**A Systematic Review of Coastal Vulnerability Assessment Studies along Andhra Pradesh, India: A Critical Evaluation of Data Gathering, Risk Levels and Mitigation Strategies**

**Komali Kantamaneni 1,\*, N.N.V. Sudha Rani 2, Louis Rice 3, Koyel Sur 4,5, Menaha Thayaparan 6, Udayangani Kulatunga 6, Rajshree Rege 7, Komali Yenneti <sup>8</sup> and Luiza C. Campos <sup>9</sup>**


Received: 3 October 2018; Accepted: 6 February 2019; Published: 23 February 2019

**Abstract:** The establishment and alteration of any coastal feature is largely dependent upon complex hydrological and geomorphologic processes. Therefore, understanding hazard factors and threat risk level is crucial for mitigating risk in coastal zones. This study examines coastal vulnerability factors and their influence along the Coastal Andhra Pradesh (CAP) region in India. CAP has been exposed to frequent hydrological and meteorological hazards due to variations in the geographical, geological, and bathymetric characteristics. Despite substantial vulnerabilities, the risk to the coastline of Andhra Pradesh has not been rigorously evaluated. The current research systematically reviews the drivers and effects of hazards and vulnerabilities in CAP. Findings indicate that urban cities have a considerably higher risk of cyclones and floods due to their locations on the Bay of Bengal tectonic plate, the topology of this coastal region, and higher population density. The study revealed that the Coastal Vulnerability Index (CVI) data along CAP is mostly gathered using low-resolution satellite data and/or field observation surveys. The study further revealed that there are very few existing mitigation strategies developed or discussed within the obtained results. However, more accurate data gathering techniques for coastal vulnerability factors are available such as Unmanned Aerial Vehicles (UAVs): Air-borne and LiDAR sensors, which provide very high resolution data and low-cost accessibility to physically inaccessible places, making them suitable for vulnerability data collection in coastal locations. These findings are useful for stakeholders seeking to reduce or ameliorate the impact of coastal disasters and their impacts on the CAP economy, environment, and population. The study further helps to reduce the existing shortcomings in the assessment techniques used previously.

**Keywords:** coastal vulnerability; natural hazards; disasters; coastal Andhra Pradesh; Unmanned Aerial Vehicles

#### **1. Introduction**

Coastal environments are particularly vulnerable to the effects of disasters; partly due to denser urbanization and human populations and related economic activities such as agriculture, aquaculture, tourism, industries, trades and transportation in these locations [1–5]. Coastal areas tend to be relatively highly urbanized with higher concentrations of human populations [6,7]. Approximately 40% of the global human population lives within 60 km of the coast [8] and more than 25% of the human population in India lies within 50 km of the coast [9]. Natural processes, particularly those affected by climate change, combined with human activities, are becoming significant factors for coastal vulnerabilities. Risks at coastal regions in India are rapidly increasing; major threats for vulnerability are mean sea level rise, extreme events like flash floods, and cyclone induced storm surges [10–13]. Increased green-house gas emissions, predominantly from human activities, are predicted to expedite climate change and sea level rises that are already making coastal zones vulnerable in the 21st century [14–18]. Warming oceans lead to sea level rises, which in turn provide a source of thermal expansion of sea water, contributing further to intensified and more frequent cyclone-induced storm surges [19–22] addressed the potential climate change impacts owing to drought, floods and cyclone events with more intensity and asserted that India has very high risk levels (class 9 among risk class 0–10). As a result of the changes related to climate change, there are significant challenges for assessing coastal vulnerability and subsequent adaptation and mitigation strategies [23,24]. Systematic coastal vulnerability assessments [25–28] are essential for managing coastal threats. According to Indian Coastal Zone (CRZ) regulations, the buffer zones, also known as active zones, usually cover a region of 500 m from the shoreline; these zones are more usually focused on coastal vulnerability studies. However, coastal areas up to 100 km perpendicular from the shoreline should be considered as vulnerable areas for assessment studies. Coastal vulnerability resistance developments and strategies done to date are limited; they are only applicable to the coastal buffer zone (yet the effects of disasters impact on areas beyond this zone) and mostly concern coastal geomorphology issues but do not pertain to other extreme events such as cyclones and inundation [29–31].

Comprised of different geological, ecological, biological, urban and socio-economic features, the Indian coast is subject to varying degrees of exposure to multiple hazards. In the current scenario, the environmental stresses on coastal zones are increasing significantly in terms of social, physical and economic variables. Social parameters such as population density, drinking water demands owing to over-pumping of fresh-water, communication networks, roads with transportation, drainage, infrastructure, agriculture, aquaculture and industrialization are the prime factors that are providing harmful feedback loops to existing coastal vulnerability. Physical parameters such as sea level rise, slope, cyclones and storm surges are the natural contributing factors (or hazards) for increasing vulnerabilities along the coast. Sand mining is another crucial parameter in terms of coastal vulnerability along this coast. The increasing levels of sand mining along CAP will further result in ecosystem damage in terms of habitat, erosion, riverine system changes, and increased destruction due to extreme events. Therefore, coastal vulnerability assessment techniques require a multi-hazard assessment methodology dependent upon the location at a regional level integrated with global level parameters. Several studies of coastal vulnerability have been conducted at various geographical locations along the Indian coast using a variety of approaches to mapping and data collection methods. Kumar et al. (2006) [32] used mapping techniques to evaluate the potential vulnerability implications, due to sea level variations, along the Cochin coast. Kanakara et al. (2007) [33] used an integrated MIKE-21 model for the estimation of oil spill vulnerability for different locations along the Gulf of Katchchh by using land-use, land cover changes, shoreline changes, rates of erosion and accretion,

and sediment transport parameters during pre-monsoon, post-monsoon seasons. Chandrasekar et al. (2011) [34] studied coastal geomorphological vulnerability along the Tamil Nadu coast and revealed that intensive sand mining and coral mining in barrier coral islands are adversely changing beach morphology. Mani Murali et al. (2013) [35] considered a Physical Vulnerability Index (PVI), Social Vulnerability Index (SVI) and calibrated Coastal Vulnerability Index (CVI) for the Puducherry coast. Saxena et al. (2013) [36] developed a composite vulnerability index for coastal hazards along the Cuddlore coast, Tamil Nadu. The parameters considered for this study were erosion mapping, flood hazard mapping, and sea level rise combined together and then integrated with socio-economic influences. Appelquist et al. (2015) [37] proposed a new methodology (adapted from a method initially proposed by Appelquist (2013) [38] to assess and manage the Karnataka coast with a new multi-hazard technique known as the Coastal Hazard Wheel (CHW). The prime focus of this methodology is for screening the coastal area at local, regional and national levels of hazard, mainly in regions where available data is inadequate. These vulnerability assessment studies were mostly conducted using low resolution spatial data; and there were very few mitigation strategies evident. Therefore, it is necessary to adopt a more holistic and integrated approach to overcome these limitations. The concepts' vulnerability, hazard and risks are interrelated, but address distinct issues. According to Gilard (2016) [39], hazards are mostly natural phenomena, whereas vulnerability is associated with socioeconomic parameters affected by hazards and natural processes. For instance, when a vulnerable region is impacted by a hazard, there exists a potential risk that needs to be addressed. The risks are given a rating to estimate the vulnerability level (for example: high, medium, low). Therefore, when a particular coastal vulnerability is estimated, an integrated approach is required that can distinguish between these three concepts.

The main aim of this work is to review the existing coastal vulnerability assessment studies along CAP to identify the vulnerable locations with its main focus on CVI methodology approaches. A systematic literature review of CVI was carried out along the coastal Andhra Pradesh. The literature review for this region was organized according to each of the CVI parameters; including the data-collection technique, the magnitude of the threat, drivers for vulnerability, and the prevalence of mitigation strategies. Furthermore, the current study also identified gaps in knowledge and some recommendations for future research.

#### **2. Snapshot of Coastal Vulnerability Methodologies**

The primary technique used to monitor coasts for vulnerability assessments is spatial data with Geographic Information System (GIS). The method for gathering spatial GIS data is satellite imagery; this is due to its relative accessibility, economic affordability, and regular repetitive coverage. GIS-based spatial data is used in a number of different analytical methodologies, which include 'Decision Support Systems' (DSS) and index based techniques. DSS approaches include 'Community Vulnerability Assessment Tool' (CVAT), DINAS-coast, Dynamic Interactive Vulnerability Assessment (DIVA), Digital Shoreline Assessment System (DSAS) and vulnerability assessment [4]. Index-based techniques such as the Coastal Vulnerability Index (CVI) are also used widely across the world [28,40–50] 'Integrated Valuation of Ecosystem Services and Tradeoffs' (InVEST) is an open source software model that has a wide range of models to analyze a range of coastal vulnerabilities, including social, geographical, biological and economic factors [27]. Indian coasts have been studied mostly through the use of CVI methods for physical assessments [51–54] most of these studies required a certain amount of field data for the evaluation, due to limitations in the available spatial and temporal satellite data resolution [55] Remote-sensing approaches evident in the existing literature range from air-born to space-borne data gathering techniques; however, coastal vulnerability studies along the Indian coast were mostly restricted to the use of earth observing space-borne sensors [56–62]. Table 1 illustrates the range of parameters used across India to evaluate coastal vulnerability in diverse circumstances.


**Table 1.** Vulnerability assessment and parameters considered for different Indian geographic locations (modified from Rani et al. 2015). [3]

### **3. Study Area**

Coastal Andhra Pradesh (CAP) is located along the Eastern Indian coastline towards the Bay of Bengal at latitude 13◦24' N to 19◦54' N and longitude 80◦02' E to 86◦46' E (Figure 1) and is

approximately 966 km in length. CAP has a hot and humid climate and consists of nine coastal districts with an area of 95,442 km2, which is 60% of the total state area [66]. The Andhra Pradesh (Figure 1) coastline is the second largest shoreline after Gujarat in India. The population is 34 million, which is 70% of the total state population; the major cities are Vijayawada, Visakhapatnam, Guntur, Rajahmundry, Kakinada, Eluru, Nellore, and Ongole [66]. CAP has many hectares of fertile agricultural land due to the Krishna and Godavari deltas. Agriculture and fishing, which are the main fiscal activities in this region, are often disrupted by monsoonal variations, changes in climate, rapid changes in coastal geomorphology, local and sub-local elevations, and the rate of shoreline changes [67].

**Figure 1.** A spatial map of India with CAP indicated.

The rainfall is influenced by both South-West and North-East monsoons and the average annual rainfall along CAP is 1078.0 mm [68]. Andhra Pradesh, particularly coastal Andhra, is prone to cyclones, storm surges and floods. Moderate to high intensity cyclones occur every two to three years. Figure 2 shows the detailed spatial map for Andhra Pradesh coastal districts with the total number of cyclones occurring during the time period of 1951 to 2010. According to UNEP [6], more than 40% of the state is vulnerable to tropical storms, hurricane winds, severe floods and other associated natural hazards. The Godavari and Krishna River floods often contribute to serious disasters in the East and West Godavari and Krishna districts. The death toll and infrastructure damage due to these disasters is higher than other Indian coastal areas. According to the National Disaster Management Authority (NDMA) report (2016) [69], Andhra Pradesh coast is the second-most cyclone affected Indian region after Odisha and second-largest vulnerable region to floods after Kerala. The Ministry of Environment & Forests, Govt. of India (2012) [70] reported more frequent and intensified cyclonic activity and associated storm surges in CAP as a result of the impacts of climate change on air and sea-surface temperatures.

**Figure 2.** Total number of cyclones and population growth rate. Developed based upon the statistics of Revenue Disaster Management Andhra Pradesh (2017) [71].

#### **4. Methodology**

The current study is a systematic review of existing literature on coastal vulnerability assessments of the Indian coastline with emphasis on coastal Andhra Pradesh. The review systematically searched the following databases: ScienceDirect, Google Scholar, Web of Science (WoS), and Scopus to recognise the relevant literature using the keywords *coastal vulnerability, India, coastal Andhra Pradesh, erosion, GIS, spatial analysis*. Before finalising the existing studies to carry out the analysis, the current study performed a scoping exercise to recognise main themes, gaps and trends in the literature on coastal vulnerability studies. The review process comprised three main steps: searching (1), screening (2) finalization and analysis (3) (Figure 3). Initially, the first the review identified 125 articles and documents that met the search criteria. During the screening procedure, the current study applied

exclusion and inclusion criteria to the search results to categorize the selected studies. Accordingly, the current study appraised the article titles and abstracts to evaluate their relevance to the study and, consequently excluded 75 articles deemed to be irrelevant. At this stage, 50 articles were selected for the full-text assessment. However, after full text analysis, a further 25 more articles were excluded due to irrelevance. Accordingly, only 25 articles from 125 articles have been finalized for the analysis. The omitted papers discussed coastal vulnerability in general, but their emphasis was not on the Coastal Andhra Pradesh region. Furthermore, other excluded papers examined CVI procedures at some point, but from integrated models' viewpoints, which are not relevant to the current study. In addition, the study also explored various government websites such as AP State of Portal, Government of India National Disaster Management Authority (NDMA), Indian Metrological Department (IMD), Regional Metrological Centre (RMC) and the Mumbai and National Cyclone Risk Mitigation Project (NCRMP) for grey literature. This grey literature was searched and evaluated according to the same criteria used for the selection of journal articles and other associated documents. Accordingly, 32 of these reports have been included for analysis. In total, 57 documents regarding coastal vulnerability studies have been analyzed.

**Figure 3.** Methodological procedure.

*Coastal Vulnerability Index Formulations and Parameters*

Several CVI approaches, formulations and parameters are used to evaluate coastal vulnerability. Listed below are the main methodological frameworks and parameters. The first application of CVI (Equation (1)) was developed by Gornitz in 1990 with a focus on sea level rise, which considered the following parameters:


The principal formula for calculating CVI is given as follows:

$$CVI = \frac{\sqrt{a \times b \times c \times d \times e \times d \times e \times f \times g}}{7} \tag{1}$$

The above equation is the standard way to assess coastal vulnerability; however, it is not the only possible way to estimate the CVI quantitatively. Gornitz (1990) [72] proposed various ways to calculate the CVI, though the above equation is most widely used, as the geometric average is the most sensitive to individual parameters. Most of the studies along CAP used the above equation for vulnerability assessment. However, Rao et al. 2010 [73] used a different approach by using the summing of variables with prescribed weightage given to each parameter. For this assessment, five parameters were considered and the equation is as follows (Equation (2)):

$$
\mathbb{C}VI = 4a + 4c + 2b + f + e \tag{2}
$$

where the variables *a*, *c*, *b*, *f* and *e* have their usual meanings defined as above.

The majority of studies [32,50–53] focused on basic factors as major drivers for coastal vulnerability along the Indian coast. Krishnan et al. (2018) [74] proposed a cumulative vulnerability index (CuVI) framework to map the coastal vulnerability along the Maharashtra coast. The CuVI is a function of exposure, sensitivity and adaptive capacity and uses the exposure-index (EI), sensitivity-index (SI) and adaptive-capacity-index (ACI). It covers all physical, social, and economic variables cumulatively for the preparation of a spatial decision-making map. Table 2 sets out the data-gathering approaches (in terms of conventional and spatial data) used for various CVI assessments as well as others that can be used. Here, conventional data is defined as point level observations or in-situ data that is collected throughout the year, whereas spatial data is satellite/aircraft based. There are various social parameters such as population, population growth rate, transportation with road networks, mobile-television facilities for each home for hazard warning awareness, age, gender, etc. CVI analyses require the data to be classified into different risk ranges; these are illustrated in Table 3.

**Table 2.** Different types of data that can be used for CVI studies in India.



**Table 3.** Coastal Vulnerability Index ranges with their risk values considered for CAP. Source: [9,58].

#### **5. Results**

Based on the literature reviewed, several of the existing studies evaluated Indian coastal vulnerability. However, very few studies were performed along CAP and most of these studies were based on the conventional CVI technique developed by Gornitz et al. (1990) [72]; however, these studies differ from each other in the selection of parameters.

#### *5.1. Coastal Geomorphology*

The coastline of Andhra Pradesh is segregated into three distinct types: a rocky coast starting from North of Godavari delta, a vegetation coast in the Krishna-Godavari deltaic region, and a sandy coast from south of Krishna delta to Pulikat lake (SAC, ISRO report, 2011); a more detailed breakdown of land-use cover is shown in Figure 3, which revels that the entire CAP is predominantly agricultural. The geomorphology of the CAP is very diverse and economically important due to several features such as the Krishna-Godavari river delta, which plays a huge role in agriculture activities; developed/urbanized coastal areas along Visakhapatnam and Machilipatnam port; and ecologically sensitive places like mangroves and mudflats. Pramanik et al. (2016) [75] described the region as sensitive to muddy and soft sand and categorised the entire coast into two risk rates (higher and lower). Rao et al. (2010) [73] used IRS P6 AWiFS (The Advanced Wide Field Sensor) data to map geomorphology and its risk rankings, showing that the Krishna-Godavari deltaic regions are highly vulnerable due to the presence of mud flats, mangroves and the beach ridge complex. Overall, the mangroves and mudflat dominated areas of Kakinada, Mummidivaram, Avanigadda and Repalle were categorised as very high risk.

#### *5.2. Shoreline Change-Rate*

Shoreline change rate can be defined as the rate at which shore gets eroded or accreted due to wave-action, sea level-rise or other hazards and processes that affect the land. Although much of this is a natural process, added urbanization, infrastructure development and increased population can exacerbate the process. The CAP region is particularly prone to erosion due to its location and spatial distribution along the ocean. Factors such as sand mining, dredging and hard engineered protection for CAP are making it more susceptible to erosion. Almost all of the studies for the CAP to assess the shoreline erosion used Landsat data. For example, Basheer Ahammed et al. (2016) [76] used Landsat satellite data for the time-period of 1972–2015, and showed that most erosion was observed along the Krishna-Godavari delta and highest accretion was confined to the estuary outlet of the Krishna-Godavari deltaic region. The reason for this accretion could be mostly due to sediment transport from the estuary rather than that of coastal accretion, whereas sand mining could be one of the parameters for extensive erosion patterns. Vivek et al. (2016) [9] used the Landsat data for the period of 1973–2015 and found that some regions, such as Bheemunipatnam, central Chebrolu, Veerappa Konda and south of Tuni, experienced a (maximum) exposure rate of almost 10 meters per year. Rani et al. (2018) [67] used 20 years (1997–2017) of spatial data (Landsat 4 to Landsat OLI) as well as conventional data (such as PSMSL, Survey of India toposheets) and showed that an approximately 2.5 km shoreline shift, in terms of erosion and 1.82 km shift in terms of accretion, was observed along the Vizianagaram–Srikakulam coast. These studies are useful to measure the shoreline erosion statistically; however, there is a need to improve the spatial, as well as temporal resolutions, for more accurate assessment.

#### *5.3. Coastal Slope*

Coastal elevation is the degree of steepness of the region with respect to the ocean; its measurement helps to identify the vulnerability of the coast. Rao et al. (2010) [73] used SRTM DEM data to evaluate the coastal slope range and its risk rate. Basheer Ahammed et al. (2016) [76] extracted coastal slope data from GEBCO (General Bathymetric Chart of Oceans) to identify that the lower southern part of CAP has steeper slopes, which leads to higher vulnerability as well as other risks. The available literature revealed different forms of data for coastal slope and its vulnerability mapping: Rani et al. (2018) [67] used ASTER DEM (Advanced Spaceborne Thermal Emission and Reflection Radiometer Digital Elevation Model) data, while SRTM DEM was used by Pramanik et al. (2016) [75] (both SRTM and ASTER DEM have a resolution of 30m for the Indian region). These studies further noted that CAP steep slopes are at high risk due to shoreline erosion and inundation due to storms. One more parameter that can classify the risk along the coastal zone is bathymetry, particularly in relation to wave height. The higher the bathymetry, the higher will be the risk to the coast owing to sea level rise and surge height. Rani et al. (2018) [64] used SRTM DEM with Directorate General of Hydrocarbons (DGH). While Pramanik et al. (2016) [75] used DGH data and revealed that Mummidivaram and Avanigadda zones are under very high risk followed by Pithapuram, Kakinada, Amalapuram and Machilipatnam, lower risk rates were identified for Tuni, Razole and Narsapur.

### *5.4. Relative Sea Level Rise*

According to Unnikrishnan et al. (2007) [77], sea level changes can be due to two phenomena—one is global (mean sea level change) while the other is regional (extreme sea level changes). Therefore, sea level rise is a major phenomenon that combines global and regional processes that affects climate change induced coastal vulnerability along any coast. There are several ways to retrieve data for sea level change, for example from satellites and observations. Vivek et al. (2016) [9] brought together data from GLOSS (Global Sea Level Observing System) and found that the highest sea level rise was 5 cm/year along the Nellore coast and the lowest sea level rise of 0.8 cm/year was recorded in Pulicat coast. Pramanik et al. (2016) [75] used the IPCC global average sea level rise data and found 40% of CAP's population vulnerable to potential flooding by mid-century.

#### *5.5. Significant Wave Height*

Significant wave height, when associated with a tropical cyclone-induced storm surge, creates disaster at higher levels and CAP is highly prone to tropical cyclones. CAP wave height records from wave rider buoys were taken from the National Institute of Oceanography (NIO) to calculate the significant wave height [75]. They considered the significant wave height values of the pre- and post-monsoon seasons for 2013, which were found in the range of 0.5 m to 2.5 m. Rao et al. (2010) [73] used the MIKE-21 (Spectral Wave) model to measure the wave height from the European Center for Medium Range Weather Forecast (ECMWF) wind velocity component data. There are other more accurate ways to measure significant wave height, such as Wavewatch III, but this has not been utilized for the CAP region so far.

#### *5.6. Mean Tidal Range, Tropical Cyclone and Induced Storm Surge*

The tidal range is the vertical difference between high and low tide. For the CAP region, Rao et al. (2010) [73] used hydrographic chart data. According to the study, the tidal range for CAP varies between 0.7 to 1.4 m, which would be classified as a low vulnerability region. The influence of cyclones and the related induced storm surges are more prevalent along the East coast of India, particularly along Andhra Pradesh and Odisha [78]. Severe cyclones since 1977 are listed in Table 4, while Figure 4 shows land use land cover distribution along CAP. As shown in Figure 4, CAP predominantly suffered from agricultural and the cyclone-induced storm surge and its inundation, which resulted in socio-economic losses. There has been a reduction of risk and loss of life in CAP since 1977, due to improved early warning protection systems that give accurate predictions for cyclonic landfall; however, economic loss has been increasing (Table 4). Surge inundation damages the soil system through penetration of saltwater into the coastal upper aquifer and damages agricultural and biophysical ecosystems. Rao et al. (2010) [73] created a vulnerability map for CAP in the context of storm surges and cyclone winds and determined that Nellore district is at highest risk followed by Srikakulam and Visakhapatnam districts; the lowest vulnerable risk locations are located at Krishna and Guntur districts. Using Expert Decision Support System (EDSS), Matta et al. (2015) [79] concluded that existing CAP cyclone shelters can only accommodate 8% of the rural population, meaning that 92% of the rural populace is still at risk from cyclones. None of the CVI-based systems (that included other parameters) used tropical cyclones as a vulnerability parameter.

Tropical cyclone induced storm surge is a significant threat to CAP, but it has not been rigorously accounted for in CVI assessments previously. One exception is the research by Vivek et al. (2016) [9] who used tsunami run up as one of the parameters for CVI along CAP. Land subsidence, depletion of groundwater, and saltwater intrusion into groundwater are additional parameters that can be incorporated into coastal vulnerability assessments for better results. The resultant CVI categorized Nellore and West Godavari as highly vulnerable. In contrast, low vulnerability ranked areas were observed along the Visakhapatnam region [9,76]. The studies conducted by Vivek et al. (2016) [9] differed because of the inclusion of the parameter 'tsunami' into the CVI studies. The slight differences in the resultant maps from available literature can be accounted for by variations in the parameters and the parameter observation time. Rao et al. (2010) [73] produced a detailed range of risks along CAP, and the analysis shows that over 70% of CAP is vulnerable. Diverse studies evaluated CAP vulnerability by implementing various parameters and techniques under CVI formation; this is shown in Table 5.


**Table 4.** Severe cyclonic storm and their impact since 1977 along CAP.



*Water* **2019**, *11*, 393



#### **6. Discussion**

Coastal management along CAP is very important due to its agricultural productivity and socioeconomic significance. However, the importance of coastal zone management and work towards adaptation and mitigation strategies along CAP is still in its infancy unlike other southern states of India. Objective coastal vulnerability assessments are necessary [73] due to the influence of a range of parameters along the coast.

#### *6.1. CVI Analysis*

The current study revealed the importance of coastal zone risks, management, and the primary challenges associated with it. Coastal Andhra Pradesh is mostly vulnerable to cyclones with their induced storm surges, floods and coastal erosion [81,82]. The importance is evidenced by historical hazards like the devastating 1997 Andhra Pradesh cyclone, 1999 Orissa cyclone, 2004 tsunami or most recent 2017 Chennai and 2018 Kerala floods. However, cyclone hazard assessments were not included in CVI assessments for CAP. For the CAP region, CVI criteria included basic parameters like shoreline-change assessment, sea-level-rise, coastal elevation, coastal geomorphology, bathymetry, significant wave height. Most of them were missing the following factors: tropical cyclone, tsunami and storm-surge run-up. However, even if these additional parameters are included in the evaluation, the studies are further limited as a result of the exclusion of social and economic factors. Disasters are often associated with human activities and the resultant stresses. Saxena et al. (2013) [36] described that the magnitude and risk of coastal disasters due to hazards should be addressed with the sensitivity and resilience of the exposed population, rather than the intensity of natural hazard. The protection of an area is deemed vulnerable if the area is sufficiently important to economic, cultural and environmental (ecological and biological) aspects [3,59]. Therefore, the CAP region needs better understanding of the human and economic damage/loss associated with it and studies should focus on a more integrated approach that combines natural, anthropogenic and climate change-induced vulnerabilities.

Rao et al. 2010 [73] showed that the risk levels were very high at Penna and Krishna-Godavari delta regions. These are very low-lying and almost flat areas with mudflats, mangrove swamps, and lagoons/backwaters. Whereas, a study by Basheer Ahammed et al. 2016 [76] showed that the Nellore region and Narsapur area are highly vulnerable. The reason for this difference in the risk rate evaluation can be the difference in the years of data as well as the approach. The CVI assessment by Rao et al. 2010 [69] was performed using a summation of parameters with weightage given to each factor; further SLR was considered as an indicator to give weightages to the parameters; whereas, for Ahammad et al. 2016, CVI analysis was based on the basic Gornitz 1998 formula. Rao et al. 2008, 2010 [58,73] provided the vulnerability maps for the storm surges and cyclone wind effect along CAP. Surge inundation is a very serious issue, as CAP has a history of inundation that lead to huge loss of life and property during the 1977 cyclone. His study showed that the East Godavari to Guntur stretch and Southern Nellore coast are under the widest surge inundation, while wind speeds are highest in Visakhapatnam, though the entire CAP stretch is not far behind. Cyclone vulnerability and risk analysis performed by Matta et al. 2015 [80] studied the population and land region exposed to cyclone and wind speed during the cyclones. These Mandals level maps revealed that out of 430 mandals, 190 are highly exposed to cyclone winds with speeds of 235 km/h. Such studies can be very useful for policy makers to mitigate future damage at a social level. Given the importance of cyclones and their influences on the CAP region, there exists a large gap in CVI estimations with cyclones as there were very few studies reported in this context.

Figure 5 provides some of the district-wise social parameters along CAP. From the figure, it is notable that almost all the districts are equally exposed to socioeconomic parameters. Nevertheless, households without proper drainage (West Godavari district) are at a higher risk of cyclone-induced storm surges and their inundation. Drinking water facility is not fully covered in any of the districts as most of the coastal villages depend upon wells. Additionally, over-pumping of well water along the coastal regions might further lead to salt water intrusion into ground water and land subsidence. Population density accelerates the vulnerability risk rate further. In terms of population density per square km, the highest risk was observed for Krishna district, followed by East Godavari, Srikakulam and Guntur districts.

#### *6.2. Mitigation Strategies and Approaches*

Andhra Pradesh coastal management has developed some adaptive strategies for stresses such as provision of cyclone shelters, disaster response force for responsive activities, cyclone and tsunami warning centers [83]; likewise, protection works for rivers and seacoasts were implemented to improve resilience to flooding (National Disaster Management Authority, 2016 [69]. Mitigation of coastal land loss due to erosion is mostly through hard engineering interventions such as the construction of groynes, dykes and seacoasts (Bhattacharyya et al. 2016 [84]. However, the literature revealed no such mitigation strategies employed over the CAP region, instead most focused on the rate of vulnerability with respect to the parameters considered. Nevertheless, these engineering structures can help to reduce the erosion in the short-term due to longshore drift; however, in the longer-term, these engineering interventions may increase erosion rates further down the coast. Furthermore, maintenance of these artificial engineering structures is relatively expensive [85]. Therefore, a more effective and more affordable alternative might be required, such as the use of ecosystem-based green-infrastructure solutions, to improve resiliency and to reduce the impact of some coastal activities such as sand mining.

#### *6.3. Data Gathering Techniques*

The most common weakness of the many studies of hazards on the Indian coast is the relatively poor resolution of data. The major focus of existing research has mostly used data gathered from field observations or satellites. CAP, in particular has less data gathered compared to states like West Bengal, Odisha, Tamil Nadu, Kerala and Karnataka. The studies focusing on mitigation and adaptation strategies to CAP hazards are very few to none. Furthermore, temporal variations of coastal processes vary from a daily time period to many years at a climatic level; however, the current data gathering techniques do not accurately capture this temporality sufficiently. At a global level, there have been significant improvements in approaches to data gathering since the majority of these studies were undertaken, particularly using new digital technologies. There is a need to study CAP with better data and constant observation; emerging technologies such as Unmanned Aerial Vehicles (UAVs) with airborne and LiDAR sensors can help to achieve a better understanding of CAP. Such digital technologies can provide spatial data at a much higher level of resolution than existing approaches, allowing data specificity at a much more local level. Moreover, the cost of these technologies has reduced recently and is now a cost-effective means for gathering data.

#### **7. Conclusions**

The current study research aims to generate up-to-date knowledge for use in the development and implementation of local, regional and national policies, and disaster and coastal risk management procedures. The current study systematically reviews existing literature to examine the occurrence, magnitude and impacts of the natural hazards affecting the Indian coast with special focus on Coastal Andhra Pradesh. Data collection methods and data sources are critically appraised as well as existing mitigation strategies proposed as a result of the analyses. The most recurrent natural hazards within the coastal Andhra Pradesh region are tropical cyclones, coastal inundation, monsoon-floods along the rivers and torrential rains. The present study highlights that, due to its geographical location on the Bay of Bengal Tectonic Plate, Coastal Andhra Pradesh is highly prone to natural coastal disasters. This study also distinguishes the probable long-term influence of sea level rise due to human-induced global warming and rapid urbanization on Coastal Andhra Pradesh in specific areas. The predicted sea level rise and temperatures are likely to result in the loss of residential and agricultural land, an upsurge in salinity and the decline of water quality, especially in agricultural fields, with predominantly negative impacts on the economy and population of coastal Andhra Pradesh.

The review reveals that the majority of existing data is collected from either satellite data or field observations that are very low in resolution, and are not continuous. This research can be used as a primary assessment for the design of a resiliency assessment framework for natural disaster management for the Andhra Pradesh region, and beyond. This resiliency assessment framework should play a vital role in decision-making within the reduction of coastal area vulnerability to natural hazards such as cyclones and floods, and also might improve its resilience capacity. Its application is likely to enable positive results regarding effective disaster management and control, by reducing harm to economies and safeguarding the security of local and regional communities. The current review of the coastal vulnerability of coastal Andhra Pradesh will be of interest to researchers, coastal engineers, environmentalists and government officials, who want to evaluate the prospective consequences of multiple disasters for coastal geographical areas across the world. The study shows that very little has been done to study the overall coastal vulnerability along CAP with only a few of the CVI parameters being accounted for, and based on relatively low resolution data. More research is required to observe, measure, and assess adaptation and mitigation measures in this area to improve resiliency.

**Author Contributions:** K.K., N.N.V.S.R. and L.R. designed the study. K.K. designed the methodology. K.K., N.N.V.S.R. and L.C.C. wrote the paper; L.R., K.S., M.T., U.K., K.Y and R.R. modified the manuscript and advised on physical and socio-economic coastal vulnerability studies.

**Funding:** This research was funded by the GCRF (Global Challenges Research Fund), United Kingdom.

**Acknowledgments:** Particular acknowledgement is made to Muhammad Kamran Naeem (Solent University) for his assistance in generating the pie chart using MetLab.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
