Definition and mapping of indicators of vulnerability to natural disasters

Definition and mapping of indicators of vulnerability to natural disasters Definición y cartografía de indicadores de vulnerabilidad frente a desastres naturales DOCTORAL THESIS Sumita Gayen Departamento de Geografía Física y Análisis Geográfico Regional Facultad de Geografía e Historia Universidad de Sevilla March, 2023 This page is intentionally left blank Definition and mapping of indicators of vulnerability to natural disasters Definición y cartografía de indicadores de vulnerabilidad frente a desastres naturales THESIS PRESENTED TO QUALIFY FOR THE TITLE OF Doctor in Geography by Sumita Gayen Supervisor: Dr. Luis Ismael Vallejo Villalta Departamento de Geografía Física y Análisis Geográfico Regional Facultad de Geografía e Historia Universidad de Sevilla March, 2023 This page is intentionally left blank Dedicated to my family This page is intentionally left blank Acknowledgements I am so happy and thrilled to write the acknowledgements of my doctoral thesis. This is a resultant of over 8 years of endless research in my field. This phase of my life, not only have developed me as a researcher but also grown me up as a human being and a professional. Now it is my privilege to acknowledge the fellow people whom I interact during this most interesting phase of my life. First and foremost, I would like to thank from my bottom of heart to my supervisor Professor Dr. Luis Ismael Vallejo Villalta who provided me the opportunity to be of his student. It’s my honor to be part of your research group. I am very grateful to you for your guidance in each and every step of this research work. Professor Sk. Mafizul Haque (University of Calcutta, India), guided and helped me immensely in several of my research works and papers. Thank you Mafida. I would also like to thank the all the distinguished professors and staffs of Universidad de Sevilla who helped me in many ways. In the beginning of my PhD, due to language barrier, I had a lot of problems. But Spanish society is so much open, friendly and rich; everyone provided me a lot of time and tried to understand what I want to say. Thank you Spain, my 2nd home. I have been able to come to this place by taking the lessons taught by my teachers and professors. Specially I want to mention the name of Dr. Ramkrishina Maiti (Vidyasagar University, India) who sowed the seeds of research in my dreams. I pay my obeisance to all my teachers and professors. I also want to say thanks to my friends, colleagues, and well-wishers. Specially, I want to mention my friend Azahara Albendiz. Your encouragement and guidance helped me a lot in this journey. Without the support and encouragement from my family members, this doctorate degree would not have been possible. I was not able to meet my parents in India frequently in last eight years as I was doing my PhD in Spain. It is very painful for my parents too. My father everyday asked me about the progress of my thesis, which is a great encouragement for me. Love of my mother, prayers and support always give me the strength and confidence. My sister and brother also give me the mental support. Thanks for both of my siblings. I never seen you, my grandfather. My grandmother, my ‘Thakuma’, I miss you every day. I know both of you are always blessing me from the heaven. My parents in law, uncle in law (dada) are also praying for my PhD. I am also grateful to them. To my Dip, thank you for your support and encouragement from the day of applying for PhD. You helped me a lot. It is impossible to write everything. In every moment of PhD you give me the strength and stand beside of me. Without your support and help, I could not be able to complete my research. My little son, my Nino, you suffered a lot for my thesis work. You have to stay many time without your mama because of my study. I will never forgot those days, when you were crying for me. Sorry betu. Never ending loves for you. Malaga Sumita Gayen March, 2023 This page is intentionally left blank Abstract In this thesis the design and generation of indicators of vulnerability to extreme natural phenomena is proposed. To this end, a series of publications have been carried out that develop these indicators based on the use of two types of techniques, such as the standardization of variables and the reduction of variable dimensions [Principal Components Analysis (PCA)]. In turn, these techniques have been implemented in two geographical areas with very different levels of development, namely Purba Medinipur district (West Bengal), in India, and the province of Malaga (Andalusia), in Spain. The units of analysis in each case are equivalent from the point of view of their administrative level, being blocks, in the case of Purba Medinipur, and municipalities, in the case of Malaga. The thesis presents the main groups of contents shared by the developed works. Basically, these contents include:  Firstly, an introduction which contextualizes the issue of vulnerability within the broader framework of studies on natural hazards and disasters.  Secondly, the objectives of the work are specified.  Thirdly, a chapter on methodology explains the strategies that have been followed in the different studies for the application of the aforementioned techniques (standardization and PCA).  Fourthly, a chapter with two fundamental parts is presented. On the one hand, the main results derived from the published works are presented. On the other hand, the opportunity to compile all these works and their results gives rise to an extensive discussion section in which the application strategies of the techniques used and the results obtained are critically analyzed, leading to a series of alternative proposals that attempt to improve the results obtained in the first instance.  Fifthly, a series of general conclusions and brief proposals for future work in this line are presented. The final part of the thesis contains the original publications that make up the compendium presented, as well as a selection of other activities and productions derived from the doctoral training period. Definition and mapping of indicators of vulnerability to natural disaster xvi Table 4.17. Location of 103 municipalities in different zones …………………….. 61 Table 4.18. Vulnerability score of seven factors for twenty-five blocks of Purba Medinipur district ……………………………………………… 65 Table 4.19. Total variance explained ………………………………………………. 75 Table 4.20. Dimension of five factors ……………………………………………… 75 Abbreviations ESPON: European Territorial Observatory Network FHI: Flood Hazard Index FRI: Flood Risk Index GIS: Geographic Information System HDI: Human Development Index HDR: Human Development Report IECA: Instituto de Estadística y Cartografía de Andalucía IMR: Infant Mortality Rate IPCC: Intergovernmental Panel on Climate Change ISDR: International Strategy for Disaster Reduction NASEM: National Academies of Sciences, Engineering, and Medicine PCA: Principal Component Analysis SPSS: Statistical Package for the Social Sciences SVI: Social Vulnerability Index UNDP: United Nations Development Programme UNISDR: United Nations International Strategy for Disaster Reduction WMO: World Meteorological Organization This page is intentionally left blank PART 1 THESIS This page is intentionally left blank Chapter I Introduction Almost all countries of the world are affected by natural hazards in each and every year and most of them bring about some sort of natural disaster. Along human history natural disasters have caused many damages and losses to human societies with many examples in which single settlement or even complex cultures or civilization have collapsed (Diamond, 2005; Leroy, 2020). According to several international sources (Jones et al, 2022), the natural disasters are in an increasing trend over the last decades in all over the world (Figure 1.1). Figure 1.1. Number of natural disaster events (Source: EM-DAT, 2021) In the last two decades a wide range of disasters have taken place with a death toll reaching approximately 60,000 people in each year; this represents 0.1% of global deaths (Ritchie and Roser, 2014). Firstly, it should be specified that two types of factors concur in the development of a disaster; on the one hand, the occurrence of a natural phenomenon of an extreme Definition and mapping of indicators of vulnerability to natural disaster 4 Risk = Hazard × Vulnerability (Eq. 1.1) nature is essential and, on the other hand, it is necessary for this occurrence to produce direct and indirect damage in a certain territory and in the society located on it. Secondly, it is convenient to transfer both circumstances to a potential situation that would take us to a time prior to the occurrence of the natural disaster. This potential situation is called natural risk, and the possibility of the occurrence of the extreme natural phenomenon is translated into natural hazard, while the damage and losses that can be expected as a result of such an occurrence would lead us to speak of vulnerability. Thus, risk was determined by these two components: hazard and vulnerability (Barroca et al, 2006) and according to International Strategy for Disaster Reduction (ISDR) 2004, the equation of risk (Eq. 1.1) is Returning now to the question of the increase in disasters at the global level that all studies seem to indicate, it is worth asking about the reasons behind these circumstances. Broadly speaking, there are three groups of factors that could be behind this increase, which refer, respectively, to factors related to available information, to causes linked to the behaviour of natural systems (hazard), or to circumstances related to human processes (vulnerability). In the first case, there are studies that highlight the role that may have been played by the implementation of different international programmes for the collection of data on natural disasters (Figure 1.2), which have led to greater availability of information, to which technological improvements in the systems for acquiring, processing and disseminating this information have also contributed (Guha-Sapir and Vos, 2011). Aside from this evident increased availability of information, an issue that is often highlighted is the shortcomings and discrepancies that can exist between different sources of information, which in some cases is related to a certain lack of consistency and consensus at the conceptual and criterion level when defining the term natural disaster (Hoeppe, 2016; Jones et al, 2022). Introduction 5 Figure 1.2. Number of disaster and global disaster databases creation (Source: Guha-Sapir and Vos, 2011) With regard to the second group of factors, related to changes in the behaviour of natural systems, it is clear that the arguments in this regard tend to refer to an increase in the number and severity (greater danger) of extreme events linked to meteorology and climate derived from the new scenarios associated with climate change [National Academies of Sciences, Engineering, and Medicine (NASEM), 2016; Intergovernmental Panel on Climate Change (IPCC), 2021]. Thus, as can be seen in Figure 1.3, it would be disasters linked to hydrological, meteorological and climatological causes that show a clear upward trend, while those with geophysical origins would present a more homogeneous behaviour with a less evident trend (Hoeppe, 2016). However, as mentioned above, it should be borne in mind that most of the studies reported here are a compilation of disasters and not strictly extreme events, so that factors linked to changes in socio-economic patterns and behaviours may lie also behind these trends, rather than only to a more severe natural environment. Definition and mapping of indicators of vulnerability to natural disaster 6 Figure 1.3. Relative trends (1980) of number of disaster events by types (Source: Hoeppe, 2016) Thus, the third group of factors mentioned above are related to certain changes in human activities that may be behind an increase in the number and importance of disasters (greater vulnerability). This section includes, in a very simplified way, two types of basic processes; either there is a new occupation of hazardous areas by the population, infrastructures or economic activities, thus creating new risk situations, or, in already existing risk situations, socio-economic and political changes take place which lead to an increase in this risk. In the latter case, these changes can be brought about by multiple and complex processes, some of the most obvious being an increase in population densities, population impoverishment, the physical deterioration of housing stock and infrastructure, or the neglect of health and safety services and activities. The impact of these types of social elements and factors (vulnerability) in shaping risk and in the triggering and consequences of natural disasters is evident at multiple levels and scales. At the global level, for example, as can be seen in Table 1.1, there are notable differences in the consequences of disasters depending on the type of territory that is affected; thus, when disasters are ordered according to the loss of human lives, it is poor countries that occupy the top positions, while if economic losses are taken into account, it is the developed countries, headed by the United States, that lead the ranking. This same model at the global level often works at other scales referring to territories such as regions, countries or urban environments, where socio-economic disparities can explain a large part of the differential effects of natural disasters [Cutter and Finch, 2008; Baker, 2012; Wang et al, 2020; European Territorial Observatory Network (ESPON), 2021]. Introduction 7 Table 1.1. Climate/ weather related disasters ranked according to A) deaths or B) economic losses (1979 – 2019) A) Disaster type Year Country Deaths 1 Drought 1983 Ethiopia 300,000 2 Storm (Bhola) 1970 Bangladesh 300,000 3 Drought 1983 Sudan 150,000 4 Storm (Gorky) 1991 Bangladesh 138,866 5 Storm (Nargis) 2008 Myanmar 138,366 6 Drought 1973 Ethiopia 100,000 7 Drought 1981 Mozambique 100,000 8 Extreme temperature 2010 Russia 55,736 9 Flood 1999 Bolivia 30,000 10 Flood 1974 Bangladesh 28,700 B) Disaster type Year Country Economic losses (US billion) 1 Storm (Katrina) 2006 United States 163.61 2 Storm (Harvey) 2017 United States 96.94 3 Storm (Maria) 2017 United States 69.39 4 Storm (Irma) 2017 United States 58.16 5 Storm (Sandy) 2012 United States 54.47 6 Storm (Andrew) 1992 United States 48.27 7 Flood 1998 China 47.02 8 Flood 2011 Thailand 45.46 9 Storm (Ike) 2008 United States 36.63 10 Flood 1995 DPR Korea 25.17 Source: World Meteorological Organization (WMO), 2021 This paper focuses on this aspect of vulnerability as a fundamental component of risk and thus as an essential factor in the genesis and development of disasters. In what follows, this concept is explored in more depth. Vulnerability derives from the Latin word ‘vulnerare’ (to be wounded), describes the potential to be harmed physically and psychologically. Based on this brief definition, the term has been used by multiple disciplines to refer to a set of characteristics that make something or someone more likely to suffer some kind of harm in the face of a negative circumstance (Coudouel and Hentschel, 2000; Paul, 2013). This circumstance can be a war, an economic crisis, an epidemic or an earthquake. It is in the latter case, i.e. in relation to the occurrence of extreme events in the physical environment, that This page is intentionally left blank Chapter III Methodology This chapter focuses on two crucial aspects: on the one hand, the justification of the selected study areas and their description, and on the other hand, the presentation of the different techniques used for the treatment of variables and the generation of social vulnerability indicators. 3.1. Study Area 3.1.1. Selection criteria for study areas: As discussed in chapter 2, one of the objectives of this work is focused on the application of social vulnerability indicators in environments with a very unequal level of development. For this purpose, two units of the second administrative level of India (district) and Spain (province) have been selected. Specifically, these units are the district of Purba Medinipur, belonging to the state of West Bengal; and the province of Malaga, which belongs to the autonomous community of Andalusia, respectively. A. Similarities i. Administrative Division: India is a large country composed of 28 states and 8 union territories (including national capital). The states and union territories are divided into several districts. The block is the next main level of administrative division after the district (Table 3.1). Spain consists of 17 autonomous communities and 2 autonomous cities. Autonomous community is the first level of political division of Spain. The autonomous communities are divided into provinces. Each province is divided in several municipalities. Table 3.1. Main administrative level in India and Spain Main administrative level India Spain First State (West Bengal) Autonomous community (Andalucia) Second District (Purba Medinipur) Province (Malaga) Third Block Municipality Definition and mapping of indicators of vulnerability to natural disaster 16 Purba Medinipur district and Malaga both are situated in the second administrative level. This comparative study between Purba Medinipur and Malaga has been done by analysing the data of third administrative level i.e. blocks for Purba Medinipur district and municipalities for Malaga province. ii. Coastal Region: Purba Medinipur district is situated in the southern part of West Bengal and coast of Bay of Bengal. It has a 27% coastline of West Bengal. The length of coastline in Purba Medinipur district is 60 km. The coastline of the district consists of flat, sand beaches and sand dunes. Malaga province is situated in the south of Spain. It is located in the Mediterranean coast. The province of Malaga has 160 km of coastline (14.53% coastline of Andalucia), running from Manilva to Nerja. As the coastline is vast, many types of beaches are found in here. Vast sandy beaches, rugged beaches and cliffs are found in Malaga. iii. Tourism: Flat sea beach, temples, museum, Indian tamarisk (Jhau) trees and the picnic spots are the main attraction of Purba Medinipur district for tourists. In all seasons, sea beaches are crowed with tourists. The maximum number of tourist comes from different parts of the state of West Bengal itself. Malaga is a famous tourist spot in Europe due to its beaches, sun kisses weather, sunshine of 300 days or more/ year, monuments, cultural heritage. The concentration of golf courses is another reason for tourist attraction. July, August and September are the main three months for tourist arrival. B. Dissimilarities i. Location: Purba Medinipur district is located in the southern part of West Bengal and eastern part of India (Figure 3.1). Geographically Purba Medinipur district is situated between 21°3635N - 22°5710N latitude and 86°3350E - 88°1240E longitude (Source: Census of India, 2011). Thus, Purba Medinipur district is located in the Torrid or Tropical zone and North-Eastern hemisphere. Methodology 17 Figure 3.1. Location map of Purba Medinipur district. (A) Map of India showing different states. The state of West Bengal is highlighted in yellow. (B) Map of West Bengal showing different districts within the study area; Purba Medinipur district is marked in yellow. (c) Map of study area, Purba Medinipur, showing its division into blocks. Malaga province is located in the southern part of the Andalucia and Spain (Figure 3.2). It is situated between 363N - 3729N latitude and 376W - 565W longitude. Malaga province is located in the Temperate zone and North-Western hemisphere. Figure 3.2. Location map of Malaga Province. A) Map of Spain showing different autonomous communities. The autonomous community of Andalusia is highlighted in yellow . B) Map of Andalusia showing different provinces. Malaga province is highlighted in yellow. C) Map of study area, Malaga province, showing its division into municipalities. ii. Human Development Index (HDI) score: The components of HDI 2020 are life expectancy of birth, expected years of schooling, mean years of schooling and per capita gross national income. India is categorized in medium HDI with 0.645 HDI score and ranked 131. HDI of Spain is very high and HDI score of Spain is 0.904 having 25 rank. India belongs to developing country and Spain belongs to developed country. Definition and mapping of indicators of vulnerability to natural disaster 18 iii. Total population: According to 2011 Census Data, population of Purba Medinipur district is 5,095,875. The population of Malaga is 1,594,808 in 2011 [Source: Instituto de Estadística y Cartografía de Andalucía (IECA), Junta de Andalucía]. Malaga province has a 1.54 times more area than Purba Medinipur district, but Purba Medinipur district has a 3.20 times more population than Malaga province. This indicates the vast population of Purba Medinipur district (Figure 3.3). Figure 3.3. Total population of Purba Medinipur district and Malaga province iv. Old aged persons (Above 60 years): The starting age of old is varied between countries and over time. The age of 60-65 is roughly equivalent to retirement ages in maximum countries. So it said to be the beginning of old age. The United Nations announced 60+ years aged population refers to the older population. The percentage of old aged persons (age 60+) in Purba Medinipur district is less than half of Malaga province. The average old aged people are only 8.67% out of which male is 8.20% and female is 9%. Low birth rate and the increasing life spans both are cause of rising of the older population in Malaga. In 2011 data of Malaga shows the population of aged 60+ is 20.52%. In Malaga, female old population is more than male population. The percentage of male and female older population is 9.34% and 11.18% respectively. 0 1000000 2000000 3000000 4000000 5000000 6000000 Purba Medinipur district Malaga province Total Population Methodology 19 v. Agriculture: The main crops grown in Purba Medinipur district are paddy, wheat, mustard, jute, potato, chillies, ginger etc. (Source: District Statistical Handbook, Purba Medinipur, 2015). The main crop is paddy which occupied more than 50% of total production. The experience of flood is common in almost every year which also affects crop production. Besides agricultural crops, Purba Medinipur district also famous for flower production. Olive is the most important crop in Malaga province. The other crops are wheat, barley, oats, grain legumes, beans etc. Fruits are also well grown in here. vi. Livestock: Livestock farming plays an important role in the rural economy of Purba Medinipur district. The main livestock animals are cattle and goat. Besides these animals; buffalo, sheep and poultry are raised in here. Buffaloes are used in agriculture as a labour. Poultries are found in Purba Medinipur in a large number. Livestock framings are raised to produce food and fiber. The main livestock animals in Malaga province are cow, sheep, goat, horse, pig, bird and rabbit. Cow, sheep, goat and pig are found in a large number and almost in all the municipalities. Goat farming and cheese making has a long history and tradition in Malaga. vii. Fishery: Inland or sweet water fishes are more popular in Purba Medinipur district. Fisheries are common in every block of the district. Fresh water fishing is increasing more rapidly than marine fishing in here. 84,323 ton fishes caught in the district in 2011-2012. Generally sea fishes are available in Malaga province. The province is famous for sea food. In 2011, total 10,431 ton fishes cough from sea. The maximum fishes are molluscs types. Only the coastal municipalities are famous for fishing. 3.1.2. Description of study areas A. Purba Medinipur district Purba Medinipur district is located in the southern part of West Bengal, India (Figure 3.1). It was formed on 1st January 2002 after the partition of Medinipur district into Purba Medinipur district and Paschim Medinipur district. It is surrounded by Rupnarayan and Hooghly rivers in the east, Paschim Medinipur district in the west, Howrah and Paschim Medinipur district by the north and south by the Bay of Bengal. In the southwest corner, this district shares a common border with the state of Orissa. The district Definition and mapping of indicators of vulnerability to natural disaster 20 comprises of 25 blocks and 5 municipalities. The district headquarters is located in Tamluk. The district is a part of lower Ganga plain. Costal features like estuary, tidal flats, mud flats, salt pans and sand dunes etc. are observed in Purba Medinipur distict. Haldi, Rasulpur, Keleghai, Kanshabati, Rupnarayan are main rivers of the district. All the rivers of Purba Medinipur district flows from west to east according to the slope of the region. Climate of Purba Medinipur district is of tropical monsoon type with hot and humid summer and dry winter. There are five main seasons observed in herespring, summer, rainy, autumn and winter. According to 2011 census data, population of Purba Medinipur district is 5,095,875 and it covers 5.58% of population of West Bengal. The district has a population density of 1081 inhabitants per square km. Purba Medinipur district ranks 8th in terms of population and population density of the state. The northern parts of the district are more densely populated than the other parts of the district. The population growth rate is 15.36%. The sex ratio of Purba Medinipur district is 938 females per 1000 males. Out of the total population 11.6% of population lives in urban area and 88.4% population lives in rural area. Whereas India’s urban population is 31.2% (Kumar, 2015). The literacy rate of Purba Medinipur district is 87.02% (2011). Only 37.49 % of population of the district is workers, whereas 62.51% of population is non-workers. The coastal blocks are famous for fish cultivation. Many people are engaged with fishing from the sea. Tourism also plays an important role in the economy of the coastal blocks. B. Malaga province Malaga province is located in southern Spain, under autonomous community of Andalucia (Figure 3.2). It is situated on northern side of the Mediterranean sea and surrounded by the Cadiz province to the west, Sevilla and Cordoba provinces to the north. The Alhama mountains separate Malaga from Granada to the east and Mediterranean sea to the south. The province contains 103 municipalities and the capital of the province is Malaga. From a geological point of view, Malaga province belongs to the Betic System, that forms the most western segment of the Mediterranean alpine orogeny. The relief gains in elevation from the coast to the inner lands, although one of its main characteristics is some sort of segmentation by which mountain ridges and depressions Methodology 21 take place. The mountains in Malaga are steep and rocky with peaks of around 500 meters. The highest peak of Malaga province is La Maroma at a height of 2,066 meter. It is located in the Natural Park of the Sierras Tejeda, Almijara and Alharma. Guadalhorce river is the main river of the province of Malaga, both in terms of its length and the size of its catchment area. Other rivers of the province are Guadalmedina, Guadiaro, Genal, Rio Grande, Chillar, Rio Velez, Hozgarganta etc. The climate of the province is Mediterranean type. Malaga enjoys an average of more than 320 sunny days in each year. Warmest months are July and August and sometimes temperature reaches over 40C. The winter is rainy and sometimes minimum temperature falls near to 0C. The population of Malaga is 1,594,808 and population density is 218.20/ sq. km (2011). Malaga has highest population density in Andalucia, and this population density is high mainly in the coastal zone. The female population is more than male population. Male: female ratio of Malaga province is 1000:1025 in 2011. The province has a high literacy rate. The average literacy rate is 95.60% in 2011. In the past, Malaga’s economy was based on various industries such as agriculture and textiles and it was even reported to be one of the richest provinces in the 19th Century. At present, the most important sector of economy in Malaga province is tourism. Other important economic sectors are construction, agriculture and technology services. 3.2. Research techniques: This chapter of thesis is divided into three sections. The first section investigates a comparative analysis between three standardization techniques. Comparative analysis is a combination of different methodologies for analysing data. It involves the analysis and synthesis of the similarities, differences and relationship between several methodologies. Normally two approaches are being used to standardize the indicators. First one is deductive approach which is based on theoretical understanding. Second one is inductive (empirical) approach which is based on statistical relationship of a large number of indicators set. Deductive approach is a ‘top-down’ approach and inductive approach is a ‘bottom-up’ approach. Cutter et al. (2003), Boruff et al. (2005), Rygel et al. (2006), Boruff and Cutter (2007), Holand et al. (2011), Chen et al. (2013), Heß (2017), Mavhura et al. (2017) used inductive approach in their studies. Whereas, Cutter et al. (2000), Bernard (2007), Cutter et al. (2010), Sajjad and Jain (2014), de Mello Rezende (2016) chose deductive approach for social vulnerability measurement. This comparative research is based on the three Definition and mapping of indicators of vulnerability to natural disaster 22 deductive approaches (maximum value transformation, z score transformation, min-max rescaling transformation). The second section explains the methodology of overall social vulnerability assessment in Malaga province and Purba Medinipur district. Social vulnerability assessment is a process of identifying, quantifying and analysing the vulnerability factors in a particular area. This assessment consists of 18 socio-economic indicators for construction of a Social Vulnerability Index (SVI). SVI is an additive model measuring the overall vulnerability of a region based on some of selected indicators (Cutter et al, 2003). Min-max standardization technique and PCA methodology have been employed to calculate SVI. The third section has explained the methodology of identification of flood risk zone in Purba Medinipur district. The district is one of the most flood prone districts of West Bengal (advocated by West Bengal Disaster Management & Civil Defence Department) and it underwent severe flood disaster events in last two decades. As a coastal region district, Purba Medinipur is flood prone as several coastal storms approach this area regularly. Besides, 75–80% of the total annual rainfall during the monsoon season, siltation, small catchment area and water release from Mukutmanipur dam, DVC (Damodar Valley Corporation) barrage etc. are the main cause of flood in Purba Medinipur district. For all the section, relationship between indicator and vulnerability are explained in two ways. For positive relationship, vulnerability of local people increases with the increasing value of indicator whereas for negative relationship, vulnerability decreases with increasing value of indicator. 3.2.1. Methodology for comparison of indicator standardization techniques Three deductive methodologies (maximum value transformation, z score transformation and min–max rescaling transformation), which are basically three different mathematical equations, have been adopted for this comparative analysis. A. Standardization Methods Theory of three different standardization methods are- i. Maximum value transformation In maximum value transformation (Eq. 3.1 and 3.2), data are transformed, such that the features lie within a specific range (0-1). It is the ratio between actual value and Methodology 23 maximum value of the variable. Higher value means vulnerability is high. Wu et al. (2002), Chakraborty et al. (2005) and Koks et al. (2015) used maximum value transformation standardization method in their study. For positive relationship, For negative relationship, Where, xi= Actual value, max= Maximum value ii. Z score transformation Z-score is numerical measurements (Eq. 3.3) that describe the position of a raw score in terms of its distance from the mean when measured in standard deviation. Z-scores may be positive or negative. A positive value indicates the score is above the mean and a negative value signifies it is below the mean. Zero z-score value represents that the data point's score is identical to the mean score. Zhang and Huang (2013), Evans et al. (2014), Lee (2014) used this standardization method for analysis SVI. The formula for calculating a z score is 𝑥= (q - µ) / σ Where, q = Actual value, µ= Mean, σ= Standard deviation For positive relationship, final value adds with total value and for negative relationship final value is deducted from total value. iii. Min-max rescaling transformation Min-max rescaling is a standardization method (Eq. 3.4 and 3.5) that rescales a set of data. The formula is as follows: For positive relationship, For negative relationship, Where, xi= Actual value, min= Minimum value, max= Maximum value (Eq. 3.1) (Eq. 3.2) (Eq. 3.3) (Eq. 3.4) (Eq. 3.5) Definition and mapping of indicators of vulnerability to natural disaster 30 hydro-geomorphological characteristics, the vast area of the district has recurring inundation with varied magnitude. In the present study, the inundation condition was segmented into three groups, i.e., large floods which have devastating nature and require relief and support, medium floods which lead to agricultural failure and small floods which show the simple inundation condition. To measure the flood frequency in the study area, all the large and medium floods as recognized by district administration were considered. Comparatively large floods occur less frequently than medium and small floods. In this study, the number of floods that occurred between 2002 to 2019 were considered as flood frequency for each block. The idea of flood frequency is used here to predict the structural failure by flood event along the riverbank, like damage to buildings, roads, highways, embankments etc. Analysis of flood frequency also helps engineers design safeguards to prevent structures from being damaged and protects against economic losses. Flood Hazard Index (FHI) is an indicator used to explain the different intensity or frequency of a flood in a particular geographical setting. Using the hydrological model and statistical methods in Geographic Information System (GIS) platform, the FHI was calculated by Asare-Kyei et al. (2015) and the resultant value of FHI varied from 1 (very low flood hazard intensity) to 5 (very high flood hazard intensity). Kabenge et al. (2017) used geoinformation based FHI for the Nyamwamba watershed in Uganda. In the present work, resultant values of FHI were classified into five different categories, i.e., very low, low, medium, high and very high and weighted values were defined as 1, 2, 3, 4 and 5 respectively (Table 3.3). Table 3.3. Flood hazard index Frequency of flood (in eighteen years) Hazard classes Hazard index <6 Very low 1 6–7 Low 2 8–9 Medium 3 10–11 High 4 >11 Very high 5 Source: Gayen et al, 2022 B. Vulnerability assessment Vulnerability assessment is an integral part of flood risk analysis (Rehman et al, 2019). Reduction of vulnerability is a core element in reducing disaster-related risk and it was Methodology 31 identified as the most important precondition for resilience of disaster-prone areas (di Girasole and Cannatella, 2017). To measure the vulnerability in the selected study area, 25 indicators were employed (Table 3.4). Table 3.4. Different indicators of vulnerability assessment Sl. no Indicators Description Relationship with vulnerability Reference 1 Population density Vulnerability is high in densely populated areas as many people live there. Positive Rasch, 2016 2 Population growth rate High population growth rate leads to high vulnerability of the society. Positive Bahinipati, 2014 3 Female population Women are more vulnerable due to their family responsibilities and low salary. Positive Tate et al, 2021 4 Child population (0 – 6 Years) Children are always dependent on another person. Positive Müller et al, 2011 5 Rural population Rural people suffer poor communication and medical facilities due to the remoteness from the urban area. Positive Bahinipati, 2014 6 Literacy Education and employment are inextricably linked. Higher education leads to better job opportunities. Negative Ahmad et al, 2016 7 Female literacy Female literacy increases the chance of having a job for women. Negative Bahinipati, 2014 8 Primary school density Higher primary school density enhances chance of education for every child. Negative Jha and Gundimeda, 2019 9 Employment rate Employed people have good economic condition which means good standard of living. Negative Kuhlicke et al, 2011 10 Households with bad house condition Bad condition of house increases the probability of damage from flood. Positive Jha and Gundimeda, 2019 11 Households without electricity To use modern technologies, electricity is absolutely needed. Positive Gayen et al, 2020 12 Households without sanitation Sanitation facility can reduce health related problems. Positive Gayen et al, 2020 13 Households without sewage Sewage system helps to get water out. Having no sewage system in house increases the impact of flood. Positive Gayen et al, 2020 14 Households having Safe drinking water is essential for health, as Negative Hazarika et Definition and mapping of indicators of vulnerability to natural disaster 32 source of safe drinking water it prevents exposure to unappealing pollutants, bacteria, viruses and parasites. al, 2018 15 Households having kitchen Households with a kitchen are hygienic. Negative Jha and Gundimeda, 2019 16 Households having banking service Banking facility supports economic condition. Negative Maiti et al, 2017 17 Cultivator Effect of flood has a huge impact on agriculture. Damage of crops is directly related to cultivator. Positive Bahinipati, 2014 18 Agricultural labor Agricultural laborers have no work during the flood period. Positive Bahinipati, 2014 19 Area covered by irrigation (ha.) More area covered by irrigation means higher adaptive capacity. Negative Bahinipati, 2014 20 Seed storage/ 10 sq km If agricultural productions are damaged by flood, seed stores are very important for replanting. Negative Proposed in this research work 21 Average number of co-operative societies/ 0.1 million population The co-operative society helps in product marketing, storage facilities, processing, transport and availing intensive cultivation by modern techniques. Negative Bahinipati, 2014 22 % of people having membership in cooperative societies Co-operative societies grant agricultural loan to the members. Negative Bahinipati, 2014 23 Number of bank/ 0.1 million population Number of bank/ 0.1 million population facilitates economic infrastructure of the region. Negative Bahinipati, 2014 24 Road density Roads are important for rescue purpose during any type of hazard. Negative Jha and Gundimeda, 2019 25 Permanent flood shelter/ 10 sq. km Flood shelter is extremely important before and after disaster. It is a temporary home for the floodaffected people. These shelters are used to manage relief and rehabilitation activities in an organized way. Negative Hazarika et al, 2018 Source: Gayen et al, 2022 There are many methods for data normalization like z score, min-max rescaling transformation, maximum value transformation etc. (Gayen et al, 2020). The normalization of vulnerability indicators is carried out through the min-max Methodology 33 normalization method followed by Žurovec et al. (2017) and Gayen et al. (2021) (equation 3.4 and 3.5). To reduce the dimension of indicators, the PCA technique was employed. PCA is considered as a better empirical method than other alternative methods of reducing dimensionality in the data, like correspondence analysis, multivariate regression, or factor analysis (Vyas and Kumaranayake, 2006). The criteria to retain all the factors were selected based on Kaiser’s rule (Kaiser, 1960). Horn (1965), Crawford (1975), Velicer (1976) also suggested some rules to retain the factors. However, Kaiser’s rule is easy to calculate and it is available in many statistical packages. The varimax rotation method was used in PCA. The SPSS software (26) was used to calculate the PCA. After analysis, the vulnerability score was obtained by summing up all the factors, as described in Eq. 3.8 (Aksha et al, 2019). Vulnerability score= Fac1 + Fac2+Fac3 + Fac4 + Fac5 + Fac6 + Fac7 Vulnerable regions were also classified into five categories (very low, low, medium, high and very high), adopted from Gautam (2017) and Aksha et al. (2019), and weighted value was 1, 2, 3, 4 and 5 respectively (Table 3.5). Furthermore, de Mello Rezende (2016) averaged out all the standardized (method min-max) scores for vulnerability index calculation and classified them into five different levels. Here, vulnerability index was classified through standard deviation calculation of the data. Table 3.5. Vulnerability index Standard deviation Vulnerability classes Vulnerability index <−1.5 Very low 1 −1.5–−0.50 Low 2 −0.50–0.50 Medium 3 0.50–1.5 High 4 >1.5 Very high 5 Source: Gayen et al, 2022 C. Flood risk assessment The flood risk assessment was calculated through a Flood Risk Index (FRI) by multiplying the indexes obtained for the hazard and the vulnerability. Phongsapan et al. (2019) calculated FRI by the normalization of the values and categorized low (<5%), moderate (5–10%) and high (>10%) zones. Buta et al. (2020) analysed FRI by (Eq. 3.8) Definition and mapping of indicators of vulnerability to natural disaster 34 multiplying the Flood Hazard Index (FHI) by the Flood Vulnerability Index (FVI) and classifying it into four categories. In this study, the risk index was classified into five different classes for the detailed explanation of flood risk zones. FRI was categorized as very low (1–2), low (3–4), medium (5–9), high (10–15) and very high (16–25), as displayed in Table 3.6. Flood hazard map, vulnerability map, and risk map were created by using ArcGIS software (version 10.2). Table 3.6. Flood risk index Vulnerability index Hazard index Very low (1) Low (2) Medium (3) High (4) Very high (5) Very low (1) VL 1×1=1 VL 1×2=2 L 1×3=3 L 1×4=4 M 1×5=5 Low (2) VL 2×1=2 L 2×2=4 M 2×3=6 M 2×4=8 H 2×5=10 Medium (3) L 3×1=3 M 3×2=6 M 3×3=9 H 3×4=12 H 3×5=15 High (4) L 4×1=4 M 4×2=8 H 4×3=12 H 4×4=16 VH 4×5=20 Very high (5) M 5×1=5 H 5×2=10 H 5×3=15 VH 5×4=20 VH 5×5=25 Chapter IV Results and discussion 4.1. Results This part is divided into three sections according to the methodology. First section examines three standardization techniques. Second section shows the overall SVI in Malaga province and Purba Medinipur district. The third section explains flood risk zone in Purba Medinipur district. All results are shown by maps and categorized in five classes based on standard deviation score. Red colour indicates higher vulnerability and green colour shows low level of vulnerability. 4.1.1. Comparison of indicator standardization techniques The result shows the comparative analysis between three different deductive standardization methods (maximum value transformation, z score transformation, minmax rescaling transformation) and explains their relationships. A. Purba Medinipur district To determine the methodologies of this work, the main prerequisite was selection of indicators. 12 different indicators have been selected. Table 4.1 provides classification and the name of the indicators and some basic information about those. Table 4.1. Indicators and their description Sl. no. Indicators Description Relationship with vulnerability Literature review 1 % of female population Women are more vulnerable than men due to low salary, family responsibilities and emotion. Positive Müller et al, 2011 2 % of child population (0-6 Years) Children are vulnerable due to lack of education, experience and knowledge. Positive Girasole and Cannatella, 2017 3 % of rural population Rural population is more vulnerable because of lower income and depending on urban for medical treatments. Positive Cutter et al, 2003 Definition and mapping of indicators of vulnerability to natural disaster 36 4 % of minority population Minority people are less educated than other. Positive Aksha et al, 2019 5 % of literate population High standard of education present greater chance of employment. Educated people can use the social network comparatively in better way, which is important tool for emergency planning and preparedness. Negative Ahmad et al, 2016 6 % of households with no electricity Without electricity, it is impossible to access modern communication gadgets like television, mobile, internet. Electricity also helps to make working environment better and increases working hours. Positive Mavhura et al, 2017 7 % of households with no sanitation Sanitation facility helps to reduce diseases. Positive Chen et al, 2013 8 % of households with no car Personal vehicles are important in emergency situation. Positive Bergstrand et al, 2015 9 % of households with no kitchen Absence of kitchen in home decrease food safety and increase health risk. Positive Ge et al, 2017 10 % of households with no sewage Without sewage people can face unhygienic conditions in surrounding areas. Waste water can contaminate the environment and cause diseases. Positive de Mello Rezende, 2016 11 Employment rate Employment influences the ability to recover after a disaster. Negative Dwyer et al, 2004 12 Infant mortality rate High infant mortality rate indicates infants are unsafe from different disease. Positive Lee, 2014 Source: Gayen et al, 2020 i. Vulnerable zone construction To understand the degree of social vulnerability, all the 25 blocks have been categorized in five different vulnerable zones (Figure 4.1). The results of the analysis show that most socio-economically vulnerable blocks are located in the eastern part of the Purba Medinipur district. Results and discussion 37 Figure 4.1. Social vulnerability map of Purba Medinipur district based on three different methodologies (Source: Gayen et al, 2020) Table 4.2. Very low vulnerable blocks Standardization methods Name of the blocks Maximum value transformation Nandigram-II Z score transformation Contai-I Min–max rescaling transformation Contai-I Table 4.2 shows the very low vulnerable blocks in Purba Medinipur district. Contai-I block is identified as a very low vulnerable category. High employment rate, less child and female population, better sanitation and sewage make Contai-I block very low vulnerable. According to maximum value transformation theory, very low vulnerable block is Nandigram-II. Although this block has high female, child and rural population, but high literacy and employment rates and low Infant Mortality Rate (IMR) make Nandigram-II very low vulnerable. Table 4.3. Very high vulnerable blocks Standardization methods Name of the blocks Maximum value transformation Sutahata Z score transformation Nandigram-I, Sutahata, Khejuri-II Min–max rescaling transformation Nandigram-I, Sutahata, Khejuri-II Definition and mapping of indicators of vulnerability to natural disaster 38 Sutahata block is very high vulnerable (Table 4.3) because of high rate of rural population, less employment rate, poor sanitary and electricity problems. By using z score transformation and min–max rescaling transformation very high vulnerability have also been found in Nandigram-I and Khejuri-II blocks. The main causes of vulnerability in Nandigram-I and Khejuri-II are similar to Sutahata block. Out of the 25 blocks, 12 blocks are categorized with same vulnerable zone irrespective to the methodologies adopted. This also signifies a strong relationship between the three different methodologies. ii. Relationship measurement Result obtained from the correlation indicates both positive and negative relationship between different indicators. It is also important thing that no relationships are strongly correlated with each other. Value of 0 between minority population and households with no sanitation indicates that there is no linear relationship between these two indicators. Table 4.4. Pearson’s correlation coefficient between three methodologies Z score transformation Maximum value transformation Min–max rescaling transformation Z score transformation 1 Maximum value transformation 0.74 1 Min–max rescaling transformation 0.99 0.79 1 Source: Gayen et al, 2020 The correlation coefficient between z score and min–max rescaling transformation is 0.99 (Table 4.4). This value shows positively "very strong" (Chan, 2003) relationship between z score and min–max rescaling transformation. The correction value between z score and maximum value transformation is 0.74, whereas the same between maximum value transformation and min–max rescaling transformation is 0.79. The values of 0.74 and 0.79 signify the relationships are positive and "moderately strong" (Chan, 2003). Such a "very strong" and "moderately strong" positive relationships provide an important implication that there lies no significant differences between three standardization methods employed. Results and discussion 39 B. Malaga province For the comparative analysis of Malaga province, 23 indicators (physical, demographic, economic, social, and health) have been used. The comparative indicator standardization techniques of Malaga province has been analyzed by using all the indicators separately as shown below in Table 4.5. Table 4.5. Description of indicators Sl. no. Indicators Description Relationship with vulnerability Literature review 1 Distance from capital Capital plays an important role for different administrative services. Economic growth rate is higher in capital compared to other regions. Capital region is well connected by other regions and also job opportunities are more in here. Positive Proposed in this research work 2 Gender Women are more vulnerable than men. Women workers are paid less salary than man for the same work. Positive Müller et al, 2011 3 Child population (below 5 years) Children have less physical strength and they are dependent on others for survival. Positive Török, 2018 4 Aged population (65 years and above) The older persons have knowledge, but sometimes they are physically disabled and challenged to work for activity. They have lack of advance knowledge like communicating others people by internet. Positive Reid et al, 2009 5 Dependency ratio Higher dependency ratio indicates lower collection of government tax, because working population has a greater responsibility to pay tax. Higher dependency ratio could create labor shortages that can reduce agriculture and industrial production. Positive Kalaycıoğlu et al, 2006 6 Single parent nuclei Child nurturing is very difficult for single parent under any circumstances and parent has sole liability for all activities to support their livelihood. Positive Flanagan et al, 2011 7 Median age High median age reduces the capability towards the preparedness to disasters i.e it is negatively linked with vulnerability of a region. Negative Burton, 2010 8 Foreigners Foreigners are not well familiar about the local Positive Aksha et al, Definition and mapping of indicators of vulnerability to natural disaster 46 Table 4.10. Total variance explained Total variance explained Comp onent Initial eigenvalues Extraction sums of squared loadings Rotation sums of squared loadings Total % of variance Cumul ative % Total % of variance Cumul ative % Total % of variance Cumul ative % 1 4.893 27.181 27.181 4.893 27.181 27.181 3.445 19.137 19.137 2 3.183 17.685 44.866 3.183 17.685 44.866 2.854 15.857 34.995 3 2.386 13.256 58.123 2.386 13.256 58.123 2.426 13.479 48.473 4 1.818 10.102 68.224 1.818 10.102 68.224 2.415 13.419 61.892 5 1.448 8.046 76.271 1.448 8.046 76.271 1.955 10.862 72.754 6 1.123 6.239 82.510 1.123 6.239 82.510 1.756 9.755 82.510 7 0.921 5.118 87.627 8 0.645 3.582 91.209 9 0.429 2.383 93.592 10 0.324 1.799 95.391 11 0.263 1.463 96.854 12 0.211 1.170 98.023 13 0.161 0.892 98.916 14 0.085 0.473 99.388 15 0.063 0.350 99.739 16 0.034 0.189 99.928 17 0.009 0.051 99.979 18 0.004 0.021 100.000 Source: Gayen and Villalta, 2022 The values of "Initial eigenvalues" and "Extraction sums of squared loadings" columns are same. The "Rotation sums of squared loadings" represents the distribution of the variance after the varimax rotation. Total percent of variance explained, is same (cumulative value for factor 6 is 82.51%) for all of three (initial eigenvalues, extraction sums of squared loadings, rotation sums of squared loadings). But rotation changes the extracted factors of eigenvalues. So, after rotation each extracted factor counts for a different percentage of variance explained. Results and discussion 47 Figure 4.3. Scree plot (Source: Gayen and Villalta, 2022) Scree plot (Figure 4.3) is a line of the eigenvalues against components in an analysis. It always displays the eigenvalues in a downward curve. It helps to determine the number of factors remain and the relative importance of the factors. The x-axis contains component number and y axis represents the eigenvalue. From scree plot, the components having score above one (total six) have been considered. At the end of the graph, line is almost flat, which signifies that on that portion each successive component is accounting smaller amounts of the total variance. Table 4.11. Dimension of six factors Name % of variance explained Dominant indicator Component loading Factor 1 27.181 Distance from capital 0.906 Population density -0.901 Households without sanitation -0.854 Households without electricity 0.655 Factor 2 17.685 Literacy rate 0.859 Minority population 0.697 Child population 0.694 Factor 3 13.256 Employment rate 0.864 Households without kitchen -0.677 Population growth rate 0.612 Factor 4 10.102 Infant mortality rate -0.856 Rural population 0.767 Definition and mapping of indicators of vulnerability to natural disaster 48 Pressure of population/ hospital 0.506 Factor 5 8.046 Households without sewage 0.577 Pressure of population/ bed 0.491 Households without car 0.042 Factor 6 6.239 Households having banking service 0.843 Female population -0.608 Source: Gayen and Villalta, 2022 Six factors have been derived from the principal component analysis and contribute to social vulnerability assessment in Purba Medinipur district (Table 4.11). Following Cutter et al. (2003) methodology, all the factors summed up to obtain final or SVI score (Table 4.12) Results and discussion 49 Table 4.12. SVI scores and vulnerable zones for twenty five blocks of Purba Medinipur district Blocks Fac1 Fac2 Fac3 Fac4 Fac5 Fac6 SVI score Rank Level of vulnerability Tamluk -1.86849 -0.20082 0.16243 0.04979 -0.4291 -1.1474 -3.43353 22 Low Sahid Matangini -1.82438 -0.63505 0.19564 0.08275 -1.2901 0.0469 -3.42426 21 Low Panskura -1.37659 0.76426 1.64011 0.34084 0.0381 -0.5808 0.82593 13 Medium Kolaghat -1.52912 -0.42288 1.00205 0.19139 -1.0346 -0.2039 -1.99703 19 Low Moyna -0.80505 -0.10743 1.03917 -0.13885 1.4736 2.0857 3.54718 1 Very high Nandakumar -1.03044 0.79712 0.5931 0.02964 0.9916 -0.1686 1.21239 10 High Chandipur -0.44171 -0.15019 -0.76743 0.74738 0.5707 0.958 0.9168 12 Medium Mahishadal -0.61012 -0.54197 -0.68288 1.23374 -0.0127 -0.5481 -1.16198 18 Medium Nandigram-I 0.07469 0.76035 -1.63292 1.55552 -0.096 1.1707 1.8323 7 High Nandigram-II 0.24488 -0.39963 -1.63729 1.05733 0.2808 -0.312 -0.76591 17 Medium Sutahata -0.72797 0.7796 -1.25186 -1.7638 -0.8517 0.1226 -3.69318 23 Very low Haldia -0.50471 -0.22737 -1.77458 -3.11207 1.3926 -0.41 -4.63614 25 Very low Potashpur-I 0.85323 -0.44343 1.55536 -0.50193 2.0969 -0.7477 2.81241 3 High Potashpur-II 1.20059 0.46007 0.69389 -0.52517 0.1459 0.5286 2.50384 4 High Bhagawanpur-I -0.18615 -0.15798 -0.2148 0.63091 1.4437 0.612 2.12763 5 High Egra-I 0.96283 1.11313 1.47748 -1.03348 -2.0222 0.1886 0.68633 14 Medium Egra-II 1.17336 -0.3313 0.58555 -0.04488 -0.4642 0.9249 1.84346 6 High Khejuri-I 0.65751 0.02055 -0.43742 0.96441 -0.7632 0.7786 1.22052 9 High Khejuri-II 0.71749 3.39769 -0.72671 0.07372 0.156 -0.3328 3.28542 2 High Definition and mapping of indicators of vulnerability to natural disaster 50 Bhagawanpur-II 0.71635 -0.81693 0.04907 0.56674 -0.5456 0.6354 0.60508 15 Medium Ramnagar-I 0.83607 -0.05633 -0.55242 0.59694 -0.4429 -3.0018 -2.62043 20 Low Ramnagar-II 1.18157 -0.85295 0.39648 0.07395 0.0646 -1.1586 -0.29495 16 Medium Contai-I 0.67953 -2.18189 -0.72098 -1.20303 -1.2198 0.7513 -3.89485 24 Very low Deshapran 0.75834 0.34394 0.55576 -0.34431 -0.5515 0.5728 1.33508 8 High Contai-III 0.84828 -0.91056 0.45323 0.47245 1.0688 -0.7643 1.16789 11 Medium Source: Gayen and Villalta, 2022 Results and discussion 51 Figure 4.4 shows different zones of social vulnerability in Purba Medinipur district. This study reveals that most vulnerable blocks are lying on the central, east and western part of the district. Figure 4.4. Social vulnerability map of Purba Medinipur district (Source: Gayen and Villalta, 2022) The result shows that the most vulnerable block is Moyna. The block is located near to the district capital Tamluk. But rural population is high (97.09%) in Moyna. Social parameters like households without electricity, no latrine, no kitchen in house and no sewage system etc. make Moyna the most vulnerable block in the district. Social parameters of vulnerability like households having no sanitation, no sewage system, no kitchen are related to health problems. Sewage system and sanitation facilities are important to reduce health related problems. Definition and mapping of indicators of vulnerability to natural disaster 52 Almost one in two people in the developing countries don’t have improved sanitation (HDR, 2006). After a decade of twenty-first century more than 18.29% households have no sanitation facility in Moyna block. Even some people do not use latrine, though they have latrine. Sewage systems carry waste water by pipes. From the dumped sewage water, ground water could be contaminated and water borne diseases such as hepatitis, cholera, and typhoid could spread. Mainly middle part of the district is high vulnerable. A large number of blocks of Purba Medinipur district (36%) come under high vulnerable category. Khejuri-II block belongs to this category and is the second most vulnerable block in the district. Here population density is quite low, but values of other indicators are high. Khejuri-II has highest rural population, minority population, illiteracy rate and households without electricity within the district. Another important cause of vulnerability is high population growth rate. Population growth is considered as increase of vulnerability (Cutter et al, 2003). Decadal population growth of Purba Medinipur district is 15.54%, which is considered as quite high. Khejuri-II block has no urban population; but population growth rate (18.75) is extremely high. Nandakumar, Bhagawanpur-I, Patashpur-I, Patashpur-II, Egra-II, Deshapran, Khejuri-I and Nandigram-I blocks are also categorized in high vulnerable zone. High rate of child and rural population, high percentage of without electrified household and household with no sewage system are the main causes of vulnerability for maximum blocks. Nandakumar, Patashpur-II, Egra-II, Khejuri-I blocks have 100% rural population. Normally rural population has lack of good communication system. The vulnerability of a region also depends on its location and spatial interaction with other regions. Communication is important for resource mobilization, accessibility and area development. Good connectivity provides better relief facilities during the hazards. Nandigram-I block has highest population growth rate (18.97%) and highest child population (13.66%) within the district. Children are the most vulnerable groups in the society during disaster events (Cutter et al, 2003). They are vulnerable due to lack of their education, experience and knowledge during the time of hazard (Gayen et al, 2021). 32 % of blocks fall into the medium vulnerable zone. Panskura, Mahishadal, Chandipur, Nandigram-II, Bhagawanpur-II, Contai-III, Egra-I, Ramnagar-II blocks are under this category. Three blocks of northern part (Kolaghat, Sahid Matangini, Tamluk) and one block of southern part (Ramnagar-I) in Purba Medinipur district have low vulnerability. Except Ramnagar-I block, other blocks have high population density. Tamluk is district Results and discussion 53 capital and Kolaghat, Sahid Matangini blocks are near to district capital. All the health and medical indicators like, IMR, pressure of population/ hospital and pressure of population/bed contribute to reduce the vulnerability. After disaster, accessibility of medical service is very important (Chen et al, 2013). Haldia, Sutahata and Contai-I blocks are in very low vulnerable category. These blocks have low female population, child population, rural population, high literacy rate and employment rate. Several researchers (Morrow, 1999; Dwyer et al, 2004; Tapsell et al, 2005; Kuhliche et al, 2011) mentioned women section of society is more vulnerable than men. Economically disadvantaged peoples are more affected by disasters (Flanagan et al, 2011). Good economic condition leads population to live better and longer lives as well as good standard of living. Good economic condition means high rate of employment that translated into the consumption of higher quantity and better quality nutrients. As well, it helps to cope with natural hazard and to a faster recovery from those. Besides these conditions, number of hospitals and nursing homes and bed availability are quite good than other blocks of the district. In Contai-I block female population and child populations are lowest and literacy rate is highest in the district, but distance from capital is far. Haldia block is least vulnerable within the district. Good demographic, social parameter and health conditions are the main contributors of low vulnerability for Haldia block. Another important thing is Haldia is the most industrialized block of the district. So, employment rate of Haldia block is high. Though Haldia block has high population growth rate, overall vulnerability index is less. Purba Medinipur district comprises of five municipalities: Tamluk, Panskura, Haldia, Egra and Contai. All the municipalities included blocks are situated between very low (Contai, Haldia), low (Tamluk) and medium (Panskura, Egra) vulnerable zones, because municipalities are facilitated with different possibilities (communication, more job opportunities, hospital facilities etc.) than other regions. Definition and mapping of indicators of vulnerability to natural disaster 54 Figure 4.5. Social vulnerability index of different blocks within the district in radar diagram (Source: Gayen and Villalta, 2022) The radar chart (Figure 4.5) shows position of every block within the district in a diagram. B. Malaga province The results of the principal component based analysis of all the indicators are shown below. The analysis has highlighted socio-economic picture of Malaga province. Table 4.13 shows the 18 indicators that are used in this analysis. Table 4.13. Indicators and their range, mean, standard deviation and variance Sl. no. Indicators Range Mean Standard deviation Variance 1 Distance from capital 153 66 35.06 1217.34 2 Population density 7114.62 260.63 824.63 673411.79 3 Female population 9.16 49.09 1.46 2.10 4 Child population (below 5 years) 6.76 4.39 1.34 1.77 5 Aged population (65 years and above) 26.45 26.52 6.37 40.18 6 Dependency ratio 38.97 54.15 6.87 46.78 7 Median age 16.77 45.12 3.69 13.50 8 Large household (5 or more) 12.85 3.75 4.18 17.26 9 Single living person 18.99 10.42 3.93 15.33 -6.00 -4.00 -2.00 0.00 2.00 4.00 Moyna Khejuri-II Potashpur-I Potashpur-II Bhagawanpur-I Egra-II Nandigram-I Deshapran Khejuri-I Nandakumar Contai-III Chandipur Panskura-IEgra-I Bhagawanpur-II Ramnagar-II Nandigram-II Mahishadal Kolaghat Ramnagar-I Sahid Matangini Tamluk Sutahata Contai-I Haldia Radar Diagram Results and discussion 55 10 Single parent 33.98 13.60 6.93 47.59 11 Foreigners 54.4 13.68 11.91 140.49 12 Literacy rate 15.09 95.82 2.51 6.23 13 Employment rate 28.32 77.11 6.64 43.69 14 Agricultural population 31.37 9.03 7.04 49.03 15 Primary health care/ 1000 population 5.82 1.03 1.29 1.65 16 Number of pharmacy/ 1000 population 5.52 0.96 1.11 1.23 17 IMR 333.33 4.80 33.44 1107.47 18 Per capita income (annual in euro) 7070.17 4058.50 1326.07 1741401 The table (Table 4.14) shows the eigenvalues and percentage of variance explained. The five factors are explaining 70.482% of the variance. The first factor explained 37.955% of the variance, the second factor 10.894% of the variance, the third factor 7.988% of the variance, the fourth factor 7.686% of the variance and the fifth factor 5.959% of the variance. Table 4.14. Total variance explained Total variance explained Comp onent Initial eigenvalues Extraction sums of squared loadings Rotation sums of squared loadings Total % of Varian ce Cumulati ve % Total % of Varian ce Cumul ative % Total % of Varian ce Cumul ative % 1 6.832 37.955 37.955 6.832 37.955 37.955 4.550 25.280 25.280 2 1.961 10.894 48.849 1.961 10.894 48.849 3.184 17.687 42.967 3 1.438 7.988 56.836 1.438 7.988 56.836 1.858 10.322 53.288 4 1.384 7.686 64.523 1.384 7.686 64.523 1.828 10.156 63.445 5 1.073 5.959 70.482 1.073 5.959 70.482 1.267 7.037 70.482 6 0.889 4.941 75.423 7 0.845 4.696 80.120 8 0.642 3.567 83.687 9 0.576 3.198 86.885 10 0.520 2.888 89.774 11 0.479 2.662 92.436 12 0.399 2.216 94.652 13 0.294 1.631 96.282 14 0.256 1.423 97.705 Definition and mapping of indicators of vulnerability to natural disaster 62 Low vulnerable zone comprises of 16 municipalities. These municipalities are clustered in southern part of the province. 5.38% municipalities are come down in very low vulnerable category. Four western municipalities (Benalauria, Atajate, Cartajima, Juzcar) and three eastern municipalities (Totalan, Macharaviaya, Salares) fall in this zone. Socio-economic parameters like population density, child population, dependency ratio, foreigners, IMR etc. are very less and median age, literacy are high. Only per capita income is very less for all these low vulnerable municipalities. Even Juzcar municipality has the lowest income/ person within the province. Except income/ person, other values of indicators are quite low, make these municipalities lying under very low vulnerable zone. Population density, child population, primary health care, pharmacy, per capita income influence social vulnerability of the Malaga province. Population density is a sign of potential affected inhabitants (Balica, 2012). It has a significant impact on the temporal trends of social vulnerability (Cutter and Finch, 2008). The densely populated municipalities are mostly located along the coast because of coastal fisheries, tourism and recreation. Marbella, Fuengirola, Benalmadena, Torremolinos, Malaga, Rincon de la Victoria etc. coastal municipalities are situated in high and very high vulnerable zone and they have more than 1000 population/ sq. km. Other factors that affect vulnerability include primary health care and pharmacy. Presence of primary health care minimizes the social vulnerability (Gayen et al, 2021). Similarly, presence of pharmacy also important to reduce social vulnerability in emergency situation. Higher per capita income expands the health care facilities and quality of life. Average annual income/ person of the province is 4058.50 Euros. Nineteen municipalities have below 3000 Euros annual income/ person. Average old aged population of the province is 26.54% which signifies high number of old aged persons in Malaga province. 4.1.3. Flood risk analysis Flood risk zone identification is done by multiplying the hazard and vulnerability score through a FRI developed by Tchórzewska-Cieślak et al. (2018), Jelínek et al. (2009) and Aguirre-Ayerbe et al. (2018). Results and discussion 63 A. Flood hazard scenario and its spatial variation The flood hazard map (Figure 4.8) shows the areas susceptible to floods in the Purba Medinipur district. In the present study, this was determined based on the frequency information of floods. The map indicates that the highly flood prone area is in the northern and western parts of the district. These parts of the district are characterized as being gently sloping (0–5 degree) and flat, mostly composed of alluvial soil at the surface. The presence of a high amount of clay particle (60%) makes this soil less permeable. Besides due to the convergence of rivers, this region is vulnerable to water logging during monsoon season owing to the presence of alluvial soil. Figure 4.8. Flood frequency map of Purba Medinipur district (2002–2019) [Source: Gayen et al, 2022] Panskura, Moyna, and Patashpur-I blocks are very high flood prone and have experienced more than eleven floods from 2002 to 2019. Panskura block is mainly flooded by the New Kasai (also known as New Cossye) river (Das and Sahu, 2017) and Patashpur-I is usually flooded by the Keleghai river. More than 70% of the soil in the Patashpur-I block is clay-like in texture (Sahu1, 2014). Moyna block is surrounded by rivers and a canal on three sides. The Kasai river (also known as the New Cossye river) Definition and mapping of indicators of vulnerability to natural disaster 64 flows in the east; the Chandia river flows in the west; the Chandia and Keleghai rivers flow in the south; the Baksi canal is on the northern part of the block (Sahu2, 2014) and forms a basin-like topography. The whole Moyna block is situated in a low land (Laha1 et al, 2014). All these factors make the Moyna block extremely vulnerable to flood and waterlogging. The New Cossye and Kaliaghai rivers are mainly responsible for the floods in the Purba Medinipur district. The reoccurrence interval of high magnitude flood events in the Kaliaghai river is 1 in 2.1 years (Laha2 et al, 2014). From 2002 to 2019, the New Cossye river crossed the danger level thirteen times (Figure 4.9). The water level of Kaliaghai also crossed the danger level eleven times (Figure 4.10) within the study period (2002–2019). The Rupnarayan and Rasulpur rivers are also responsible for floods in the district. Figure 4.9. Highest gauge height in New Cossye river at Panskura rain gauge station (2002–2019) [Source: Gayen et al, 2022] Figure 4.10. Highest gauge height in Kaliaghai river at Amgachia rain gauge station (2002–2019) [Source: Gayen et al, 2022] 4 5 6 7 8 9 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Water height (meter) Year DL- 5.79 meter PDL- 5.18 meter EDL- 6.4 meter 6 7 8 9 10 11 12 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 Water height (meter) Year EDL- 9.9 meter DL- 9.29 meter PDL- 8.69 meter EDL - Extreme Danger Level DL - Danger Level PDL - Primary Danger Level EDL - Extreme Danger Level DL - Danger Level PDL - Primary Danger Level Results and discussion 65 During the study period, the Tamluk, Kolaghat, Nandakumar, Patashpur-II, Bhagawanpur-I, Egra-I and Ramnagar-I blocks suffered ten to eleven floods that were grouped as high category. Only the Nandigram-II (five floods from 2002 to 2019) block was in the very low flood category. B. Understanding the vulnerability scenario and its spatial variation Another component of risk assessment is vulnerability analysis. Flood maps alone are not sufficient to measure the risks to people, property, infrastructure, and services (Armenakis et al, 2017). The damage of floods also depends on the vulnerable condition of the region. Seven factors have been determined from selected indicators using PCA and contributed to understanding the vulnerability in the Purba Medinipur district. Seven factors explained 83.84% of the total cumulative variance. The first three factors have more than half of the variance. Block wise scores of seven factors were summed up to obtain the total vulnerability score. Total vulnerability scores were classified into five categories. Table 4.18 represents the total vulnerability score of each and every block of the district. The Khejuri-II block has the highest vulnerability and the Tamluk block has the lowest vulnerability. A high vulnerability score indicates low economic status and underdeveloped infrastructure of the society. Table 4.18. Vulnerability score of seven factors for twenty-five blocks of Purba Medinipur district Blocks Fac1 Fac2 Fac3 Fac4 Fac5 Fac6 Fac7 Total vulnerability score Khejuri−II 1.035 1.191 1.973 1.224 0.164 0.948 0.011 6.546 Nandigram−I 0.667 1.821 1.124 0.147 1.440 −0.512 −0.006 4.681 Moyna −1.213 −1.489 −0.212 0.364 1.424 1.257 1.934 2.065 Chandipur −0.527 0.598 −0.172 0.218 1.173 0.701 0.058 2.048 Ramnagar−I 0.428 0.744 0.064 1.994 −2.853 0.459 0.687 1.523 Sutahata −0.748 1.325 0.199 −0.293 0.089 −0.725 1.567 1.415 Nandigram−II −0.052 1.461 −0.275 0.728 0.273 0.487 −1.216 1.407 Potashpur−II 1.488 −0.780 0.354 −0.276 0.516 0.096 −0.007 1.391 Khejuri−I 0.324 0.541 −0.887 1.003 1.304 −0.777 −0.212 1.296 Deshapran 1.068 0.167 −0.487 0.118 0.311 −0.542 0.653 1.288 Panskura −0.565 −0.978 1.670 −0.204 −0.618 0.987 0.507 0.798 Bhagawanpur−II 0.087 −0.438 −2.093 1.099 0.464 0.250 0.810 0.178 Haldia −0.217 1.551 0.044 −3.093 −0.674 1.684 0.518 −0.189 Definition and mapping of indicators of vulnerability to natural disaster 66 Source: Gayen et al, 2022 The goal of vulnerability analysis is to understand how the components of vulnerability are impacted by a flood. The result of this attempt focuses on the weaknesses of the society that cause threats in livelihood. According to the result shown in Figure 4.11, the Khejuri-II and Nandigram-I blocks are very vulnerable. High population growth rate, a large share of child and rural population, less number of agricultural society and membership, low number of banks/ 0.1 million population, etc. are the main driving factors of high vulnerability at Khejuri-II and Nandigram-I blocks. Most of the blocks of the district appear under the medium vulnerable category. Owing to the availability of better facilities and amenities, municipality bearing blocks, i.e., Tamluk and Contai-I, are categorized in the very low vulnerable zone. These two blocks have a low population growth rate, less of the share of the child population, a high concentration of female literacy, a lower number of cultivators and less agricultural labor. Nandakumar −0.865 −0.268 0.805 −0.020 0.567 0.107 −0.565 −0.238 Egra−I 2.015 −1.549 1.631 −0.961 −0.098 −1.276 −0.266 −0.504 Potashpur−I 0.199 −1.729 −0.309 0.048 −0.833 1.880 0.104 −0.639 Ramnagar−II 1.054 −0.183 −1.021 0.125 −1.078 −0.128 0.373 −0.858 Egra−II 1.072 −1.087 −0.318 −0.368 0.901 −1.155 0.046 −0.909 Bhagawanpur−I −0.593 −0.421 −0.429 −0.431 1.217 1.261 −1.708 −1.103 Mahishadal −0.530 0.282 −0.126 0.058 −0.451 −0.384 −0.064 −1.215 Sahid Matangini −1.752 −0.261 −0.060 0.607 −0.157 −1.627 0.500 −2.750 Kolaghat −1.212 −0.543 0.919 −0.023 −0.962 −1.260 0.185 −2.896 Contai−III 0.433 −0.138 −1.161 0.288 −0.800 0.430 −2.603 −3.552 Contai−I 0.301 0.557 −1.763 −1.986 −0.743 −1.058 0.534 −4.158 Tamluk −1.898 −0.374 0.529 −0.364 −0.577 −1.102 −1.842 −5.628 Results and discussion 67 Figure 4.11. Vulnerability map of Purba Medinipur district (Source: Gayen et al, 2022) C. Flood risk analysis of the study area Figure 4.12 represents the flood risk map, based on Eq. 1.1 overlayed with the flood hazard map and vulnerability condition. The zone number of flood hazard and the zone number of vulnerability were multiplied (Table 3.6) and then five risk classes were identified to be categorized into five different zones. Furthermore, 12% of blocks lie in the very high risk zone, 32% of blocks are in the high risk zone, 40% of blocks are in the medium risk zone, 12% of blocks stay in the low risk zone and only 4% of blocks lie in the very low risk zone. The map indicates that the northern and western part of the district are more risky to floods. Moyna and Patashpur-II, the two most risky blocks are situated in the western part. In addition, most of the high risk categorized blocks (Panskura, Nandakumar, Bhagawanpur-I, Chandipur, Patashpur-I and Egra-I) are located in the northern and western part of the district. Comparatively, the eastern part of the district is Definition and mapping of indicators of vulnerability to natural disaster 68 less flood risk prone. Mainly very low, low and medium types of flood risk zones are observed in this portion. Only two blocks (Khejuri-II and Nandigram-I) with high risk are located in the eastern part of the district. The map displays that one very high risk block, Ramnagar-I, is clustered in the southern part of the district. Moyna, Patashpur-II Figure 4.12. Flood risk map of Purba Medinipur district (Source: Gayen et al, 2022) and Ramnagar-I come under the very high risk category as the probability of flood is very high and also the blocks are socio-economically highly vulnerable. Flood risk is highest in the Moyna block of the Purba Medinipur district because it is in the very high flood hazard category and the high vulnerable category. Moyna is affected by flood almost every year. Contai-I is the least risky block within the district. The Contai-I block is low flood prone and very low vulnerable and this means it is relatively safe from flood risk. Results and discussion 69 Tamluk and Kolaghat blocks fall under the high flood prone zone. However, as they fall in very low (Tamluk) and low (Kolaghat) vulnerable zones, the risk level is low (Tamluk) and medium (Kolaghat), respectively. On the contrary, Nandigram-II has a high level of vulnerability, but flood hazard probability is very low. So, the Nandigram- II block appears to have a very low level of flood risk. Thus this study explains that hazard frequency not sufficient to identify risky region. Indicators of vulnerability are also important to identify the risky region and the value of indicators also effect to identify the risky region for any hazard. 4.2. Discussion This section provides a critical review of some of the results presented in the previous sections. This review is followed by a series of specific proposals that seek to improve the results obtained. Finally, some general aspects related to this type of vulnerability studies are discussed. 4.2.1. Purba Medinipur district In the case of Purba Medinipur, there is a significant discrepancy in the assignment of vulnerability classes between the different techniques applied (standardization and PCA), as can be seen in the maps produced (Figure 4.13). Figure 4.13. Vulnerability maps for Purba Medinipur district obtained by standardization and PCA techniques The following graph attempts to reflect these differences. Standardization PCA Definition and mapping of indicators of vulnerability to natural disaster 70 Figure 4.14. Differences between class assignment by the different applied techniques Figure 4.14 shows the classes assigned by each of the techniques, as well as the mean and the deviation between the scores received by each block. Overall, the three standardization techniques show some consistency, although the min-max value technique seems to perform worse than the other two; blocks typically only move between contiguous classes. The largest discrepancies appear with the scores assigned by PCA, with striking cases such as Sutahata. Figure 4.15. Regression graphs for the different techniques Results and discussion 71 Consistent with the appearance of previous graph, the correlations between PCA and each of the standardization techniques are very low (Figure 4.15), with the highest correlation being with the z score technique at 0.43 (0.1739 significance coefficient). Different reasons have been put forward to explain these differences. One of the main reasons is different applied methodology for measuring vulnerability (Bucherie, 2022). Yoon (2012) compared deductive approach (z score transformation, maximum value transformation, min–max rescaling transformation) and inductive approach (factor analysis by using PCA method). In his study name and number of indicators are fixed. The result of social vulnerability scores using three different standardization methods are highly correlated with each other (like our result), but result between standardization methods and factor analysis are quite different. Name and number of the indicators that are used in two analysis are also different. This is another important reason for changing the vulnerable blocks. For standardization methods 12 indicators have been used and PCA method analysis have been done by using 18 indicators. Aksha et al. (2019) and Gautam (2017) have worked with 2001 and 2011 data with different indicators for Nepal and their results are different. Another cause is using of data in different years, even if there is not much temporal distance. For PCA analysis we have used new data from district statistical handbook (2012-2015) and other source. For the particular case of Sutahata block a high disparity is found, which could respond to specific circumstances. The main industrial region of Purba Medinipur district is Haldia industrial belt (with petrochemicals and other industries). This is very near from Sutahata block (within few kms). Due to the lack of land in Haldia block, Sutahata block is rapidly changing. It has been observed that in the Sutahata block as a whole, percentage of households without electricity, percentage of households without sanitation, percentage of households without kitchen, IMR have reduced. IMR is an important health indicator. But it can change severely even within a year. Two new indicators have been addednumber of hospitals and bed number in hospitals. Sutahata block has high value for both of these indicators. Overall these conditions makes this block very high to very low vulnerable. 4.2.2. Malaga province Given that the discrepancies noted above are also evident in the case of the province of Malaga, and partly in order to better understand the implications of using different techniques for the calculation of vulnerability indicators. Three different techniques have Definition and mapping of indicators of vulnerability to natural disaster 78 this case is more linked to a low number of inhabitants, since this determines these higher ratios and not a high number of centers. As far as component 3 is concerned, population density, presence of foreigners, high level of education and income all point to urban environments close to the capital, which is also the case to some extent for component 4. Component 5 has been discarded, as it only includes the IMR, the significance of which is very low in an environment such as this, as will be discussed below. For a better understanding of the process explained, the situation of the municipalities is shown in relation to components or factors 1 (positive for vulnerability) and 3 (negative for vulnerability). If 2 municipalities are chosen (Figure 4.20) from two of the quadrants (Parauta and Torrox), the process of considering the components differently can be followed. In the case of Parauta, in the lower right quadrant, it obtains a value of 2 in factor 1, i.e. it presents data which makes it vulnerable according to the variables which govern this factor. However, with respect to factor 3, considered as an inhibitor of vulnerability, the value obtained is -2. If both values are added together, this municipality would obtain a value close to 0, probably corresponding to a medium vulnerability; on the other hand, if factor 3 is multiplied by -1, the result obtained would be 4, which would lead to a high vulnerability. Conversely, Torrox, in the opposite quadrant, obtains a negative value in the first factor, and a positive one in factor 3; following the same technique, in this case we have found a municipality that would obtain a negative vulnerability value reflecting a low level. Results and discussion 79 Figure 4.20. Variable contribution and position of municipalities in relation to factors 1 and 3 Finally, having considered 4 factors, the formula applied to obtain the vulnerability indicator is as follows: IV(PCA2) = F1 + F2 + [F3 * (-1)] + [F4 * (-1)] The resulting map is as follows (Figure 4.21): (Eq. 4.1) Definition and mapping of indicators of vulnerability to natural disaster 80 Figure 4.21. Resulting map by application of PCA2 As can be seen, the pattern of distribution of vulnerability has a certain territorial logic, with two environments (eastern Antequera and southern mountain range of Ronda) where the highest values are concentrated. On the contrary, it is the coastal areas and their proximities, together with the inland towns of Ronda and Antequera, which have the lowest levels. D. A new integrated vulnerability indicator In this last section, a new type of vulnerability indicator is designed, based on three different aspects.  The first of these encompasses territorial variables including physical characteristics (altitudes and slopes), population and accessibility (dispersed population and presence/ density of roads) and conditions of the built environment (age and condition of dwellings).  The second aspect refers to socio-demographic variables, largely already used in previous calculations (age, income, education etc.).  Finally, the existence of or distance to certain health services (hospitals) and security/ emergency services (security forces and fire stations) have been Results and discussion 81 considered as a way of approaching this essential factor for understanding the capacity to respond to disaster situations. For each case, PCA analyses have been carried out, applying the same technique as in the previous section. In each case, vulnerability classes have been established in the same way as throughout the work and, finally, the average vulnerability has been calculated for each municipality, which is considered as an integral vulnerability indicator. In the first case, the PCA analysis showed two opposing factors. In the first factor, considered as a positive contribution to vulnerability, the variables relating to the age (prior to 1950) and poor condition of dwellings, as well as high values for dispersed population and large municipal areas with slopes of more than 30% were weighted heavily; on the other hand, the variables relating to road accessibility (number of different roads which connect to a population center and road density) were weighted very little in this factor. In the second factor, which was considered to be negative for vulnerability, the weights were practically the opposite. The map obtained is as follows (Figure 4.22): Figure 4.22. Territorial vulnerability for an integrated vulnerability indicator Definition and mapping of indicators of vulnerability to natural disaster 82 With regard to social vulnerability, the process followed has been similar to that described in the previous section (4.4.2.C), although the consideration of different variables has led to some modifications (Figure 4.23). Figure 4.23. Social vulnerability for an integrated vulnerability indicator Finally, in the case of accessibility to health, security and emergency services, the analysis made it possible to characterize the municipalities according to the average distance of their population centers from this type of services considered essential in the event of the occurrence of extreme natural phenomena (Figure 4.24). Results and discussion 83 Figure 4.24. Emergency coping capacity for an integrated vulnerability indicator The result of obtaining the average vulnerability from the above maps provides the following map (Figure 4.25): Figure 4.25. Integrated vulnerability indicator Definition and mapping of indicators of vulnerability to natural disaster 84 E. Comparison of results Although it is clear that more work is needed on the definition and design of vulnerability indicators, it has been estimated that the last of the indices obtained, due to the number and diversity of variables considered, can adequately reflect this important condition at the municipal level. For this reason, the data corresponding to this last indicator have been used to compare them with the rest of the results obtained from other techniques. In a way, despite the need to go deeper into these questions, this comparison can be considered as a certain form of validation of the results, in the sense that it allows us to know which techniques yield consistent results despite of a certain variation in the variables selected. The first comparison is with the results obtained using the standardization technique. [Figures 4.26 a) and b)] As can be seen, there are very important differences between the classes assigned to the municipalities by the two approaches. Only 25 municipalities remain in the same class, while many move between two classes. In several cases there are jumps between 3 classes, the most striking being the case of Fuengirola, which moves from a very high level of vulnerability to a very low one. This is reflected in the scatter plot, with a coefficient of determination and correlation of 0.146 and -0.38, respectively. Results and discussion 85 Figure 4.26. Comparison among vulnerability classes by a) standardization and b) integrated vulnerability indicator a) b) Definition and mapping of indicators of vulnerability to natural disaster 86 Figure 4.27. Comparison among vulnerability classes by a) PCA1 and b) integrated vulnerability indicator a) b) Results and discussion 87 Figure 4.28. Comparison among vulnerability classes by a) PCA2 and b) integrated vulnerability indicator a) b) Definition and mapping of indicators of vulnerability to natural disaster 94 5.1. On the concept of vulnerability With regard to the theoretical framework of vulnerability, it has been possible to observe an extraordinary diversity of approaches and contexts from which this concept is addressed, and this is only limited to the field of natural risks. In a very synthetic way, it is possible to distinguish several points that are necessary when approaching a study on vulnerability to natural hazards. First of all, it is necessary to clarify whether the study is going to take into consideration any type of extreme natural phenomena or whether, on the contrary, a particular phenomenon or small group of phenomena with similar characteristics is being considered. In the first case we will talk about generic vulnerability, while in the second case we will refer to specific vulnerability (hazard specific). Secondly, the dimensions or areas on which the study focuses must be mentioned. In this respect, there is a great diversity of approaches, which, moreover, are combined in some cases. Thus, we speak of physical, social, economic or environmental vulnerability, although this list does not exhaust the fields for which it is possible to find particular analyses (technological vulnerability, patrimonial vulnerability, etc.). Finally, it is very important to define what is to be measured within each of these domains. Broadly speaking, these are exposure, sensitivity/ fragility, response capacity and resilience. The consideration of all these aspects is guided by the field of work and the objectives from which a study is approached and logically, greatly conditions the variables that will be taken into account when defining the vulnerability indicators. In this thesis, the work that has been addressed has focused mainly on the assessment of a generic vulnerability that tries to address primarily the sensitivity/ fragility from a social point of view. However, a specific vulnerability indicator has also been explored, in this case referring to the phenomenon of flooding, and some variables have been considered to measure response capacity besides sensitivity. The main conclusions that can be drawn from the work carried out relate to the following aspects:  It is extremely important to specify from the outset the type of vulnerability to be assessed, taking into account the different typologies and areas referred to above. This precision should allow for a more accurate selection of the variables to be used. Conclusion 95  In this sense, it seems much more convenient to opt for a specific type of vulnerability. Thus, although there are a series of territorial (isolation), socio/economic (ageing) or infrastructural (substandard housing) variables that adequately reflect fragility in the face of any extreme natural event, there is no doubt that knowledge of the behavior and main effects of a specific natural phenomenon (flood, earthquakes…) helps to make a better selection of the variables to be used. 5.2. On the techniques employed Two groups of techniques for obtaining indicators have been used throughout the work carried out: standardization techniques and techniques for the reduction of variable dimensions (PCA). In the case of standardization techniques, studies have been carried out comparing the performance of three different techniques: maximum value transformation, z-score transformation and min-max rescaling transformation. In all three cases, the positive or negative contribution of each variable to vulnerability is decided beforehand, so that in the formula for obtaining the indicator, the result of standardization is introduced in the correct sense. In the comparative studies, no major differences between the three types of techniques are apparent, although min-max rescaling transformation provides more robust results. In any case, recognizing the ease of use of this type of technique, it seems evident that it is too simple an approximation. Probably, the use of some kind of weighting in certain variables would make the use of this type of technique more effective. As far as variable dimension reduction technique (PCA) is concerned, it is clear that it offers better results, although it is more difficult to apply. In this work, two different types of applications have been carried out. The first is based on a standardization of the variables (min-max rescaling transformation) which implies their introduction into the PCA with the assignment of their contribution to vulnerability; the indicator is obtained by adding up the factors or components obtained. The second application involves the introduction of the original variables, leaving the PCA to carry out the standardization; the subsequent reading of the contribution of each variable to the factors obtained means that these factors as a whole are interpreted as positive or Definition and mapping of indicators of vulnerability to natural disaster 96 negative with respect to vulnerability, adding some and subtracting others to obtain the final indicator. The main conclusions in this respect are as follows:  Of the two groups of techniques employed, the use of PCA is much more advisable.  Once this technique has been chosen, it seems more appropriate to use the option of including the original variables, and to estimate the contribution of the factors once the influence of each variable has been analyzed.  As this is a complex technique, it is also appropriate to design synthetic indicators based on the integration of several sub-indicators that would be obtained by using groups of variables of the same type, whose smaller number and greater affinity would facilitate interpretation. 5.3. On the application of the concepts and techniques Some of the reflections that have just been expressed are logically derived from the concrete applications that have been carried out in the different works that make up this thesis. Based on these experiences and, above all, on their overall vision, it has been possible to carry out an integrated and critical analysis of the methodological processes followed and the results obtained, which is expressed to a large extent in the discussion section in chapter 4. In the case of Purba Medinipur district, the main causes of high vulnerability rates are related to poor public health services, significant deficiencies in household facilities and high population concentrations in predominantly rural settings. Examples of such circumstances include average infant mortality figures of 20.30, a population density of 1320, a rural population percentage of 89% or in the case of households, percentages of 54% of dwellings without electricity or 92% for the absence of sewage. Number of government hospitals are also very less, and private hospitals and nursing homes are mainly concentrated in municipality related blocks like Tamluk, Haldia, Egra- I and Contai-I, which show generally low levels of vulnerability. On the other hand, blocks such as Khejuri-II, Nandigram-I, Moyna, Nandakumar, show high vulnerability values irrespective of the method used to generate the indicators. As far as the case of the province of Malaga is concerned, the design of different strategies for the generation of vulnerability indicators has led to some interesting conclusions. Thus, apart from a more or less successful implementation of the Conclusion 97 techniques, the use of certain variables in a developed environment is highly debatable. Two cases are particularly striking. On the one hand, the IMR, which so well reflects certain precarious health and poverty conditions in India, is not very representative in a context such as Malaga; although its use as a long-term variable (for several years) may be interesting, its annual consideration has led to the introduction of significant inconsistencies in the data. On the other hand, population density and all those variables that imply a population distribution of endowments also require a certain amount of caution. Thus, in the first applications of this type of variable, the disproportionate weight of population densities on the coast compared to the more dispersed population in the interior of the province, led to an important bias of higher vulnerability towards the coastal strip; in the same way, primary health services are highlighted in the interior, due to the per capita distribution in very small populations; this again reinforces the previous bias. With regard to the issue of population density, it is clear that its consideration needs to be given serious thought. Thus, while in developing environments its contribution to vulnerability is obvious (overcrowding, environmental deterioration etc.), in developed regions its importance is not so clear. On the one hand, if economic losses are taken as a central issue of vulnerability, urban centers will concentrate a greater number of goods whose damage will entail high losses; the same applies to the greater concentration of people who, in the first instance, may entail a greater number of victims. On the other hand, orderly urban environments, with modern infrastructures and a high concentration of public services, may in many circumstances contribute to lower personal damage, while rural environments with ageing populations and poorly communicated and served may be areas prone to personal losses at times of the action of certain extreme natural phenomena (floods, snowfalls, forest fires...). 5.4. Some proposals for future works As mentioned in the discussion, the work carried out and its critical analysis allow us to advance some lines of future works that may contribute to the achievement of more solid results. Some of these suggestions are listed below, in addition to those that have been specified in the previous conclusions.  Development of specific vulnerability studies that select a phenomenon (natural hazard) or group of phenomena with similar characteristics.  Based on the spatial scope of these phenomena, determine the area of study and the possible scales of work or reference units that can be adopted. Definition and mapping of indicators of vulnerability to natural disaster 98  For the selection of the variables that will generate the indicators, one should go beyond the simple use of references to explore real cases of disasters in the area of study or in equivalent environments, in order to try to analyse the territorial, socio-economic, endowment or socio-political organisation aspects that best explain the distribution of the damage and losses observed.  Inventory, from available data sources, the set of variables that can be used and, if necessary, generate new variables, especially of a spatial nature.  Group the selected variables into sets that represent different components, dimensions or types of vulnerability.  For each of these sets, study the convenience of increasing or decreasing their contribution to vulnerability on the basis of weightings.  Select more than one technique for the generation of indicators, so that the comparison of the results obtained demonstrates a certain robustness and consistency. Among these, the PCA is highly recommended.  Update the studies developed as new data become available and try to validate the results obtained when there is an opportunity to contrast the results with real disaster event situations. References 1. Adger, WN. (1999): Social vulnerability to climate change and extremes in coastal Vietnam. World Development, 27(2): 249-269. 2. Adger, WN. (2006): Vulnerability. Global Environmental Change, 16, 268-281. 3. Aguirre-Ayerbe, I.; Sánchez, JM.; Aniel-Quiroga, Í.; González-Riancho, P.; Merino, M.; Al-Yahyai, S.; González, M.; Medina, R. (2018): From tsunami risk assessment to disaster risk reduction—The case of Oman. Natural Hazards and Earth System Sciences,18, 2241-2260. 4. Ahmad, HF.; Bhat, MS.; Alam, A.; Ahmad, S. (2016): Flood hazard zonation and vulnerability assessment of Greater Srinagar, J&K, India. International Journal of Advanced Research, 4, 1679- 1690. 5. Aksha, SK.; Juran, L.; Resler, LM.; Zhang, Y. (2019): An analysis of social vulnerability to natural hazards in Nepal using a modified SVI. International Journal of Disaster Risk Science, 10, 103-116. 6. André, G. (2004): Cartographie du risque naturel dans le monde. Etude comparative entre une approche d’ordre social et une approche d’ordre économique de la vulnérabilité. Environnement, Nature, Paysage. https://doi.org/10.4000/cybergeo.2614 7. Annual Flood Report (2002–2019): Irrigation and Waterways Directorate, Government of West Bengal, Kolkata, India. 8. Armenakis, C.; Du, EX.; Natesan, S.; Persad, RA.; Zhang, Y. (2017): Flood risk assessment in urban areas based on spatial analytics and social factors. Geosciences, 7, 123. 9. Asare-Kyei, D.; Forkuor, G.; Venus, V. (2015): Modeling flood hazard zones at the sub-district level with the rational model integrated with GIS and remote sensing approaches. Water, 7, 3531-3564. 10. Bahinipati, CS. (2014): Assessment of vulnerability to cyclones and floods in Odisha, India: A district-level analysis. Current Science, 107, 1997-2007. 11. Baker, JL. (2012): Climate Change, Disaster Risk, and the Urban Poor: Cities Building Resilience for a Changing World. Urban Development. Washington, DC: World Bank. https://openknowledge.worldbank.org/handle/10986/6018 License: CC BY 3.0 IGO. 12. Balica, S. (2012): Approaches of understanding developments of vulnerability indices for natural disasters. Environmental Engineering and Management Journal, 11(5). 13. Barroca, B.; Bernardara, P.; Mouchel, JM.; Hubert, G. (2006): Indicators for identification of urban flooding vulnerability. Natural Hazards and Earth System Sciences, 6, 553-561. 14. Beccari, B (2016): A comparative analysis of disaster risk, vulnerability and resilience composite indicators. PLOS Currents Disasters. doi: 10.1371/currents.dis.19f9c194f3e3724d9ffa285b157c6ee3. 15. Bergstrand, K.; Mayer, B.; Brumback, B.; Zhang, Y. (2015): Assessing the relationship between social vulnerability and community resilience to hazards. Social Indicators Research, 122, 391-409. 16. Bernard, GS. (2007): Measuring social vulnerability in Caribbean States. 8th SALISES Annual Conference Crisis, Chaos and Change: Caribbean Development Challenges in the 21st Century, Trinidad and Tobago, 26-28 March, 2007. https://sta.uwi.edu/conferences/salises/documents/St%20Bernard%20%20G.pdf 17. Birkmann, J. (2006): Measuring Vulnerability to Promote Disaster-resilient Societies: Conceptual Frameworks and Definitions. In J. Birkmann (Ed.), Measuring Vulnerability to Natural Hazards: Towards Disaster Resilient Societies, Tokyo: United Nations University Press, 9-54 pp. Definition and mapping of indicators of vulnerability to natural disaster 100 18. Boruff, BJ.; Emrich, C.; Cutter, SL. (2005): Erosion hazard vulnerability of US Coastal Counties. Journal of Coastal Research, 21(5): 932-942. 19. Boruff, BJ.; Cutter, SL. (2007): The environmental vulnerability of Caribbean Island Nations. Geographical Review, 97(1): 24-45. 20. Bucherie, A.; Hultquist, C.; Adamo, S.; Neely, C.; Ayala, F.; Bazo, J.; Kruczkiewicz, A. (2022): A comparison of social vulnerability indices specific to flooding in Ecuador: Principal component analysis (PCA) and expert knowledge. International Journal of Disaster Risk Reduction, 73, 102897. 21. Burton, CG. (2010): Social vulnerability and hurricane impact modeling. Natural Hazards Review, 11(2): 58-68. 22. Buta, C.; Mihai, G.; Stănescu, M. (2020): Flood risk assessment based on flood hazard and vulnerability indexes. Ovidius University Annals Series: Civil Engineering, 22, 127-137. 23. Caliskan, S.; Taubenböck, H.; Hinz, S.; Roth, A. (2006): Earthquake vulnerability indicators and vulnerability assessment using remote sensing, Istanbul. 1st EARSeL Workshop of the SIG Urban Remote Sensing. Humboldt-Universität zu Berlin, Germany, 2-3 March, 2006. 24. Cançado, V.; Brasil, L.; Nascimento, N.; Guerra, A. (2008): Flood risk assessment in an urban area: Measuring hazard and vulnerability. 11th International Conference on Urban Drainage, Edinburgh, Scotland, UK, 31st August–5th September 2008. 25. Census of India (2011): District Census Handbook, Purba Medinipur. Village and Town Wise Primary Census Abstract. Directorate of Census Operations, West Bengal, Series-20, Part XII-B. http://censusindia.gov.in/2011census/dchb/1919_PART_B_DCHB_PURBA%20MEDINIPUR.pdf 26. Chakraborty, J.; Tobin, GA.; Montz, BE. (2005): Population evacuation: Assessing spatial variability in geophysical risk and social vulnerability to natural hazards. Natural Hazards Review, 6(1): 23-33. 27. Chan, YH. (2003): Biostatistics 104: Correlational analysis. Singapore Medical Journal, 44(12): 614- 619. 28. Chen, W.; Cutter, SL.; Emrich, CT.; Shi, P. (2013): Measuring social vulnerability to natural hazards in the Yangtze River Delta Region, China. International Journal of Disaster Risk Science, 4(4): 169- 181. 29. Contreras, D.; Chamorro, A.; Wilkinson, S. (2020): Review article: The spatial dimension in the assessment of urban socio-economic vulnerability related to geohazards. Natural Hazards and Earth System Sciences, 20: 1663-1687. 30. Coudouel, A.; Hentschel, J. (2000): Poverty Data and Measurement. Preliminary Draft for a Sourcebook on Poverty Reduction Strategies. The World Bank, Washington DC. 31. Crawford, CB. (1975): Determining the number of interpretable factors. Psychological Bulletin, 82(2): 226-237. 32. Cutter, SL. (1993): Living with Risk: The Geography of Technological Hazards. Edward Arnold, London, 214 pp. 33. Cutter, SL.; Mitchell, JT.; Scott, MS. (2000): Revealing the vulnerability of people and places: A case study of Georgetown County, South Carolina. Annals of the Association of American Geographers, 90(4): 713-737. 34. Cutter, SL.; Boruff, BJ.; Shirley, WL. (2003): Social vulnerability to environmental hazards. Social Science Quarterly, 84(2): 242-261. Reference 101 35. Cutter, SL.; Emrich, CT. (2006): Moral hazard, social catastrophe: The changing face of vulnerability along the Hurricane coasts. The Annals of the American Academy of Political and Social Science, 604, 102-112. 36. Cutter, SL.; Finch, C. (2008): Temporal and spatial changes in social vulnerability to natural hazards. Environmental Sciences, 105 (7): 2301-2306. 37. Cutter, SL.; Burton, CG.; Emrich, CT. (2010): Disaster resilience indicators for benchmarking baseline conditions. Journal of Homeland Security and Emergency Management, 7, 51. 38. Das, SC.; Sahu, AS. (2017): Flood zonation mapping using RS and GIS techniques: A case study in Panskura of Purba Medinipur district, West Bengal. Indian Journal of Landscape Systems and Ecological Studies, 40, 126-135. 39. de León, JCV. (2011): Risks in Central America: Bringing Them Under Control. In:, et al. Coping with Global Environmental Change, Disasters and Security. Hexagon Series on Human and Environmental Security and Peace, vol 5. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978- 3-642-17776-7_70 40. de Mello Rezende, GB. (2016): Social vulnerability index: A methodological proposal for application in the cities of Barra do Garcas—MT, Pontal Do Araguaia—MT and Aragarcas—GO, Brazil. Open Journal of Social Sciences, 4, 32-45. 41. de Oliveira Mendes, JM. (2009): Social vulnerability indexes as planning tools: Beyond the preparedness paradigm. Journal of Risk Research, 12(1): 43-58. 42. Diamond, J. (2005): Collapse : How Societies Choose to Fail or Succeed. Penguin Books, New York, USA, 589 pp. 43. di Girasole, EG.; Cannatella, D. (2017): Social vulnerability to natural hazards in urban systems. An application in Santo Domingo (Dominican Republic). Sustainability, 9, 2043. 44. District Disaster Management Plan, Purba Medinipur, (2012-20): District Disaster Management Section, Purba Medinipur. http://wbdmd.gov.in/pages/district_dm_plan.aspx 45. District Human Development Report, Purba Medinipur (2011): Development & Planning Department, Government of West Bengal. https://www.wbpspm.gov.in/publications/District%20Human%20Development%20Report 46. District Statistical Handbook, Purba Medinipur (2005-2015): Bureau of Applied Economics & Statistics, Department of Statistics & Programme Implementation, Government of West Bengal: Kolkata, India. http://wbpspm.gov.in/publications/District%20Statistical%20Handbook 47. Dodman, D. (2009). United Nations Population Fund (UNFPA) Analytical Review of the Interaction between Urban Growth Trends and Environmental Changes Paper 1 URBAN DENSITY AND CLIMATE CHANGE 1. 48. Dwyer, A.; Zoppou ,C.; Nielsen, O.; Day, S.; Roberts, S. (2004): Quantifying social vulnerability: A methodology for identifying those at risk to natural hazards. Geoscience Australia Record, 2004/14. 49. Emrich, CT.; Cutter, SL. (2011): Social vulnerability to climate-sensitive hazards in the Southern United States. Weather, Climate, and Society, 3(3): 193-208. 50. ESPON (2021): ESPON-TITAN Territorial Impacts of Natural Disasters. ESPONTITAN_Scientific%20Report-Annex%203_Vulnerability%20Analysis.pdf 51. Evans, JM.; Hauer, M.; Hardy, D.; Pippin, JS. (2014): Assessing social vulnerability using ‘‘SoVI- Lite’’: A demonstration study at Glynn County, GA. Athens: Carl Vinson Institute of Government, University of Georgia. DOI: 10.13140/2.1.2329.3765 Definition and mapping of indicators of vulnerability to natural disaster 102 52. Fabrigar, LR.; Wegener, DT.; MacCallum, RC.; Strahan, EJ. (1999): Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods, 4(3): 272-299. 53. Fekete, A. (2009): Validation of a social vulnerability index in context to river-floods in Germany. Natural Hazards and Earth System Sciences, 9: 393-403. 54. Fekete, A. (2019): Social Vulnerability (Re-)Assessment in Context to Natural Hazards: Review of the Usefulness of the Spatial Indicator Approach and Investigations of Validation Demands. International Journal of Disaster Risk Science, 10, 220–232. 55. Flanagan, BE.; Gregory, EW.; Hallisey, EJ.; Heitgerd, JL.; Lewis, B. (2011): A social vulnerability index for disaster management. Journal of Homeland Security and Emergency Management, 8(1): 3. 56. Gautam, D.(2017): Assessment of social vulnerability to natural hazards in Nepal. Natural Hazards and Earth System Sciences, 17, 2313-2320. 57. Gayen, S.; Villalta, IV.; Haque, SM. (2020): Comparative social vulnerability assessment in Purba Medinipur District, West Bengal, India. European Journal of Geography, 11(1): 093-108. 58. Gayen, S.; Villalta, IV.; Haque, SM. (2021): Assessment of social vulnerability in Malaga Province, Spain: A comparison of indicator standardization techniques. Revista de Estudios Andaluces, 41, 87- 108. 59. Gayen, S.; Villalta, IV. (2022): Assessment on Social Vulnerability to Adapt the Hindrances of Natural Hazards in Purba Medinipur District, West Bengal, India. In: Shit, PK.; Pourghasemi, HR.; Bhunia, GS.; Das, P.; Narsimha, A. (Eds.), Geospatial Technology for Environmental Hazards. Advances in Geographic Information Science. Springer. https://doi.org/10.1007/978-3-030-75197- 5_11 60. Gayen, S.; Villalta, IV.; Haque, SM. (2022): Flood Risk Assessment and Its Mapping in Purba Medinipur District, West Bengal, India. Water, 14, 1049. 61. Ge, Y.; Dou, W.; Dai, J. (2017): A new approach to identify social vulnerability to climate change in the Yangtze River delta. Sustainability, 9, 2236. 62. Gilard, O.; Givone, P. (1997): Flood risk management: New concepts and methods for objective negotiations. In Leavesley, GH.; Lins, HF.; Nobilis, F.; Parker, RS.; Schneider, VR; van de Ven, FHM. (Eds.), Destructive Water: Water-caused Natural Disasters, their Abatement and Control, IAHS Press, Oxfordshire, 145-155. 63. Ginige, K.; Amaratunga, D.; Haigh, R. (2014): Tackling women's vulnerabilities through integrating a gender perspective into disaster risk reduction in the built environment. Procedia Economics and Finance, 18, 327-335. 64. Girasole, EG.; Cannatella, D. (2017): Social vulnerability to natural hazards in urban systems. An application in Santo Domingo (Dominican Republic). Sustainability, 9(11): 2043. 65. Guha-Sapir, D.; Vos, F. (2011): Earthquakes, an Epidemiological Perspective on Patterns and Trends. In: Spence, R.; So, E.; Scawthorn, C. (Eds.), Human Casualties in Earthquakes. Advances in Natural and Technological Hazards Research, vol 29. Springer, Dordrecht. https://doi.org/10.1007/978-90-481-9455-1_2 66. Guillard-Gonçalves, C.; Cutter, SL.; Emrich, CT.; Zêzere, JL. (2015): Application of Social Vulnerability Index (SoVI) and delineation of natural risk zones in Greater Lisbon, Portugal. Journal of Risk Research, 18(5): 651-674. Reference 103 67. Hazarika, N.; Barman, D.; Das, AK.; Sarma, AK.; Borah, SB. (2018): Assessing and mapping flood hazard, vulnerability and risk in the Upper Brahmaputra River valley using stakeholders’ knowledge and Multicriteria Evaluation (MCE). Journal of Flood Risk Management, 11, S700-S716. 68. HDR (2006): Beyond Scarcity: Power, Poverty and the Global Water Crisis. United Nations Development Programme, New York, USA. http://hdr.undp.org/sites/default/files/reports/267/hdr06- complete.pdf 69. Heß, VDC. (2017): Weigh(t)ing the dimensions of social vulnerability based on a regression analysis of disaster damages. Natural Hazards and Earth System Sciences. doi.org/10.5194/nhess-2017-74 70. Hoeppe, P. (2016): Trends in weather related disasters – Consequences for insurers and society. Weather and Climate Extremes, 11, 70-79. 71. Holand, IS.; Lujala, P.; Rød, JK. (2011): Social vulnerability assessment for Norway: A quantitative approach. Norsk Geografisk Tidsskrift - Norwegian Journal of Geography, 65(1): 1-17. 72. Horn, JL. (1965): An empirical comparison of methods for estimating factor scores. Educational and Psychological Measurement, 25(2): 313-322. 73. Hufschmidt, G. (2011): A comparative analysis of several vulnerability concepts. Natural Hazards, 58, 621–643. 74. Hummell, BML.; Cutter, SL.; Emrich, CT. (2016): Social Vulnerability to Natural Hazards in Brazil. International Journal of Disaster Risk Science, 7, 111–122. 75. Instituto de Estadística y Cartografía de Andalucía (IECA,) Junta de Andalucía. http://www.juntadeandalucia.es/institutodeestadisticaycartografia/iea/consultasActividad.jsp?CodOpe r=104&sub=38120 76. IPCC (2001): Climate Change 2001: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change, [McCarthy, JJ.; Canziani, OF.; Leary, NA.; Dokken, DJ.; White, KS. (Eds.)], Cambridge University Press, Cambridge, UK. 77. IPCC (2007): Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, [Parry, ML.; Canziani, OF.; Palutikof, JP.; van der Linden, PJ.; Hanson CE. (Eds.)], Cambridge University Press, Cambridge, UK, 976 pp. 78. IPCC (2021): Climate Change 2021: The Physical Science Basis. Working Group I Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change, [Masson-Delmotte, V.; Zhai, P.; Pirani, A.; Connors, SL.; Péan, C.; Berger, S.; Caud, N.; Chen, Y.; Goldfarb, L.; Gomis, MI.; Huang, M.; Leitzell, K.; Lonnoy, E.; Matthews, JBR.; Maycock, TK.; Waterfield, T.; Yelekçi, O.; Yu, R.; Zhou, B. (Eds.)], Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 2391 pp. 79. ISDR (2004): Living with Risk. In A Global Review of Disaster Reduction Initiatives; United Nations: Geneva, Switzerland, Volume 1. https://www.undrr.org/publication/living-risk-global- review-disaster-reduction-initiatives 80. Jelínek, R.; Eckert, S.; Zeug, G.; Krausmann, E. (2009): Tsunami Vulnerability and Risk Analysis Applied to the City of Alexandria, Egypt. JRC Scientific and Technical Reports, Tsunami Risk ANd Strategies For the European Region (TRANSFER). https://publications.jrc.ec.europa.eu/repository/handle/JRC53316 This page is intentionally left blank 1. Gayen, S.; Villalta, IV.; Haque, SM. (2019): Earthquake hazard scenario and seismic micro-zoning assessment in Málaga province of Spain. European Journal of Geography, 10(1): 117-133. http://www.eurogeographyjournal.eu/articles/0810012019.pdf Impact factor: 0.272 (SCOPUS SJR) Social Sciences: Geography, Planning and Development: 398/787 (Q3) 2. Gayen, S.; Villalta, IV.; Haque, SM. (2020): Comparative social vulnerability assessment in Purba Medinipur district, West Bengal, India. European Journal of Geography, 11(1): 093-108. https://doi.org/10.48088/ejg.s.gay.11.1.93.107 Impact factor: 0.285 (SCOPUS SJR) Social Sciences: Geography, Planning and Development: 396/818 (Q3) 3. Gayen, S.; Villalta, IV.; Haque, SM. (2021): Assessment of social vulnerability in Malaga Province, Spain: A comparison of indicator standardization techniques. Revista de Estudios Andaluces, 41, 87-108. https://dx.doi.org/10.12795/rea.2021.i41.05 Impact factor: 0.34 (JCI) Emerging Sources Citation Index (ESCI) Sello de calidad FECYT 4. Gayen, S.; Villalta, IV.; Haque, SM. (2022): Flood Risk Assessment and Its Mapping in Purba Medinipur District, West Bengal, India. Water, 14, 1049. https://doi.org/10.3390/w14071049 Impact factor: 3.53 (WOS JCR) Water Resources: 36/100 (Q2) Book chapter 5. Gayen, S.; Villalta, IV. (2022): Assessment on Social Vulnerability to Adapt the Hindrances of Natural Hazards in Purba Medinipur District, West Bengal, India. In: Shit, PK.; Pourghasemi, HR.; Bhunia, GS.; Das, P.; Narsimha, A. (Eds.), Geospatial Technology for Environmental Hazards. Advances in Geographic Information Science. Springer. 235-262 pp. https://doi.org/10.1007/978-3-030-75197-5_11 Impact factor: SPI Q1 111 1. Gayen, S.; Villalta, IV.; Haque, SM. (2019): Earthquake hazard scenario and seismic micro-zoning assessment in Málaga province of Spain. European Journal of Geography, 10(1): 117-133. http://www.eurogeographyjournal.eu/articles/0810012019.pdf Impact factor: 0.272 (SCOPUS SJR) Social Sciences: Geography, Planning and Development: 398/787 (Q3) 112 !"#$%&'()*$"#('+)$,)-&$.#'%/0)1$+"2&)345)6"27&#)38339:3;;5)<'#=/ >43? !@AA$=B'CB$()$,)!"#$%&'()-&$.#'%/&#A) !"#$%&'()*$"#('+)$,)-&$.#'%/0 : DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I !!" EARTHQUAKE HAZARD SCENARIO AND SEISMIC MICRO-ZONING ASSESMENT IN MÁLAGA PROVINCE OF SPAIN Sumita GAYEN Universidad De Sevilla, Departamento de Geografía Física y Análisis Geográfico Regional, Sevilla, Spain [email protected] Ismael V. VILLALTA Universidad De Sevilla, Departamento de Geografía Física y Análisis Geográfico Regional, Sevilla, Spain [email protected] Sk. M. HAQUE University of Calcutta, Department of Geography, Kolkata, India [email protected] Abstract Spain is considered as a land of moderate seismic activity compared to other Mediterranean countries. Seismic activity is more intense in the south and eastern part of Spain. Málaga province is situated in the southern part of Spain and located near the boundary between the African and Eurasian plates. Málaga has experienced strong and frequent earthquakes in past, caused severe damages. To measure the effect of earthquake ‘peak ground acceleration (PGA)’ has commonly been used by the scientists. Here, in this study the seismic zone map of Málaga province has been created depending on available PGA data (return period of 475 years). The obtained seismic micro-zoning is important for assessing hazard prediction and urban regional planning in the earthquake prone areas of Málaga. The result shows the existence of spatial variation of vulnerability of earthquake in the province, which increases from west to east of Málaga. Keywords: Earthquake susceptibility, peak ground acceleration, seismic zone, Málaga province 1. INTRODUCTION An earthquake is rapid vibrations of earth’s surface resulting from the sudden release of energy from the earth’s asthenosphere. According to plate tectonic theory the plate i.e. the segmented crust of the earth is constantly moving in various directions with different velocity. This movement can create stress which is being considered as main root cause of earthquake. Earthquake can be smaller or larger. Similarly, it is a natural phenomenon, but sometimes human activities can also lead to its magnitude. Earthquakes can last from a few seconds to several 113 Gayen S. – I.V. Villalta / European Journal of Geography 10 1 117–133 (2019) !"#$%&'()*$"#('+)$,)-&$.#'%/0):)DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I" !!#$ minutes. It occurs suddenly and it can happen at any time and at any place of its susceptive area. Earthquake can occur on land or under the sea floor. Strong earthquake can damage buildings, roads, bridges and even change the direction of river within a few minutes. Many people also loss their lives due to the earthquake damage. Seismic risk assessment is very complex phenomenon because of direct or indirect, immediate, or long-lasting effects of the earth surface (De Pascale et al, 2015). Normally plate margins seem to be earthquake prone region. The southern part of Spain is located at the boundary between Eurasian plate and African plate. Since the Cretaceous African plate is moving towards the Eurasian Plate (Di Bucci et al, 2010). This is a convergent plate boundary where two plates are colliding. When two plates collide, one plate diving beneath the other and into Earth's mantle. The boundary which marks the collision between two plates along with downward movement is known as subduction zone. In this case, the African plate is diving below the Eurasian plate. So, in southern region, earthquakes are more frequent than other parts of Spain. Here the two plates collide very slowly, 2 to 5 mm/year in the NW-SE to WNW-ESE directions (source: Solares, 2016; Actualización de mapas de peligrosidad sísmica de España, 2012). When two plates collide, initially a fault is being created. The fault releases sudden energy that accumulates along the margin surrounding and finally earthquake occurs. Earthquakes happen only on that time, when fault builds up highest tensions (Parsons, 2008) and crossed the threshold limit of the surround rocks. There are two major right-lateral strike-slip fault systems in the study area; a) Maro-Nerja fault (NW – SE) and b) Yusuf fault (WNW – ESE), and those are presently active in the Alboran sea region, located near Málaga province (Alvarez-Marrón, 1999; Fernández-Ibáñez et al, 2007). The intensity of earthquake is near to moderate in Málaga province, but large earthquakes also occurred in this region (Casado et al, 2001; Grácia et al, 2006). The Granada-Málaga-West of Alboran zone has the vast majority of superficial earthquakes with epicenters (h) less than 50 km, but a large number of earthquakes have also been generated at intermediate depths of 50-200 km (Solares, 2016). Several kinds of maps are used to analyze the earthquake hazard; likeintensity map, acceleration map etc. Intensity map represents isolines of the expected intensities for a certain return period. Acceleration is defined as changes of velocity per unit time. During the time of earthquake when the ground is shaking, it also experiences acceleration. Acceleration map shows the probability that the expected PGA will be exceeded for a certain return period by the isolines. PGA is the largest increase in velocity recorded by a particular station during an earthquake. It is the most commonly used ground motion parameter to estimate seismic hazard in a region. The purpose of this article is to classify Málaga province in micro seismic zones depending on PGA value and identify the vulnerable zones. Results of the study could potentially be used for risk analysis and to build earthquake resistance structure in Málaga province. 2. STUDY AREA Málaga is located in southern Spain, under autonomous community of Andalucia (Fig. 1). It is situated on northern side of the Mediterranean Sea. It is surrounded by the Cadiz province to the west, Sevilla and Cordoba provinces to the north, the Alhama Mountains separate Málaga from Granada to the east and Mediterranean Sea to the south. The province contains 103 municipalities and the capital of the province is Málaga. Geographically Málaga Province is situated between 36°3′N - 37°29′N and 3°76′W - 5°65′W longitude. 114 !"#$%&'()*$"#('+)$,)-&$.#'%/0)1$+"2&)345)6"27&#)38339:3;;5)<'#=/ >43? !@AA$=B'CB$()$,)!"#$%&'()-&$.#'%/&#A) !"#$%&'()*$"#('+)$,)-&$.#'%/0 : DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I !!% Figure 1. Location map of Málaga province The province of Málaga has over 160 kilometres of coastline, which falls under famous Costa del Sol (Coast of Sun). The most of beaches in both the eastern and the western part of the province are sandy and sunny. Many municipalities are situated above 500 meter height. The Mountains in Málaga are steep and rocky with peaks of around 500 meters. The highest peak of Málaga province is La Maroma at a height of 2,066 meter. It is located in the Natural Park of the Sierras Tejeda, Almijara and Alharma. Guadalhorce River is the main river of the province of Málaga, both in terms of its length and the size of its catchment area. Other rivers of the province are Guadalmedina, Guadiaro, Genal, Rio Grande, Chillar, Rio Velez, Hozgarganta etc. The climate of the province is Mediterranean type. Málaga enjoys an average of more than 320 sunny days in each year. Warmest months are July and August and sometimes temperature reaches over 40°C. The winter is rainy and sometimes minimum temperature falls near to 0°C. The population of Málaga is 1,594,808 (2011) (Intituto de Estadística y Cartografía de Andalucía) and population density is 218.20/ sq. km (2011). The female population is more than male population. Male: female ratio of Málaga province is 1000:1025 in 2011. Málaga province 115 Gayen S. – I.V. Villalta / European Journal of Geography 10 1 117–133 (2019) !"#$%&'()*$"#('+)$,)-&$.#'%/0):)DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I" !&'$ has a high literacy rate. The average literacy rate is 96.50% in 2001. In the past, Málaga’s economy was based on various industries such as agriculture and textiles and it was even reported to be one of the richest provinces in the 19th Century. At present, the most important sector of economy in Málaga province is tourism. Other important economic sectors are construction, agriculture and technology services. 3. OBJECTIVES It is essential to have at least one seismic hazard map of a region that can determine the possibility of occurrence of earthquake damage. The map is also capable of generating the spatial identity regarding the personal injury and property damage. Thus the main objective of this study is to prepare a seismic micro-zonation map of the Málaga province depending on PGA values (return period of 475 years) that will help to identify areas at susceptibility of earthquake. This map will help to estimate the probability of the earthquake generated hazard for each municipality of Málaga and to take remedial measures to minimize elements at risks and to avoid losses occurring due to earthquakes. Also another objective is to analyze the history of earthquake in Málaga province that involves the identification and description of the devastating earthquakes along with the trends of recent earthquake in the study area. Earthquake catalog is considered to be very useful for seismicity analysis and evaluation of earthquake risk. In Spain, earthquakes are concentrated mainly in the southern region (including Málaga province). Here the causes of earthquake in Málaga province have also been discussed. 4. MATERIAL AND METHODS There have been several methods to determine and analyze hazards of earthquake; Ex. by analyzing seismic zone and earthquake hazard micro-zonation (depending on spectral amplification or peak ground acceleration value), adopting Neo-Deterministic Seismic Hazard Analysis (N-DSHA) method, and as well as using Remote Sensing (Sari et al, 2017). This work is based on municipality wise secondary data. To know the history of earthquake in the province, a lot of literature surveys have been conducted. The hazard map of Spain has been collected from Instituto Geográfico Nacional (IGN). To characterize seismic action, specifically three types of data have been utilized: 1. History of past earthquakes: To estimate the seismic hazard values in any region, past earthquake data with a magnitude scale is essential (Vipin et al, 2009). The seismic catalogue of IGN is the source of history of earthquake (1581 – 2006) data for Málaga province. A catalogue for intensity ≥V or magnitude ≥ 4 has been considered to generate a broad idea of the seismic activity of the province. 2. Recent earthquakes: Data of all earthquakes that have epicenters in Málaga province over past twelve years are highlighted. 3. PGA values: PGA is widely adopted parameter in seismic hazard assessments. Several authors (Makropoulos et al, 1985; Irwansyah et al, 2013; Abdalla et al, 2004) have used this parameter to identify the seismic zone. The ‘Actualización de mapas de peligrosidad sísmica de España 2012’ is the main source of PGA data for a seismic hazard study in Málaga province. Based on PGA value, Málaga province has been classified into five different vulnerable zones. 116 !"#$%&'()*$"#('+)$,)-&$.#'%/0)1$+"2&)345)6"27&#)38339:3;;5)<'#=/ >43? !@AA$=B'CB$()$,)!"#$%&'()-&$.#'%/&#A) !"#$%&'()*$"#('+)$,)-&$.#'%/0 : DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I !&! 5. DISCUSSION 5.1 Scenario of Earthquake in Spain Probably first seismic hazard map of Spain was made by the Gómez-Guillamón (1957) and divided the Iberian Peninsula (includes the countries of Andorra, Portugal, Spain and the British Crown colony of Gibraltar and small part of France) into 17 areas assigning each a maximum seismic degree and frequency (Actualización de mapas de peligrosidad sísmica de España, 2012). Spain is considered to be a country of moderate to low seismic activity with relatively low seismic hazard in world’s perspective (Gaspar-Escribano et al, 2008). Seismic fault activity rate in Spain is of moderate type (including the historic period) (Gaspar-Escribano et al, 2010). The seismic activity is more intense in south-east area, as displayed in the map of seismicity of the National Geographic Institute (Fig. 2). The highest acceleration (> 0.20) is concentrated in two regions of south-east area of Spain. Málaga, Granada, Murcia and Alicante are the four provinces which are located in these high intensity regions. The lowest values are recorded in the whole Avila province and in parts of Palencia, Burgos, Zamora, Valladolid, Segovia, Salamanca, Guadalajara, Madrid, Toledo, Caceres, Badajoz and Ciudad Real provinces. Hazard map shows the acceleration to a return period of 475 years, which corresponds to a probability of 10% at 50 years. Generally, a destructive earthquake is recorded in every 70 years in Spain (Sanchez, 1986). Figure 2. Seismic hazard map of Spain in PGA values for a return period 475 years according to the Updating seismic hazard maps of Spain 2012 (CNIG 2013). Revised version: October 2015. Source: Instituto Geográfico Nacional Table no. 1 shows some significant earthquakes in the country. The last moderate earthquake happened on 11th May, 2011 in Lorca, situated in south-eastern part of the country. More than 300 people were injured and around 462 million Euros economic loss were reported due to this earthquake (Monsó et al, 2018). 117 Gayen S. – I.V. Villalta / European Journal of Geography 10 1 117–133 (2019) !"#$%&'()*$"#('+)$,)-&$.#'%/0):)DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I" !&&$ Table 1. List of some significant earthquakes in Spain (Since 1700) Date Latitude Longitude Epicenter Magnitude 23/03/1748 39.0333 -0.6333 Estubeny (Valencia) 6.2 01/11/1755 36.5000 -10.0000 SW. Cabo San Vicente 8.5 13/01/1804 36.0833 -3.5833 Sea of Alboran 6.7 25/08/1804 36.7667 -2.8333 Dalias (Almeria) 6.4 27/10/1806 37.2333 -3.7333 Pinos Puente (Granada) 5.3 20/03/1829 38.0833 -0.6833 Torrevieja (Alicante) 6.6 25/12/1884 37.0000 -3.9833 Arenas del Rey (Granada) 6.5 29/03/1954 37.0000 -3.6000 Dúrcal (Granada) 7.0 19/04/1956 37.1833 -3.6833 Albolote (Granada) 5.0 28/02/1969 35.9833 -10.8167 SW. Cabo San Vicente 7.8 11/05/2011 37.7000 -1.6833 Lorca (Murcia) 5.1 Source: Instituto Geográfico Nacional. "Terremotos más importantes (En España)" (in Spanish). Retrieved 2018-03-12. 5.2 Scenario of Earthquake in Andalucia According to the report of Instituto Geológico y Minero de España, (survey year: 1987 – 2001) Andalucia is the most vulnerable area of Spain (Análisis del impacto de los riesgos geológicos en España, 2004) due to earthquake in terms of economic loss. Andalucia has higher seismic activity amongst different regions of Spain (Benito et al, 2008), although in a global context the seismic activity of Andalucia lies between low to moderate. Most of the earthquakes occurred in Andalucia at shallow depth (h < 40 km), a significant number with at intermediate depth (40 < h < 150 km) and only a very few earthquakes occurred in deep depth (around 630 km) (Buforn et al, 1995). This region is closer to where the Eurasian and African plates are met. So earthquakes are more frequent than the other parts of Spain. Seismic epicenters are concentrated in eastern Andalusia of Granada, Málaga and Almeria provinces (Agencia Andaluza de la Energía) (Fig. 3[B] and 4). The intensity of earthquake is moderate, occasionally reach or surpass 6 in Richter scale. Highest PGA value has been found in Granada province of Spain. Cullar Vega, La Malaha and Las Gabias municipalities have highest PGA value (0.24) in Granada. The probability of recurrence of earthquake is more in east of Málaga province, all Granada and west of Almeria. 118 !"#$%&'()*$"#('+)$,)-&$.#'%/0)1$+"2&)345)6"27&#)38339:3;;5)<'#=/ >43? !@AA$=B'CB$()$,)!"#$%&'()-&$.#'%/&#A) !"#$%&'()*$"#('+)$,)-&$.#'%/0 : DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I !&( Figure 3. [A] Geo-tectonic environment of the study area and [B] Location of major earthquakes. Source: Hürtgen et al. 2013, Benito et al. 2008 Figure 4. Spatial distribution of large earthquakes in Andalucia from 880 to 1994. Source: Plan de actuación local ante riesgo sísmico en Benalmádena, 2012 5.3 Scenario of Earthquake in Málaga Province 5.3.1 Historical earthquakes Málaga province has suffered earthquakes greater than or equal to maximum intensity VIII (epicenter within Málaga province) in 1494 and 1680. Earthquakes were generally occurred in the southern region of the Málaga province. The earthquake of January 26, 1494, caused significant damages in whole Málaga province. It was felt from Granada and Sevilla, around 200 119 Gayen S. – I.V. Villalta / European Journal of Geography 10 1 117–133 (2019) !"#$%&'()*$"#('+)$,)-&$.#'%/0):)DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I" !('$ assessment of susceptibility of this extreme event is more applicable. Everyone who takes appropriate earthquake safety precautions can minimize damages and injuries. Most of the buildings in Málaga province were constructed before 20th century. It is essential to improve the design of the old buildings and strengthening the old houses to incorporate the latest advances in seismic and structural engineering. Improvement of the building elements that resist the forces generated by seismic waves, reconstruction of weak structures in highly seismic zones, land use control and public awareness will significantly reduce the impact of earthquake. Damage of earthquake is inversely proportional to the resistant power of the infrastructure. Earthquake resistance structures are able to protect the buildings. It decreases the probability of damage and so the loss of life could be minimized. The thematic maps on the various parameters of earthquake (i.e. historic earthquake maps, major damaged maps, vulnerable maps, map on safe zone, map regarding fast aids facilities, road network of shortest path, rescue and relief maps etc.) should be displayed in the public places. Kiosk regarding the micro-zoning of this disaster must be opened in the crowded areas. These types of information will be more useful for the tourists and outsiders. As Málaga is fully occupied by urban settlement, the microzoning of earthquake hazards should be prepared at utmost priority. This work shows the regions of past earthquakes. This data representation helps to recognize date, epicenter coordinates, magnitude or intensity, depth, fault and the direction of slips of past earthquakes. The peak ground acceleration map will be helpful to identify the earthquake vulnerable zones and the map can be used for land-use planning including building infrastructure, earthquake mitigation, emergency response, and minimizing the environmental damages. To reduce the risk of earthquake several ‘Seismic Alert Network Stations’ have been constructed in Spain. In Málaga province, this station is situated in Mijas and is responsible for sensing the ground motion, determining time and collecting data. The objectives of seismic alert network stations is to detect the initiation of an earthquake at earliest possible time by estimating the ground motion before shaking starts which can reduce damage and casualties during an earthquake. Ground rupturing is also one of the important footprints of earthquake. Detail study on ground rupturing and its associated phenomenon is the scope of the future work. The Instituto Geográfico Nacional, through the National Seismic Network (Red Sísmica Nacional), is also trying to frame a project that will be able to establish an alert procedure for tsunamis in the Atlantic and Mediterranean coastal regions. ACKNOWLEDGEMENTS We are thankful to Instituto Geográfico Nacional, Spain for providing necessary data. REFERENCES Actualización de mapas de peligrosidad sísmica de España. (2012). Centro Nacional de Información Geográfica (CNIG). http://publicacionesoficiales.boe.es (Accessed 2019-08- 13). Abdalla J. A., and A. Al-Homoud. (2004). Earthquakze hazard zonation of Eastern Arabia. 13th World Conference on Earthquake Engineering. Vancouver B.C., Canada. Paper No. 1008. Agencia Andaluza de la Energía, Concejería de Economía, Innovación y Ciencia, Anexo 6. La sismicidad en Andalucía como indicador geotérmico. 126 !"#$%&'()*$"#('+)$,)-&$.#'%/0)1$+"2&)345)6"27&#)38339:3;;5)<'#=/ >43? !@AA$=B'CB$()$,)!"#$%&'()-&$.#'%/&#A) !"#$%&'()*$"#('+)$,)-&$.#'%/0 : DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I !(! https://www.agenciaandaluzadelaenergia.es/sites/default/files/documentos/1300195655433_ anexo_6_-_sismicidad_en_andalucxa.pdf (Accessed 2019-08-13). Alvarez-Marrón, J. (1999). Pliocene to Holocene structure of the Eastern Alboran Sea (Western Mediterranean). Proceedings of the Ocean Drilling Program, Scientific Results 161: 345– 355. Análisis del impacto de los riesgos geológicos en España. Evaluación de pérdidas por terremotos e inundaciones en el periodo 1987 – 2001 y estimación para el periodo 2004-2033. (2004) Instituto Geológico y Minero de España. http://info.igme.es/SidPDF/112000/337/Tomo%203-1.%20Anexos/112337_0003.pdf (Accessed 2019-08-13). Benito B., M. Navarro, J. M. Gaspar-Escribano, F. Vidal, A. Góngora, M. J. García Rodríguez, and J. M. Martínez Solares. (2008). Seismic hazard in Andalusia region (Southern Spain). The 14th World Conference on Earthquake Engineering Beijing, China. Buforn, E., C. Sanz de Galdeano, and A. Udías. (1995). Seismotectonics of the Ibero- Maghrebian region. Tectonophysics 248: 247-261. Casado C. L., C. Sanz de Galdeano, S. M. Palacios, and J. H. Romero. (2001). The structure of the Alboran Sea: an interpretation from seismological and geological data. Tectonophysics 338: 79–95. De Pascale F., M. Bernardo, F. Muto, S. D’amico, R. Zumbo, P. Galea and M. Agius (2015). Neogeography and seismic risk perception. A comparison between two case studies: Calabria, Southern Italy and Malta. European Journal of Geography 6 (1): 64-82. Di Bucci, D., P. Burrato, P. Vannoli, and G. Valensise. (2010). Tectonic evidence for the ongoing Africa!Eurasia convergence in central Mediterranean foreland areas: A journey among long!lived shear zones, large earthquakes, and elusive fault motions. Journal of Geophysical Research. 115: B12404 (1-17). DOI:10.1029/2009JB006480. Fernández-Ibáñez, F., J. I. Soto, M. D. Zoback, and J. Morales. (2007). Present-day stress field in the Gibraltar Arc (western Mediterranean), Journal of Geophysical Research. 112: B08404 (1-25). DOI:10.1029/2006JB004683. Gaspar-Escribano, J. M., and B. Benito. (2008). Overview of recent seismic risk analysis in Spain. The 14th World Conference on Earthquake Engineering Bejing, China. Gaspar-Escribano, J. M., B. Benito, M. J. García Rodríguez, A. Rivas-Medina, L. Cabañas, J. J. Martínez-Díaz, and M. J. Crespo. (2010). Towards a new seismic hazard assessment in Spain. 14th European Conference on Earthquake Engineering Ohrid, Macedonia. Goded, T., E. Buforn, and D. Muñoz, (2008). The 1494 and 1680 Málaga (Southern Spain) earthquakes. Seismological Research Letters 79 (5): 707-715. Grácia E., R. Pallás, J. I. Soto, M. Comas, X. Moreno, E. Masana, P. Santanach, S. Diez, M. García, and J. Dañobeitia (2006). Active faulting offshore SE Spain (Alboran Sea): Implications for earthquake hazard assessment in the Southern Iberian Margin. Earth and Planetary Science Letters 241 (3-4): 734–749. Hürtgen J., A. Rudersdorf, C. Grützner and K. Reicherter (2013). Morphotectonics of the Padul- Nigüelas Fault Zone, southern Spain. Annals of Geophysics 56 (6). S0679. DOI:10.4401/ag- 6208. Insua-Arévalo J. M., J.J. Martínez-Díaz, J. García-Mayordomo and F. Martín-González (2012). Active tectonics in the Málaga Basin: evidences from morphotectonic markers (Western 127 Gayen S. – I.V. Villalta / European Journal of Geography 10 1 117–133 (2019) !"#$%&'()*$"#('+)$,)-&$.#'%/0):)DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I" !(&$ Betic Cordillera, Spain). Journal of Iberian Geology 38 (1): 175-190. http://dx.doi.org/10.5209/rev_JIGE.2012.v38.n1.39212. Irwansyah E., E. Winarko, Z. E. Rasjid, and R. D. Bekti. (2013). Earthquake hazard zonation using peak ground acceleration (PGA) approach. Journal of Physics: Conference Series 423: 012067. Instituto Geográfico Nacional; (2019); Seismic information; https://www.ign.es/web/ign/portal/sis-area-sismicidad (Accessed 2019-08-13). Intituto de Estadística y Cartografía de Andalucía; (2019); Sistema de Información Multiterritorial de Andalucía (SIMA); http://www.juntadeandalucia.es/institutodeestadisticaycartografia/iea/consultasActividad.jsp ?CodOper=104&sub=38120 (Accessed 2019-08-13). Makropoulos, K.C., and P. W. Burton. (1985). Seismic hazard in Greece. II. Ground acceleration. Tectonophysics 117: 259-294. Millán T. G. (2006). Reevaluación de daños de los terremotos de Málaga de 1494 y 1680. Trabajo de Investigación. Departamento De Física De La Tierra, Astronomía Y Astrofísica I (Geofísica Y Meteorología). Universidad Complutense de Madrid. https://webs.ucm.es/info/Geofis/g-sismolo/Goded_tesina_2006.pdf (Accessed 2019-08-13). Monsó C. G., and E. Miranda. (2018). Significance of directivity effects during the 2011 Lorca earthquake in Spain. Bulletin of Earthquake Engineering 16 (5): DOI: 10.1007/s10518-017- 0301-9. Parsons, T. (2008). Persistent earthquake clusters and gaps from slip on irregular faults. Nature Geoscience 1: 59–63. Peláez J. A., A. M. Rosa, M. S. Gómez, and I. M. García. (2011). Efectos arqueosismológicos del terremoto de Málaga de 09/10/1680 en el sur de la provincia de Córdoba. 4º CNIS (Granada). https://www.academia.edu/6099107/Efectos_arqueosismol%C3%B3gicos_del_terremoto_d e_M%C3%A1laga_de_09_10_1680_en_el_sur_de_la_provincia_de_C%C3%B3rdoba (Accessed 2019-08-13). Plan de actuación local ante riesgo sísmico en Benalmádena. (2012). Área de Gobierno de Seguridad y Emergencias Excmo. Ayuntamiento de Benalmádena (Málaga). http://www.benalmadena.es/emergencias/docs/plan-de-actuacion-local-ante-riesgo- sismico.pdf (Accessed 2019-08-13). Sánchez, F. V. (2011). El Terremoto de Alhama de Granada de 1884 y su impacto. Anuari Verdaguer 19. https://www.raco.cat/index.php/AnuariVerdaguer/article/view/262467/349631(Accessed 2019-08-13). Sanz de Galdeano C., F. J. García-Tortosa, J. A. Peláez, P. Alfaro, J. M. Azañón, J. Galindo- Zaldívar, C. López Casado, A. C. López Garrido, J. Rodríguez-Fernández, and P. Ruano. (2012). Main active faults in the Granada and Guadix-Baza Basins (Betic Cordillera). Journal of Iberian Geology 38 (1): 209-223. Sanchez, F. V. (1986). Sismotectónica de la región Béticas-Mar de Alboran. Ph.D. Thesis. Universidad de Granada. http://digibug.ugr.es/handle/10481/5951 (Accessed 2019-08-13). Sari A. M., and A. Fakhrurrozi. (2017). Earthquake hazard analysis methods: A review. IOP Conf. Series: Earth and Environmental Science 118: 012044. DOI :10.1088/1755- 1315/118/1/012044. 128 !"#$%&'()*$"#('+)$,)-&$.#'%/0)1$+"2&)345)6"27&#)38339:3;;5)<'#=/ >43? !@AA$=B'CB$()$,)!"#$%&'()-&$.#'%/&#A) !"#$%&'()*$"#('+)$,)-&$.#'%/0 : DEE6)39?>:3;F3)G)@++)#B./CA)#&A&#H&I !(( Solares, J. M. M. (2016). Earthquakes in Spain. Recent activity in the Alborán Sea. Consorseguros Revista Digital 4. http://www.consorsegurosdigital.com/en/numero-04/indepth/earthquakes-in-Spain-recent-activity-in-the-alboran-sea-en (Accessed 2019-08-13). Vipin K. S., P. Anbazhagan, and T. G. Sitharam. (2009). Estimation of peak ground acceleration and spectral acceleration for South India with local site effects: probabilistic approach. Natural Hazards and Earth System Sciences 9: 865–878. 129 2. Gayen, S.; Villalta, IV.; Haque, SM. (2020): Comparative social vulnerability assessment in Purba Medinipur district, West Bengal, India. European Journal of Geography, 11(1): 093-108. https://doi.org/10.48088/ejg.s.gay.11.1.93.107 Impact factor: 0.285 (SCOPUS SJR) Social Sciences: Geography, Planning and Development: 396/818 (Q3) 130 Volume 11, Issue 1, pp. 093 - 107 Article Info: Received: 08/05/2020; Received in revised form: 18/09/2020; Accepted: 19/10/2020 Corresponding Author: *sumga[email protected] https://doi.org/10.48088/ejg.s.gay.11.1.93.107 Comparative social vulnerability assessment in Purba Medinipur district, West Bengal, India Sumita GAYEN1*, Ismael Vallejo VILLALTA1, Sk Mafizul HAQUE2 1 University of Seville, Spain 2 University of Calcutta, India Keywords: Social vulnerability, deductive approach, indicators, comparative assessment Abstract Social vulnerability assessment is an important element for disaster risk reduction system. Social vulnerability depends on physical, social, economic, environmental factors, etc., which determine the susceptibility of community against the impact of hazar ds. Over the past two decades, social vulnerability has been measured by using different methodologies. The present work portrays a comparative assessment of social vulnerability methodologies in the Purba Medinipur district located in the eastern part of India. Different standardization techniques of data transformation like, Z score, maximum value, and minmax rescaling have been employed to compare and examine social vulnerability in the study area. The principal aim of this work is to analyse the result s between the three methodologies adopted, and identify the vulnerable blocks in the district. The result shows that relationships between three methodologies are moderately strong to very strong and Sutahata block is identified as the most vulnerable block in the Purba Medinipur district. The publication of the European Journal of Geography (EJG) is based on the European Association of Geographers’ goal to make European Geography a worldwide reference and standard. Thus, the scope of the EJG is to publish original and innovative papers that will substantially improve, in a theoretical, conceptual or empirical way the quality of research, learning, teaching and applying geography, as well as in promoting the significance of geography as a discipline. Submissions are encouraged to have a 131 European Journal of Geography vol.11(1), pp.093-107, 2020 © Association of European Geographers European Journal of Geography - ISSN 1792-1341 © All rights reserved 94 © Association of European Geographers European dimension. The European Journal of Geography is a peer-reviewed open access journal and is published quarterly. 1. INTRODUCTION In 1970´s the concept of social vulnerability within the disaster management context was introduced (Flanagan, 2011). Some populations are prone to suffer more damages and loss from different hazards. Social vulnerability examines social distribution of risk and reasons behind high level of risk associated with some communities to any hazards. It is the adaptation technique between hazard and people. Social vulnerability can be defined as the state of a community before a hazard triggers them. Under the mitigation strategy of adaptation approach, social vulnerability assessment is one of the vital attempts for the building of regional preparedness. Social vulnerability is associated to the level of wellbeing of individuals, communities and society (ISDR, 2004). Normally two approaches are being used to select and calculate the social vulnerability. First one is deductive approach which is based on theoretical understanding. Second one is inductive (empirical) approach which is based on statistical relationship of a large number of indicators set. Deductive approach is a ‘top-down’ approach and inductive approach is a ‘bottom-up’ approach. Cutter et al. (2003), Rygel et al. (2006), Boruff et al. (2005), Boruff et al. (2007), Holand et al. (2011), Chen et al. (2013), Mavhura et al. (2017), Heß (2017) used inductive approach in their studies. Whereas, Cutter et al. (2000), Bernard (2007), Cutter et al. (2010), Sajjad et al. (2014), de Mello Rezende (2016) chose deductive approach for social vulnerability measurement. Yoon (2012) had compared between deductive approach and inductive approach and developed social vulnerability index for in the Gulf of Mexico and Atlantic coastal areas. This article is based on the three deductive techniques such as z score transformation, maximum value transformation and min–max rescaling transformation, to measure comparative social vulnerability assessment in 25 blocks of Purba Medinipur district in West Bengal state of India The assessment of social vulnerability is a key issue of indicator based measurement. Indicators are necessary to determine the social vulnerability. This assessment consists of 12 socio-economic indicators for construction of a social vulnerability index (SVI). Social vulnerability index is a planning tool that helps to identify communities; those may need support in preparation for hazards or recovering after hazards. Comparative analysis is a combination of different methodologies for analyzing data. It involves the analysis and synthesis of the similarities, differences and relationship between several methodologies. It is a new novel approach of vulnerability assessment and it mainly focuses on the relationships between two or more variables. The main objective of this article is to identify the vulnerable blocks in Purba Medinipur district by employing three standardization methods and performing a comparative analysis of the data obtained. 2. SOCIAL VULNERABILITY 132 Gayen, S. et al./ European Journal of Geography vol.11(1), pp.093–107, 2020 European Journal of Geography - ISSN 1792-1341 © All rights reserved 95 Social vulnerability indicates to a population’s exposure to potential hazards. Some physical, demographic, social, economic conditions of people expose more or less to a hazard. These factors determine damage, loss and injury of people in any hazard. Social vulnerability is controlled by some indicators. There exist two types of relationship between indicator and social vulnerability. In first one, vulnerability increases with the increasing value of indicator. This is called positive relationship. Another one is negative relationship. Here vulnerability decreases with increasing the value of indicator. Generally higher value of social vulnerability indicates high number of elder population, single living person, child population, female, poverty, unemployment, illiteracy of populations etc. Some of the definitions of social vulnerability are summarized below in Table 1. Table 1. Definitions of social vulnerability Working definition(s) Exemplar reference source Social vulnerability refers to the exposure of a group or individual to stress as a result of social and environmental changes where stress indicates to unexpected changes and disruption to livelihoods. Adger, 1999 Social vulnerability is a product of social inequalities. It is defined as a susceptibility of social groups to the impacts of hazards and their resiliency or ability to adequately recover from those. Cutter et al, 2006 Social vulnerability shows the social construction to prevent a hazard. It is also partially the product of s ocial inequalities. Veen et al, 2009 The aims of social vulnerability are to identify and understand which groups of people are more sensitive and susceptible to the impacts of natural disasters and why. Tapsell et al, 2010 The concept of social vulnerability is multidimensional and not directly observable. Tate, 2012 3. METHODOLOGY 3.1 Study Area Purba Medinipur district is situated between 21°36¢35²N and 22°57¢10²N latitudes and 86°33¢50²E and 88°12¢40²E longitudes. Midnapore district was bifurcated into Paschim Midnapore and Purba Midnapore on 1st January 2002. The location map of Purba Medinipur district is shown in Figure 1. The district has a 65.5 km long coastline along the south and south eastern part. Purba Medinipur district has divided into twenty-five blocks and five municipalities. These municipalities are; Tamluk (Tamluk Block), Panskura (Panskura Block), Haldia (Haldia Block), Egra (Egra-I Block) and Contai (Contai-I Block). The district headquater is Tamluk. 133 European Journal of Geography vol.11(1), pp.093-107, 2020 © Association of European Geographers European Journal of Geography - ISSN 1792-1341 © All rights reserved 96 Purba Medinipur district is a part of lower Ganga plain. The slope of this region is from west to east. Haldi, Rasulpur, Keleghai, Kanshabati, Rupnarayan are main rivers of the district. The climate of the district is tropical monsoon type with hot and humid summer and dry winter. According to 2011 census data (CI, 2011), population of Purba Medinipur district is 5,095,875 and it covers 5.58% of population of West Bengal. The district has a population density of 1081 inhabitants per square km. Purba Medinipur district ranks 8th in terms of population and population density of the state. The northan parts of the district are more densely populated than the other parts of the district. The population growth rate is 15.36%. The sex ratio of Purba Medinipur district is 938 females per 1000 males. Out of the total population 11.6% of population lives in urban area and 88.4% population lives in rural area. Whereas India’s urban population is 31.2% (Kumar, 2015). The literacy rate of Purba Medinipur district is 87.02%. The male literacy rate is 92.32% and female literacy rate is 81.37%. The district has recorded highest literacy rate all over the state. The state literacy rate is 76.26%. Only 37.49 % of population of the district is workers, whereas 62.51% of population is non-workers. The dependent population ratio is high in the district. Agriculture is the main activity in the Purba Medinipur district. The crops grown in Purba Medinipur district are Paddy, Wheat, Mustard, Jute, Potato, Chillies, Ginger etc. The main crop is Paddy which occupied more than 50% of total production. The coastal blocks are famous for fish cultivation. Many people are engaged with fishing from the sea. Tourism also plays an important role in the economy of the coastal blocks 134 Gayen, S. et al./ European Journal of Geography vol.11(1), pp.093–107, 2020 European Journal of Geography - ISSN 1792-1341 © All rights reserved 97 Figure 1. Location Map of Purba Medinipur district 3.2 Data Base All the data have been collected from census data 2011 (CI, 2011) and district human development report 2011 (DHDR, 2011). The maps were prepared by using ArcGIS. 3.3 Indicator Selection To determine the methodologies of this work, the main prerequisite was selection of indicators. 12 different indicators have been selected. Table 2 provides classification and the name of the indicators and some basic information about those. Table 2. Indicators and their description Sl. no. Indicators Name Description Relationship with vulnerability Literature review 1 % of female PFEP Women are more Positive Müller et 135