action of high-energy waves

(Cherian et al., 2012) Similarly, the narrow and undulating berms

and terraces are formed near Kasavanputhendurai (0.013 km2),

Sanguthurai (0.03 km2) and Ganapathipuram (0.035 km2) coastal

areas These are often altering to multi-faced forms due to

fluctu-ations in sediment accumulation through swash during monsoon

(Kaliraj et al., 2014) Cusps are commonly distributed landforms

in various parts of the study area for a length of 25.44 km due to

action of breaking waves at the surf zone Fig 4 shows the

sector-wise distribution of the coastal landforms along the study

area

The discontinuous cusps developed along the coastal segments

of the western parts near Manavalakurichi–Colachel (4.6 km) and

Keezhkulam–Inayamputhenthurai (1.2 km) are due to presence

of offshore rocky outcrops moderating the wave energy resulting

in deposition of sediments along the coast Scarps are wave-cut slope or miniature cliff on the seaward slope indicating erosional activities along the coast Vertical expressions of scarps vary dis-tinctly along the Rajakkamangalam coast (0.07 km2) and in the Chinnavilai-Manavalakurichi coastal stretches (0.13 km2) due to direct exposure of the coast to high-energy wave action resulting

in wave-cuts along the beaches from several centimeters to a

Table 2 Spatial distribution and area extent of the coastal landforms in the south west coast of Kanyakumari.

Sl no Coastal landforms Areal extent of landforms

Area (km 2

) Percentage of distribution (%) i) Marine origin

Total area of marine origin of landforms

81.07 27.76 ii) Fluvio-marine origin

Total area of fluvio-marine origin

of landforms

62.15 21.28 iii) Fluvial origin

31 Pediment (moderately weathered)

38.02 13.02

33 Structural hill and Inselberg 1.54 0.53

34 Wetland shallow/waterlogged area

Total area of fluvial origin of landforms

126.98 43.49 iv) Aeolian origin

Total area of aeolian origin of landforms

21.44 7.34 v) Coastal structures

Total area of coastal structures 0.36 0.12 Net total area of coastal landforms

distribution

few meters (Samsuddin et al., 1991) The sandbars (0.05 km2) are

formed on the river mouth of the estuaries near Thengapattinam

and Manakudi coastal areas, which is open to wave action and

cur-rents producing seasonal changes especially during monsoons

Shallow fluvio-marine landforms like salt marsh, tidal flat or

mud flats are associated with estuaries near Thengapattinam coast

(0.20 km2), Manakudi coast (2.61 km2) and Colachel coast

(0.8 km2) due to deposition of fine muddy sediments from river

discharge (Table 2) The backwater creeks (0.96 km2) are found

in the Colachel, Manavalakurichi, Midalam and

Rajakkaman-galamthurai coastal zones Significantly, the coastal uplands in

the Kanyakumari-Kovalam (0.22 km2), Muttam-Kadiyapattinam

(0.69 km2) and Colachel-Kurumpanai (0.86 km2) coastal areas are

made up of bedrocks overlaid with Terisand deposits which are

exposed in the surrounding coastal landforms In the Kanyakumari

coast, the upland is characterized by thickly layered sandstone

overlain by Teri sand deposits ranging in thickness from 8 to

56 m (Jayangondaperumal et al., 2012)

Similarly, the Muttam-Kadiyapattinam coastal tract has

pres-ence of sedimentary rocks in the upland varying in thickness from

4 to 73 m Colachel upland has notable outcrops of sandstone

com-posed of sand and boulder derived from the Teri sand materials

with a thickness of 6–34 m Rocky outcrops are also seen in the

off-shore of the Kanyakumari, Muttam, Colachel and

Inayamputhen-thurai coastal areas at the distance of 0.5–5 km from the coast

These are considered as remnants of headlands detached earlier

due to wave erosion and tectonic activities (Jayappa et al., 2006)

notches and attributed their presence to sea level changes and local and regional tectonic activities They have surmised that the notches represents the past sea level stands varying from 12 to

25 m above the MSL Notches of the southern coast equated to past stand of sea level hint at slow long-term uplift along the coastal tracts of southeast coast of Tamil Nadu In the study area, the sea-ward slope of the rocky shore comprises of wave-cut notches (0.38 km2) and wave-cut platforms (0.62 km2) increasing in their shape and size in the Kanyakumari, Cape Comorin, Kovalam, Mut-tam and Colachel coasts mainly due to the slumping of the rocky shore towards the sea by the undercutting action of the waves Sand dunes are formed as parabolic dune complexes with a height

of 2–4 m along the Kanyakumari-Kovalam (3.27 km2), Manakudi-Periyakadu (2.38 km2) and Manavalakurichi-Colachel (1.30 km2) coastal areas It is observed that the evolution of the landform reflects the prevailing coastal processes sculpturing their morphol-ogy and distribution Sector wise distribution of the coastal land-forms is shown inFig 3and the areal extents of the landforms are given inTable 2

4.2 Topographical change detection and assessment

‘The topographical change indicates vertical difference of sedi-ment load in the area providing insight into the morphological expressions to coastal processes over time (Gyasi-Agyei et al., 1995; Pavlopoulos et al., 2009; James et al., 2012) As explained earlier, topographical change analysis In this study is carried out using SRTM DEM and ASTER DEM estimating the vertical difference

Fig 4 Sector wise coastal landforms of the study area.

of sediment load for various landforms during the periods of 2000–

2011.Fig 5shows the topographical characteristics of the coastal

landforms using SRTM DEM for the year 2000, and the ASTER

DEM represents the same for the year 2011 The elevation of

land-forms represented SRTM DEM is 0–157 m AMSL (Fig 3A) Whereas,

the elevation range in the ASTER DEM depicts a reduction to 0–

138 m AMSL (Fig 3B) Morphodynamics of beaches through beach

profiling shows that the coastal zone prevailing with low wave

energy gets accreted with mass balance of sediment load up to

1–2 m in pre-monsoon and decreases to 0.5–1.0 m during rest of

periods (Hentry et al., 2012) High wave energy zones experience

severe erosion causing landward movement of shoreline due to

erosion, beach subsidence and wave run-up Moreover, artificial

structures along the coast such as groins, revetments, and jetties

pose hindrance to the natural hydrodynamic processes of waves

and currents causing severe erosion on the down-drift side

(Jayappa et al., 2006; Kaliraj et al., 2013) The low-laying landforms

such as creeks, salt marshes and backwater that are above/below

MSL in their topographical setting render an elevation of 7 to

+6 m with respect to MSL in 2000; however this has been modified

subsequently to 0–9 m during 2011 This could be due siltation

from the adjacent landforms by marine and fluvial activities

The beach landforms such as berms, cusps, ridges, scarps,

ter-races and sandy beaches have undergone relief reduction in terms

of elevation from 1–3 m to 1–2 m due to erosional processes of

wave, current and wind causing loss of sediments from the coast

(Cherian et al., 2012) Moreover, in the backshore region, relative heights of the sand dunes and foredunes have been reduced from 9–13 m to 6–11 m due to removal of surface dune layer by wind winnowing during monsoons Significant removal of sediments have taken place from the dune complexes due to the effect of the Tsunami occurred on 26th, December 2004 (Chandrasekar

et al., 2012) In the inland area, even the fluvial landforms such

as alluvial plains and flood plains have undergone reduction in ele-vation from 34–55 in 2000 to 34–38 m in 2011 Severe erosion due

to river and surface runoff during monsoon is considered responsi-ble the lost of landforms

4.3 Cross-shore profile change detection and assessment The extracted cross-shore profile analysis using the DEM data-sets provide information on geomorphic change of the various landforms in vertical scale (Zandbergen, 2008; Dawson and

cross-shore profile sets (3 profiles for each sector) were separately extracted using SRTM and ASTER DEM datasets for the four sectors namely Kanyakumari, Rajakkamangalam, Muttam and Thengapat-tinam at 5 km of interval In this analysis, the total areal extend (292 km2) of the DEM datasets is gridded into 60 segments with

a grid size of 5 5 km to extract the twelve cross-shore profile based on the concept of one profile flow across the five grids for estimation of vertical changes on different types of landforms

Fig 5 Topographical change (vertical difference of elevation) detection analysis using SRTM and ASTER DEM datasets for the periods of 2000–2011.