1 Introduction

Cyclones are often regarded as the world’s most destructive and deadly natural catastrophes, claiming thousands of lives every year (Bakkensen and Mendelsohn 2019). In addition to the loss of life, cyclones often cause property and infrastructural damage and destruction of the environment. On average, a tropical cyclone (TC) passes over the Bay of Bengal (BoB) three to four times every year, making the BoB one of the world's most cyclone-prone regions (Kuttippurath et al. 2022). Being a semi-enclosed basin, the majority of tropical cyclones that form here make landfall, giving BoB TCs a hugely disproportionate social significance in comparison to the tiny number of them that occur overall (Balaguru et al. 2014). TCs in the BoB are more common during the post-monsoon months of October to December (primary TC peak season) and the pre-monsoon months of April to June (secondary TC peak season). During these months, the sea surface temperature (SST) is rather warm (28–30°C), the atmosphere is thermodynamically unstable, and the tropospheric wind shears are minimal, which are conducive to the development of TCs (Mcphaden et al. 2009; Girishkumar and Ravichandran 2012). In the past three decades, a growing trend of post-monsoon cyclones has been seen, although the trend in the pre-monsoon season is still uncertain (Sahoo and Bhaskaran 2016). The genesis of most depressions that create cyclones in the BoB occurs in the southeast (SE) Bay of Bengal (the Andaman Sea).

A group of islands located in the BoB between 92°–94°E and 6°–14°N, the ANI, are home to a diverse range of biodiversity, various fauna and flora. A total of 352 islands, 220 islets, and rocks make up the archipelago, which spans about 800 km in length from north to south and has a coastline of 1962 km (Rajan 2013). Cyclonic disturbances have traversed the ANI numerous times due to its proximity to the cyclogenesis area in the BoB. There have been times in the past when severe cyclones have had an influence on the Andaman Islands. In the 20th century, severe cyclones passed through or passed near the Andaman Islands in the years 1911, 1914, 1916, 1921, 1961, 1976, and 1989 (Kumar et al. 2008). The majority of the island group's people have settled along the coast and in the vicinity of streams to be closer to their places of employment and recreational activities. Because of the rising number of coastal villages and tourist destinations in recent decades, they have become increasingly exposed to substantial wind speeds, high waves, and storm-driven surges. As a result, it is essential to understand and quantify the risk posed by these islands during these large-scale events. This will assist administrators and policymakers in developing disaster risk reduction plans and strategies, which will ultimately support the people of these islands in safeguarding themselves from long term damage.

Although several studies have been conducted to understand and quantify the vulnerability and risk associated with tropical cyclones on the Indian mainland (Sharma and Patwardhan 2008; Sharma et al. 2009; Kumar et al. 2010; Rao et al. 2013a, b; Balaguru et al. 2014; Sahoo and Bhaskaran 2018; Basheer Ahammed and Pandey 2019; Murty et al. 2020; Mondal et al. 2021), there are no studies associated with risk assessment of tropical cyclones over the Andaman Islands. This situation has motivated the current research to develop and quantify a comprehensive risk assessment for the Port Blair region of the Andaman Islands. Port Blair is the capital of these islands and the most populated area in the islands, as well as the central hub for all tourism-related and other development activities which take place in the region, thus making this area more sensitive to disasters and hazards than any other location on the islands. The area of study is around 184 km2 (figure 1), which has a mean annual temperature of 30.1°C and mean annual precipitation is 130.8 mm. Most of the population of the study area is concentrated on the municipal boundaries of Port Blair (Port Blair Municipal Council (PBMC)) due to better availability of resources, hospitals and other recreational centres and the areas in the outskirts of the PBMC are scarcely populated.

Figure 1
figure 1

Study Area map of Port Blair, Andaman Islands.

Full size image

The primary aim of this research is to provide a comprehensive and concrete tropical cyclone risk map for Andaman Island’s Port Blair area by combining vulnerability, exposure, hazard, and mitigation capability using numerical modelling and geographic information technologies (GIS). The following objectives contribute to the study’s accomplishment: (1) mapping the research region’s vulnerability and exposure, as well as its hazard and mitigation capacity. (2) Create a risk map for tropical cyclone effects, including vulnerability and exposure, hazard, and mitigation capability, using an AHP-based multi-criteria decision-making approach.

Cyclone Lehar was considered in this research, which is categorised as a very severe cyclonic storm (VSCS). Lehar originated as a depression in the southeast BoB on 23rd November 2013. On 24th November 2014, it was converted to a cyclonic storm (CS) category. On 25th November 2013, at around 0000 UTC, this CS passed the Andaman and Nicobar Islands near Port Blair, causing significant damage. It entered the Andaman Sea on the morning of 25th November 2013 and travelled west–northwest before strengthening into a VSCS on the morning of 26th November 2013. While it continued to move west–northwest, it deteriorated rapidly in the afternoon of 27th November and passed into Andhra Pradesh near Machilipatnam as a depression at about 0830 UTC on 28 November. Since 1989, this has been the only cyclone to pass over Port Blair (Kotal et al. 2013). As a result, we considered the features of Cyclone Lehar while creating the risk, vulnerability, and exposure criteria used in this research. Figure 2 depicts the track of cyclone Lehar.

Figure 2
figure 2

Track of the cyclone Lehar and its proximity to Andaman Islands.

Full size image

2 Materials and methods

The present research used a multi-criteria assessment method based on AHP that utilised GIS and numerical modelling tools. Saaty (1990) proposed the analytical hierarchical process (AHP) as a multi-criteria decision-making process. AHP was chosen for this research owing to its ability to effectively and comprehensively combine and aggregate multi-criteria functions (Mani Murali et al. 2013; Rao et al. 2013a, b). A number of risk assessment equations and methodologies were found from previous literature studies (Sharma and Patwardhan 2008; Hoque et al. 2018, 2019; Sahoo and Bhaskaran 2018) and for the present study, the risk associated with a tropical cyclone was represented using equation (1).

$${\text{Risk}} = {\text{hazard}} \times {\text{vulnerability}} \times {\text{exposure}}/{\text{mitigation capacity}}.$$
(1)

The step-by-step method adopted in the present study is illustrated in the flow chart (figure 3).

Figure 3
figure 3

The methodological flow chart.

Full size image

2.1 Data and sources

In order to quantify the risk of the study area, a variety of data sources have been combined in the current study. The simulation of storm surge heights and significant wave heights related to the cyclone was carried out using the MIKE 21 HD and SW modules. The details of cyclone shelters in the study area were spatially integrated using information obtained from the Andaman and Nicobar Disaster Management Authority website (http://ddm.and.nic.in) and through field observations. Similarly, details of hospitals in the study area were gathered through field observations and information obtained from the official website of the Andaman and Nicobar Administration (https://dhs.andaman.gov.in). Table 1 describes the datasets and the sources of the data.

Table 1 Data types and their sources used for the study.
Full size table

2.2 Risk assessment criteria

The risk map’s development includes numerical modelling simulations of wave heights and storm surge heights, as well as their spatial integration with all other criteria layers in a GIS environment. As a result, 13 criteria layers for the risk components of vulnerability and exposure, hazard, and mitigation capacity have been created. All the raster layers were prepared and processed with Arc GIS 10.4 software with a resolution of 15 × 15 m. The two hazard components, significant wave heights for developing the wave influencing factor during the cyclone and storm surge heights for coastal flooding assessment, were modelled using MIKE 21 SW and HD modules, respectively, as previously mentioned, and the process will be explained in-depth in the subsequent section.

2.2.1 Vulnerability and exposure mapping and their criteria

A large number of factors determine the vulnerability and exposure of a certain place as a result of the diverse physiological, topographical, and socio-economic conditions that exist in that region at any given time (Hoque et al. 2018; Sahoo and Bhaskaran 2018; Basheer Ahammed and Pandey 2019). We chose six parameters as vulnerability and exposure criteria for the current study, which includes four physical parameters, viz., elevation, slope, proximity to cyclone track, and proximity to coastline of the study area and two socio-economic parameters, viz., land use and land cover (LULC) and population of the study area.

Elevation and slope are extremely crucial in defining the vulnerability of a location (Hoque et al. 2018). The location with considerably low elevation and slope is regarded as extremely vulnerable, whereas places with higher elevation and slope are regarded as low (Rao et al. 2013a, b). For the present study, we have adopted ALOS PULSAR DEM of 12.5 m resolution for creating the Digital Elevation Model (DEM) and slope for the study area (figure 4a, b). The following key components in the cyclone risk assessment are the proximity to coastline and proximity to cyclone track. Hoque et al. (2018, 2019) have successfully incorporated these two parameters for the vulnerability analysis of tropical cyclones in Bangladesh. People living close to coastal regions are more susceptible to cyclones since the cyclonic winds together with high intense precipitation can inflict catastrophic damage to people’s life, property and infrastructure. The coastline was delineated by digitising the shoreline, and multiple ring buffering technique was performed in ArcGIS. Thus we have created the criteria for the proximity to coastline (figure 4c). The proximity to cyclone track data was generated from the best track data published by IMD for cyclone Lehar and was spatially integrated using ArcGIS (figure 4d). The rationale behind assessing the proximity to the cyclone track is that there is likely greater destruction than that could occur in the places where the cyclone is actually crossing.

Figure 4
figure 4

(a) Digital elevation model, (b) slope, (c) proximity to coastline, (d) proximity to cyclone track, (e) LULC map, and (f) population map.

Full size image

The other two parameters, the LULC and the population of the study area are also vital in determining the vulnerability of the region (Sharma et al. 2009; Saravanan et al. 2018). The LULC map was created from Landsat 8 satellite images with a resolution of 30 m (figure 4e). The radiometric, geometric, and atmospheric modifications of the satellite pictures were conducted using the QGIS Version 3.14 software plugin. The eleven identified LULC categories were classified using a hybrid categorisation, which included agriculture, airport, barren land, deforestation, forest, mangrove, plantation, quarry, settlement, swamp, and water bodies. Initially, an unsupervised clustering approach was used to find the likely LULC classes, followed by the use of maximum likelihood algorithm for supervised classification (Kumar et al. 2008). The population data for the study were derived from the 2011 Census of India report and the map was prepared by spatially integrating the census data using ArcGIS (figure 4f).

2.2.2 Hazard mapping and criteria

The most important aspect of creating a hazard map is identifying the hazards associated with disasters, which are defined as any geophysical, hydrometeorological, or biological events characterised by a significant release of energy that poses a threat to human life and can cause significant damage to property and the environment (Faulkner and Ball 2007). The first stage of this research was to identify the potential risks associated with cyclone Lehar as it passed across the islands, and after a few literature searches, four key hazard components were identified as follows: (i) coastal flooding caused by the cyclone’s storm surge; (ii) wave influencing factor due to significant wave heights generated by the cyclonic wind and pressure fields, which pose a significant threat to the coastline; (iii) high-speed cyclonic wind, which can cause significant damage to people and the environment; and (iv) cyclonic precipitation intensity, which also poses a significant threat in the study area.

Storm surge and flooding of coastal areas caused by storm surges are two primary hazards linked to cyclones. In this work, we simulated storm surge using a numerical model to simulate a storm surge model utilising the MIKE 21 Hydrodynamic (HD) model. MIKE 21 HD is a generic modelling system for tidal hydrodynamics, wind-generated currents, storm surges, and tsunami simulations supplied by MIKE 21’s hydrodynamic module. The modelling system covers two horizontal dimensions and solves spatially and temporally variable equations of continuity and momentum conservation. We computed the surge height associated with cyclone Lehar at a nearshore point at Port Blair at 92.74920E, 11.642980N (figure 5a) and found a maximum surge of 0.34 m at the site. It was observed that at the time of the landfall of the cyclone on Port Blair, a low tide condition prevailed (at 0000 hours on 25th November 2013). We took this data to the Digital Elevation Model and using the raster calculation tool, we added the surge values to the raster layer, resulting in coastal flooding map (figure 5b). The areas affected by coastal flooding are considered as high, and the areas without the effect of the coastal flooding are considered as very low for the ranking and from the coastal flooding map, it is understood that with a surge of 0.34 m, 5% of the study area had experienced the impact of coastal flooding.

Figure 5
figure 5

(a) Storm surge heights at 92.7492E, 11.64298N. (b) Areas affected by coastal flooding due to storm surge.

Full size image

Another hazard linked with cyclones is the high waves created by the wind and pressure fields associated with them. Significant wave heights are the average height (trough to crest) of one-third of the highest waves valid for the stated 12-hr period. Mean significant wave heights are utilised as a proxy to measure wave energy which affects the coastal sediment movement. In order to determine the significant wave heights related to cyclone Lehar, we have simulated a wave model using MIKE 21 SW Module, which is a standard modelling tool provided in the MIKE 21 Modelling environment. The model was driven using the pressure and wind fields of cyclone Lehar. Figure 6(a) shows the spatial distribution of significant wave heights near Port Blair on 25th November 2013 at 0000 hours and this information was taken into ArcGIS for ranking by the process of georeferencing. It was spatially integrated with DEM and the influence of these wave heights was calculated using the raster calculator tool; finally, the wave influencing factor map was created (figure 6b). So, for generating this map, information from the numerical model was linked with the DEM in order to have a better idea of how these enormous waves created during a cyclone would affect the coastal regions with varied heights. The wave influencing factor map reveals that about 13% of the study area was affected due to these high waves generated during the cyclone. Here the locations which are impacted by the significant wave heights are given a very high ranking and the places which are not influenced by the wave heights are provided with a very low ranking. The cyclonic wind fields generated by cyclone Lehar wreaked havoc in the study region. As a result of the reported incidents of trees falling, vegetation destruction, and damage to infrastructure and communication networks, the wind speeds associated with cyclone Lehar were considered as the next hazard component in this study. Wind data from IMD’s Best Track Data was used to simulate the wind speeds associated with cyclone Lehar using the MIKE 21 cyclone generation tool, which is a tool within the MIKE 21 environment that simulates the pressure and cyclonic wind speeds. Figure 7(a) depicts the wind fields around the Andaman Islands during cyclone Lehar’s passage on 25th November 2013 at 0000 hours. Wind speeds generated with the MIKE 21 cyclone generation tool were processed and interpolated into the study area using ArcGIS’s inverted distance weighted (IDW) technique. Thus the wind speed map was created (figure 7b).

Figure 6
figure 6

(a) Spatial distribution of significant wave heights around Port Blair at 25th November 2013, 0000 hours. (b) Wave height influencing factor map.

Full size image
Figure 7
figure 7

(a) The wind speed of cyclone Lehar around Andaman Islands on 25th November 2013, (b) cyclone wind map of the study area, and (c) the average precipitation intensity map of the study area.

Full size image

The next hazard identified in the study area is the intensity of precipitation during cyclone days. Average precipitation data from 23 November 2013 to 26 November 2013 was retrieved from the Tropical Rainfall Measuring Mission (TRMM) database, and this data was further processed and integrated into the study area using inverse distance weighted technique (IDW) in ArcGIS. Figure 7(c) depicts the study area’s average precipitation map on cyclonic days.

2.2.3 Mitigation capacity criteria

Mitigation capacity refers to the community’s capacity to cope with any hazards or disasters that may have possible impacts on them with their own available resources. There are so many strategies available for disaster mitigation, especially in coastal communities and can be classified mainly into three types: (i) protection, (ii) accommodation and (iii) retreat. These measures include both structural and non-structural approaches.

For the present study, we have identified three mitigation capacity criteria (i.e., cyclone shelters, hospital facilities and coastal vegetation).

Cyclone shelters are really important for enhancing the mitigation capacity of the community in cyclone prone areas since they provide immediate relief for the people in the immediate hours of disaster response and protect them from the immediate effect of storm surges and associated coastal flooding due to cyclones (Mallick and Rahman 2013). The accessibility to hospitals and other healthcare facilities is regarded as one of the most important features of the mitigation capacity of a region (Ali et al. 2020). The information regarding cyclone shelters were collected from directorate of disaster management, Andaman and Nicobar Islands and the data of the hospitals from Directorate of Health services, ANI. These two features were spatially integrated by using Euclidian distance method in ArcGIS and the maps are produced (figure 8a, b). These two criteria define people who have less distance to cyclonic shelters and hospitals as very less vulnerable and who has more distance to both of them as very highly vulnerable.

Figure 8
figure 8

(a) Distance to cyclone shelter map. (b) Distance to health facilities map and (c) NDVI map.

Full size image

The next key feature in determining the mitigation capacity of an area is the vegetation cover; the Normalised Difference Vegetation Index (NDVI) determines the area under vegetation. The calculated values for NDVI should range from –1 to 1. For the areas with highest green cover, the NDVI values will be closer to 1 (i.e., 0.9, 0.8, etc.) and for those areas with less green cover will have values below 0 to –1. The Landsat 8 imagery was used for the calculation of NDVI in the present study by utilising the red and near infrared band of the satellite imagery (Band 4 and 5). The NDVI was calculated by using the equation

$${\text{NDVI}} = (({\text{NIR}}{-}{\text{VIS}}))/(({\text{NIR}} + {\text{VIS}})),$$
(2)

where NIR is the near infra-red and VIS is the visible red band.

The idea behind taking NDVI as a criterion for the mitigation capacity is that the vegetation acts as a solid barrier for protecting people and infrastructure from the high impact cyclone winds and waves (Paul and Rahman 2006). This particular idea of integrating NDVI for creating coastal vegetation map was successfully done by Ali et al. (2020). The NDVI map for coastal vegetation is given in figure 8(c).

2.3 Ranking and standardisation of criteria layers

Risk ratings were assigned and ranked for each alternative of spatial criteria on a scale of 5 points, where 1 indicates very low risk and 5 indicates very high risk (table 2). A 15-m resolution was created for each criterion, converting them from vector to raster for performing the weighted overlay technique. A pair-wise comparison matrix was created according to Analytical Hierarchy Process (AHP), a decision-making algorithm developed by Saaty (1990). Out of 13 cyclone criteria, rank values were given on a scale ranging from 1 to 9 on the basis of their relative importance. In the matrix, the rank of one sub-criteria is reciprocal (i.e., 1/2 to 1/9) (Saaty 2008) to its inverse comparison (table 3). As mentioned earlier, the AHP technique was employed for weighing the selected criteria and is the most effective and widely used multi-criteria decision-making tool and has been employed mostly in coastal and cyclone risk assessment studies (Paul and Rahman 2006; Hoque et al. 2018, 2019). The pair-wise comparison matrix was developed for each risk component (vulnerability and exposure, hazard and mitigation capacity) with the help of six experts in the field of disaster risk management and a user following the ‘scale of relative importance’ (table 3). All of the experts selected had a considerable experience in coastal disaster management and disaster risk reduction. The consistency ratio (CR) was calculated to justify the consistency of comparison assigned by the experts and the user. The comparison is considered at the accepted level if the CR value is equal or < 0.1. The CR was calculated by using the following equation:

$${\text{CR}} = {\text{Consistency index }}\left( {{\text{CI}}} \right)/{\text{Random index }}\left( {{\text{RI}}} \right),$$
(3)

where the CI is calculated as:

$${\text{CI}} = \left( {\lambda \_\max {-}n} \right)/n{-}1,$$
(4)

where λ_max is the largest eigenvalue of the matrix and n is the matrix order.

Table 2 Ranking of the cyclone criteria based on their contribution to cyclone risk.
Full size table
Table 3 Numerical scale of importance (source: Saaty 1990).
Full size table

The random index (RI) indicates the randomly generated average consistency index (CI). Table 4 shows the RI for different matrix classes. Table 5 shows the criteria weights and consistency ratio from the pair-wise comparison matrix.

Table 4 Random index table.
Full size table
Table 5 Criteria weights and the consistency ratios from the pairwise comparison matrix.
Full size table

3 Results and discussion

3.1 Vulnerability, exposure, hazard and mitigation capacity mapping

The vulnerability and exposure map generated (figure 9a) indicates that 14% of the study area is classified as low vulnerable, 48% as medium vulnerable, 33% as high vulnerable, and 5% as very high vulnerable. The very vulnerable regions are mostly concentrated in Junglighat, the study area’s most heavily inhabited region. Due to their low elevation, moderate slopes, and land use pattern, these sites are among the most vulnerable in the study area. The research area's high vulnerable areas, which comprise low elevated regions, are located along the eastern and western coastal regions. The primary factor that contributed to their vulnerability to cyclones was their low elevation and very shallow slopes. Even though the population was sparse in this region, their proximity to the sea and to the cyclonic track made them very susceptible. The bulk of the research region falls into the category of medium vulnerability, which may be ascribed to the study area’s relatively high elevation and slope, greater distance from the coastline, and greater distance from the cyclone track. The low vulnerable zones include the cliff sections of Port Blair and the Chidiyattappu region, which are less vulnerable and these can be attributed to their relatively high elevation and dense vegetation cover.

Figure 9
figure 9

(a) Vulnerability and exposure map, (b) hazard map, (c) mitigation capacity map, and (d) risk map.

Full size image

The intense rainfall, coastal inundation of land as a result of coastal flooding, and significant wave impacts were identified and thus, the hazard map was produced (figure 9b). Four percent of the study area was classified as very high hazard zones, which include the Junglighat and Shipighat regions of the study area. The reason for these can be attributed to the intense rainfall, coastal inundation of land as a result of coastal flooding, and significant wave impacts. The bulk of the research region falls within the high-hazard category (76%), which covers the vast majority of the area under consideration. The strong rainfall during the cyclone, the high cyclonic wind speeds, the coastal flooding, and the wave effects, particularly in the eastern coastal region, caused these locations to fall into the high hazard category. The remaining area (20%) is classified as medium hazard zones because it received comparably less precipitation during the cyclonic period, as well as less damage from coastal floods and waves. This makes these areas medium hazardous.

The mitigation map produced for the study area (figure 9c) reveals that 45% of the area has a very low mitigation capacity which is concentrated on the central and southern sides of the study area. The reason for this is the fact that almost all the health facilities and shelters are concentrated in the city region of Port Blair, which is on the northern side of the study area. Access to health infrastructure and cyclone shelters is very far for the people living in this very less zone. Thirty percent of the study area comes in has medium mitigation capacity, 10% has high mitigation capacity and the rest 15% has a very high mitigation capacity, this is because these areas are located very near to Port Blair city and have easy access to health and cyclone shelters.

3.2 Risk mapping

The overall risk map of the study area is depicted in figure 9(d). The risk map reveals that 13% of the study area is under a very low-risk zone, i.e., the dangers caused by a tropical cyclone in this area are very less.

Twentyone percent falls under the category of low-risk zone followed by 22% of the medium risk zone. A considerable amount of the study area (34%) falls under the category of high-risk zone and 10% of the study area comes under very high-risk zone.

4 Limitations of the study

The present research included a number of limitations that may have influenced the findings. Although several risk assessment criteria were used, lack of high-quality and up-to-date data was a constraint. For example, high-resolution topographic data such as Light Detection and Ranging (LiDar) were necessary for a more accurate risk assessment; we generated the DEM using 12-m resolution ALOS pulsar data. The LULC was categorised using readily accessible Landsat data, however, it might have been classified more accurately with high-resolution data. Another problem with the research was that there were no recent demographic statistics for the study area because India's last census was done in 2011.

5 Conclusion

The cyclone risk map for the Port Blair area of the Andaman and Nicobar Islands was created in this research utilising numerical modelling methods and multi-criteria geospatial techniques. The risk was quantified by taking into account the study area’s vulnerability and exposure, as well as its hazard and mitigation capability. The vulnerability of a region, defined as its exposure to a particular hazard, was delaminated from the current study; the study area’s overall low elevation, gentle slopes, proximity to the coast, and, in fact, the track of the cyclone, which had a significant influence due to the study area’s compact size, all contributed to the study area’s vulnerability and exposure to cyclones. All of the parameters used to determine vulnerability were appropriate and properly defined. The proximity to the coastline has contributed around 37% of the study area’s vulnerability and exposure, followed by the population (19%), demonstrating that the region’s vulnerability is primarily determined by cyclonic characteristics and demographic patterns. The remaining vulnerability and exposure factors, namely elevation, slope, and closeness to the shore, are fundamental to the study region and do not vary significantly over time. Thus, it could be said that the vulnerability of the research area to cyclonic parameters is a big factor in how they are spread out across the area.

We used four components to assess the cyclone’s hazard and map the study area; the most significant factor contributing to the cyclone’s hazard (46%) was the cyclonic wind speeds, followed by the intense precipitation (40%) that the study area received when the cyclone crossed these islands; a great deal of destruction, including falling trees and damage to infrastructure, occurred during the event, and was confirmed by the local population. The two hazard variables, coastal flooding and wave impact, had little effect on adding to the hazard in the research region, which may be related to the less noticeable surge of 0.34 m, which did not significantly inundate the study area owing to the coastal flooding caused by the storm surge. Even though the waves reached great heights, they were unable to reach farther into the habituated region, which might be explained by the research area’s varied geographical characteristics.

Despite the limitations discussed above, the risk assessment findings provided as part of a pilot study in the study area will be useful in developing and implementing effective mitigation measures and disaster management policies aimed at mitigating the effects of tropical cyclones in the region. The Directorate of Disaster Management, ANI, last updated the Disaster Management Plan for these islands in 2016, and based on the conclusions of this research, we would like to suggest that the authorities update the plan and include the study’s findings in the new plan.