Content area
Purpose
Risk assessment is imperative for disaster risk reduction. The risk is rooted to various physical, social, economic, demographic and environmental factors that determine the probable magnitude of loss during an extreme event. By way of bringing a conceptual model into practice, this paper aims to examine the flood risk of the Srinagar city.
Design/methodology/approachThe “risk triangle” model has been adopted in the present investigation evaluating parameters, reflective of hazard (intensity), exposure (spatial) and vulnerability (sensitivity) using Landsat-8 operational land imager scene (10 September 2014), global positioning system, Cartosat-1 digital elevation model and socioeconomic and demographic data (Census of India, 2011). The authors characterise flood hazard intensity on the basis of variability in water depth during a recent event (September 2014 Kashmir flood); spatial exposure as a function of terrain elevation; and socioeconomic structure and demographic composition of each municipal ward of the city as a determinant factor of the vulnerability. Statistical evaluation and geographic information system-based systematic integration of all the multi-resolution data layers helped to develop composite flood risk score of each ward of the city.
FindingsPrincipal deliverable of this study is flood risk map of the Srinagar city. The results reveal that approximately 46 per cent of the city comprising 33 municipal wards is at high risk, while rest of the area, i.e. 17 and 37 per cent, exhibit moderate and low levels of risk, constituting 23 and 12 municipal wards, respectively. It is very likely that the municipal wards expressing high risk may witness comparatively more damage (impact) during any future flood event. Thus, there is a need of planned interventions (structural and non-structural) to minimise the emergent risk.
Originality/valueVery rare attempts have been made to bring theoretical models of disaster research in practice; this is mainly because of the complexities associated with the data (selection, availability and subjectivity), methodology (integration, quantification) and resolution (spatial scales). In this direction, this work is expected to have considerable impact, as it provides a clear foundation to overcome such issues for the studies aiming at disaster risk assessment. Furthermore, using varied primary and secondary data, this paper demonstrates the relative (municipal wards) flood risk status of the Srinagar city, which is one of the key aspects for flood hazard mitigation.
1. Introduction
Flooding is the most common natural hazard distressing people worldwide. It has been estimated that about a third of the total landscape of the earth is flood-prone, consequently affecting approximately 82 per cent of the world’s population (Dilley et al., 2005). World Meteorological Organization (World Meteorological Organization (WHO), 2014) asserted that between 1970 and 2012, about 89 per cent of the reported global disasters were due to flooding and storms, resulting in deaths of millions and huge economic losses. Globally, economic losses from flooding exceeded $19bn in 2012 (Visser et al., 2012; IPCC, 2012; Ward et al., 2013). The twentieth century witnessed an unprecedented rate of flood occurrences within the most drainage basins of the world (Perry, 2000). The yearly evaluation shows that in the past 50 years (1945 to 2005), flood incidence has increased by almost 10-fold (Scheuren et al., 2008). The statistics inarguably suggests that flooding is a global environmental issue that requires consistent monitoring and mitigation.
Structural measures may significant reduce the losses associated with floods; however, they cannot completely eliminate the flood risk (Bailey et al., 1989; Alexander and Wilson, 1995; Watson and Biedenharn, 2000; UNISDR, 2011; FEMA, 2012). There are other aspects such as hazard intensity, spatial exposure, social structure and demographic composition that need to be taken into account in the process of flood risk reduction. Therefore, an integrated assessment method, including maximum possible parameters is needed to address residual risk. Considering the multi-faceted nature of risk, UNISDR defines risk assessment as “a qualitative or quantitative approach to determine the nature and extent of disaster risk by analysing potential hazards and evaluating existing conditions of exposure and vulnerability that together could harm people, property, services, livelihoods and the environment on which they depend”. Thus, to minimise both human suffering and the economic loss, the risk is be assessed in a comprehensive manner considering hazard, spatial exposure and vulnerability (Turner et al., 2003; Adger et al., 2004; Flanagan et al., 2011). Moreover, the use of a geographic scale sufficient to depict spatial differences in risk is also important (Flanagan et al., 2011). The scale implications need more attention and documentation in risk studies (Fekete et al., 2009). Portraying comparative risk at varied spatial scales on the basis of administrative, planning and political boundaries is often a generalised one. However, usage of the high spatial resolution data or evaluation at smaller planning units may minimise the effect of the generalisation considerably (Jones and Andrey, 2007). In the present instigation, we use municipal ward as a spatial unit to present relative flood risk of the Srinagar city.
1.1 Srinagar city
Srinagar city is located in Kashmir basin (Figure 1) of Jammu and Kashmir State (India). The city has two main physical divisions (Right River Division and Left River Division) with a total of 68 municipal wards. The right river division is spread over two administrative zones – east zone with eight administrative wards comprising 17 electoral wards and north zone with nine administrative wards having 17 electoral wards. The left river division is also spread over two administrative zones – west zone with eight administrative wards comprising 15 electoral wards and south zone with nine administrative wards comprising 19 electoral wards [Figure 1(b)] (SMC 2011, www.smcsite.org/zonedetails.php).
Srinagar is not only the largest urban centre both in terms of the population and areal extent but also the rapidly growing city amongst all Himalayan urban centres (Bhat, 2008), which is evident from the fact that the area of the city increased from 12 km2 in 1901 to 300 km2 in 2011 (SMC, 2011). The city has been also listed as one of the hundred (92nd) fastest growing cities of the world on basis of average annual growth rate from 2006 to 2020 (www.citymayors.com/statistics/urban_growth1.html).
Owing to its geomorphic configuration (elongated trough encompassed by mountains) and location (seismically active), the Kashmir basin is exposed to various natural hazards such as floods, earthquakes, landslides and snow avalanches; however, flooding is the most recurrent phenomenon in the basin, especially in the Srinagar city (Ahmad et al., 2009; Ahmad et al., 2013; Ahmad et al., 2014; Alam et al., 2014; Ahmad et al., 2015; Alam et al., 2015a, 2015b; Meraj et al., 2015; Ahmad et al., 2016; Alam et al., 2017). There are records of flood occurrence and associated losses in the city that date back to 3000 BC (Hassan, 1896; Raza et al., 1978; Koul, 1991; Bilham and Bali, 2014). Pertinently, simulations based on instrumental record of annual peak flows (1956-2014) from Ram Munshibagh gauging station of River Jhelum within the city suggests that the frequency of floods may increase in future (Figure 2). Thus, it is imperative to have comprehensive and reliable estimates of flood risk in the Srinagar city.
2. Data and methodology
Disaster risk is typically described as a function of hazard, exposure and vulnerability (Blaikie et al., 1994; Crichton, 1999; Kron 2002; Peduzzi et al., 2009; Tomlinson et al., 2011; Cavan and Kingston, 2012; Espada et al., 2015; Murnane et al., 2016). Here, we use the widely recognised approach, known as “the risk triangle” model (Crichton, 1999; Kron 2002) to assess the flood risk of the Srinagar city (Figure 3). According to the structure of the model, we consider three variables, i.e.
hazard intensity, as a function of water depth (event specific record of inundation);
terrain elevation as a representative variable for determining the spatial exposure to floods; and
vulnerability, represented by a set of physical, social and economic indicators, i.e. housing density, crowding index, population density, child population (0-6 year), elderly population (60 years and above), women population, working population, slum population, literate population/level of education, distribution of critical service centres (health, emergency, education and finance), obtained from Census of India (2011) reports and field surveys.
In view of the fact that the selected variables and their respective attributes have different connotations in disaster research, such as comparative higher value of one indicator (e.g. elevation) represents negative correlation with the flood risk; however, higher value of another indicator (e.g. population density) represents positive correlation with the probable damage due to floods (Table I). Therefore, ranking and inverse ranking of the data according to the nature of variable (±correlation with the risk) was done. Moreover, to develop a common evaluation scale of all indictors corresponding to the variables, the whole range of absolute statistics corresponding to each variable was reclassified and reduced to a new scale ranging from 0 to 5. For projecting a composite picture, an additive model was used, followed by z-score calculation (Cutter et al., 2003) of the scale values of each variable to determine relative flood risk score of the each ward (Table II). Here, we avoid subjective ranking of the variables to minimise the error to the risk assessment, often introduced due to biasness towards a particular variable/s. In the present assessment model, every variable is contributing equally to the overall risk status of a particular spatial unit (municipal ward).
3. Results and discussion
3.1 Hazard
Damage due to floods is always a consequence of both water depth and flow velocity (Kreibich et al., 2009; Zhang et al., 2014; Murnane et al., 2016). Spatial variation in intensity of the flood hazard is determined on the basis inundation i.e. more the flood water depth, more is the possibility of damage and vice versa. During September 2014, extreme flooding was experienced in different parts of south Asia. The transboundary flood event resulted in widespread destruction and colossal loss of life and property in the Kashmir basin of Jammu and Kashmir state. With estimated discharge of approximately 1,15,218 upstream at Sangam and 72,585 cusecs downstream at Ram Munshibagh gauging stations in Srinagar, the discharge was declared as the highest ever recorded on the trunk river (Jhelum) of Kashmir basin (Irrigation and Flood Control Department, 2014, www.jkirfc.com/). The flood resulted in death of 282 people and damage to 2.53 lakh houses across the state of Jammu and Kashmir (www.greaterkashmir.com/…/-282-people-died-2…jk-floods/181241.html). During this flood event, about 60 per cent of the Srinagar city’s area submerged under water witnessing the water depth up to 16 feet (e.g. in the localities of Rajbagh, Jawahar Nagar, Gogji Bagh) and extended inundation period of approximately 25 days in the residential areas including the Central Business District (CBD) of the city [Figure 4 (a, b, c)]. More than 0.6 million people were stranded in the submerged housing colonies of the city for more than a week without drinking water, food and other essentials.
We used water depth records of this event as an ideal input data to determine the flood hazard levels of each ward, assuming velocity and debris as constant across the flooded area. The spatial extent of the flooding was determined from the Landsat-8 (operational land imager) satellite image (e.g., Wang et al., 2002; Merwade et al., 2008; Haq et al., 2012; Zhang et al., 2014) and data on the highest flood level record of various locations was obtained through global positioning system aided post-flood survey, followed by interpolation (nearest neighbour) of the point data to develop continuous inundation layer (flood water depth map) for the whole city [Figure 4(c)]. Five classes with varying water depth were identified; the maximum depth (3-5 metres) of the flood water was measured in the wards adjacent to main river, especially on the left bank side, covering most wards of central, north-western and south-eastern parts of the city. As expected, post-flood field survey revealed that maximum damage was experienced in the areas, where flood water depth was higher.
3.2 Exposure
Exposure is spatial in context and is a direct function of geographic location. The primary factors determining exposure of a spatial unit to flood hazard is the topography (elevation). Environmental hazards to which a region is exposed, such as those presented by its surrounding topography, can have a significant impact on the future integrity of its built structures and so on its socioeconomic sustainability (Fedeski and Gwilliam, 2007). Proximity or exposure to a hazard agent, the nature of the hazard itself, is critically important in assessing and addressing population vulnerabilities to hazard impacts (Zahran et al., 2008). The human settlement and other assets located in the low-lying terrain are more exposed to inundation than the areas of elevated terrain (IPCC, 2012).
In most cases, terrain elevation may be used as proxy variable to determine flood water depth, i.e. low lying area may experience the highest flood water depth; however, some structural controls such as embankments may provide controlled passage to flood water, thus avoiding inundation of low lying zone along some river reach. It is for that reason we consider terrain elevation and flood water depth as two independent variables. Digital elevation model (DEM) is widely used as an input data set to map zones exposed to floods (Leenaers and Okx, 1989; Zhang et al., 2014). The exposure of each municipal ward of the city was determined on the basis of elevation; obtained from Cartosat-1 DEM. Five terrain elevation classes were identified from the data ranging approximately 1,478 to 1,672 metres above mean sea level (amsl) [Figure 4(d)]. In general, the gradient of the terrain in the city decreases in SE-NW direction; hence, the municipal wards from NW of the city are the most low lying, followed by some scattered depressions from west and south of the city.
3.3 Vulnerability
An evaluation of the risk to an element from a hazardous event requires a consideration of the element’s vulnerability, which expresses its propensity to suffer damage (Douglas, 2007). The vulnerability analysis has emerged as a central concept for understanding what it is about the condition of people that enables a hazard to become a disaster (Tapsell et al., 2010). The process of undertaking the assessment can contribute to better understanding of community and environmental needs with respect to capacity-building and/or the identification of adaptation actions for vulnerability reduction (Adger et al., 2004).
The factors that contribute to the vulnerability cannot be generalised (Müller et al., 2011) and the indicators selected for various studies depend on the purpose of the study, the research discipline being explored and the final application (Dwyer et al., 2004). It is important nonetheless to provide consistent frameworks for measuring vulnerability that provide complementary qualitative and quantitative insights into outcomes and perceptions of vulnerability (Adger, 2006). Comprehensive vulnerability analysis considers the totality of the system; however, that is unrealistic because availability of data and other real world constraints invariably necessitate a “reduced” vulnerability assessment (Turner et al., 2003).
Here, we use smallest planning unit (municipal ward) as a geographic scale to evaluate the vulnerability. The comparative vulnerability of spatial units was determined based on specific indicators, contributing to the overall vulnerability status of the city. Following are the selected indicators used to develop a composite index of the overall vulnerability:
3.3.1 Housing density.
Housing density refers to the number of houses per unit of land. Household density is being treated both as an indicator of low socioeconomic status and stressful situation associated with high morbidity/mortality risks (Melki et al., 2004). High housing density also represents congested environment of the human habitation. Owing to shortage of vacant land, it restricts free movement and may be a cause of human causalities and economic loss during an emergency.
3.3.2 Crowding index.
Crowding index is generally quantitative measure based on a calculation i.e. number of people in a dwelling and the dwelling size, or a proxy for size such as the number of rooms or bedrooms; the quantitative measures range from simple counts of people and rooms, to more sophisticated models that also take into account household composition and demographic information (Goodyear et al., 2011). Several decades of research have correlated a high household crowding index, denoted by the number of co-residents per room, with socioeconomically deprived urban communities (Melki et al., 2004).
3.3.3 Population density.
Population density has long been viewed as one of the main contributing factor to the vulnerability status of communities, both in developed and underdeveloped nations. The high death tolls in the Asian continent are due to high population density (UNEP, 2002). High housing density, crowding index and population density suggest the higher number of socially deprived population with low living standards/quality is vulnerable to hazards within a restrained zone.
3.3.4 Female population.
Women can have mobility constraints during an extreme event especially during pregnancy and inherent limited physical flexibility than men. Moreover, women can have a more difficult time during recovery than men, often due to sector-specific employment, lower wages, and family care responsibilities (Cutter et al., 2003).
3.3.5 Working population.
Working population especially in secondary or tertiary sectors may be considered an independent and as resilient section of the population due to possible physical (healthy) and economic (salaried) abilities. In general, high percentage of working population reflects community’s reduced vulnerability level. The proportion of workers to total population depicts the rate of economic productivity and also the rate of unemployment or dependency ratio.
3.3.6 Level of education.
Education plays an important role in building resilience of communities. Level of education has many implications in vulnerability science and influences other constituent variables of the vulnerability. For example, lower education constrains the ability to understand warning information and access to resources (Cutter et al., 2003).
3.3.7 Income.
Poverty is an important aspect of vulnerability because of its direct association with access to resources (Adger and Kelly, 1999). Levels of vulnerability are highly dependent upon the economic status of individuals, communities and nations (Tapsell et al., 2010). Income is one of the primary variables affecting the overall vulnerability status of a household, community or region.
3.3.8 Age extremes (child and elderly population).
Children and elderly are most vulnerable group of the society. Extremes of the age spectrum affect the movement out of harm’s way; children/elderly may have mobility constraints or mobility concerns increasing the burden of care and lack of resilience (Cutter et al., 2003).
3.3.9 Critical service centres (medical, emergency, education and financial service centres).
The availability and access to critical service centres play a vital role in reducing damage that may occur during a hazardous event. For example, health-care providers are important sources of relief; lack of proximate health-care service centres is vital for rescue and other health-related issues immediately after disastrous events (Cutter et al., 2003). Moreover, emergency service centres in close proximity can play significant role to reduce the response lag time during emergencies, which in turn would reduce the magnitude of loss. Similarly, the higher density of education and finance institutions in a zone is a credible reason for the access to these resources. The former is critical in knowledge dissemination (regarding various safety aspects before, during and after an extreme natural event) and the latter can guide the communities to cover disaster risks through insurance. Hence, more the number of emergency service centres, less would the impact of an extreme event.
3.3.10 Slum population.
Slums are mainly those residential areas where dwellings are in any respect unfit for human habitation by reasons of dilapidation, overcrowding, faulty arrangements and designs of such buildings, narrowness or faulty arrangement of streets, lack of ventilation, light, sanitation facilities or any combination of these factors which are detrimental to safety, health and morals (SAICA, 1956). UN-HABITAT (2005) defines a slum household as a group of individuals living under the same roof in an urban area who lack one or more of the following: durable housing of a permanent nature that protects against extreme climate conditions, sufficient living space which means not more than three people sharing the same room, easy access to safe water in sufficient amounts at an affordable price, access to adequate sanitation and security of tenure that prevents forced evictions. The inherent characteristics of the slum population are enough to describe their vulnerability to natural hazards.
While evaluating the indicators discussed above (Section 3.3), the overall vulnerability status of communities inhabiting in the different wards of the Srinagar city was found to be highly variable [Figure 5]. The interior wards reflecting high vulnerability, corresponds mainly to the old central business district (CBD) of the city. The zone is inhabited by low income group communities [Figure 5]. Without renewal and redevelopment process since the past few decades, the zone experienced urban decay, thus expressing escalated levels of vulnerability.
With the integration of all the layers water depth, elevation and socioeconomic and demographic data, corresponding to flood hazard intensity, spatial exposure and vulnerability respectively, the flood risk scenario of the Srinagar city became obvious [Figure 6]. The municipal wards that exhibit high flood risk include – Bemina East, Bemina West, Bod Dal, Lokut Dal, Rajbagh, Jawahar Nagar, Aloochi Bagh, Sheikh Dawood Colony, Wazir Bagh, Lal Chowk, Nundreshi Colony, Shaheed Gunj, Barbarshah, Syed Ali Akbar, Ganpathyar, Parimpora, Hassanabad, Madin Sahab, Kawdara, Zadibal, Idd Gah, Hazratbal, Lal Bazar, Maloora, Palpora, Laweypora, Karan Nagar, Chattabal, Magarmal Bagh, Batamaloo, Zainakot, Natipora, and Mehjoor Nagar. The wards expressing moderate flood risk are – Telbal, Qamarwari, Islamyarbal, Nawab Bazar, Jogilankar, Safakadal, Aali Kadal, Malik Aangan, Khankah-i-Moula, Zind Shah Sahib, Akil-Mir Khanyar, Khawja Bazar, Jamia Masjid, S. R. Gunj, Soura, Zoonimar, Umer Colony, Nowshera, Channapora, Rawalpora, Bhagat Barzulla, Khumani Chowk, Pandrathen; and the wards with low flood risk include – Khanmoh, Humhama, Nishat, Harwan, Dara, Alesteng, Buchpora, Dalgate, Mukhdoom Sahab, Tarabal, Zakura, and Ahmad Nagar. With a sizeable population in the low-lying areas (spatially exposed), congestion in core areas, and deprived economic conditions, and sensitive social structure over significant area, the Srinagar city is likely to experience more damages in the identified high risk areas due to floods. Selective vulnerability alleviation measures are needed to mitigate the probable adverse effects that may originate from future flood events. We specify that in addition to structural measures, depopulation of the overcrowded interior parts of the city followed by initiation of the urban renewal and poverty alleviation programmes would reduce the flood risk of the resident population. Moreover, a new land use policy must be framed and implemented for planned and flood-safe expansion of the city.
4. Conclusion
Risk assessment brings into attention the underlying causes that may turn a natural hazard into a disaster. The process is imperative for understanding probability of loss a spatial unit is likely to experience during a future extreme event. The present investigation produced broad flood risk status of Srinagar city based on an integrated assessment model, considering the major determinants of risk, i.e. hazard, exposure and vulnerability. The risk status of the city appeared to be exhibiting considerable variation in relation to the municipal boundaries. In general, the wards in the central, western and south eastern parts of the city express high level of the flood risk; however, the risk is comparatively low especially towards the peripheral zones in south and northeast. The findings of this study may be used as a decision support tool by the stakeholders while building structural and non-structural flood hazard mitigation measures for the Srinagar city. Furthermore, demonstration of adopting a conceptual framework for practice by this study is expected to be useful in determining the comprehensive risk scenario of other cities as well.
This research article is an extension of the project – Flood Hazard Evaluation and Vulnerability Assessment of Upper Jhelum Floodplain in Kashmir Valley – sponsored by Ministry of Earth Sciences (MoES), Government of India. Authors are appreciative to the MoES (India) for providing the financial support to carry out this study. However, it is pertinent to mention that MoES is not responsible for any result interpretations expressed in the paper.
The authors are indebted to the editor (Prof Richard Haigh) and anonymous reviewers for their valuable and constructive comments that substantially improved the quality and structure of the paper. Free dissemination of extensive data acquired through various earth observation programmes by the NASA (National Aeronautics and Space Administration) is really noteworthy.
Figure 1.
Location of the Srinagar city (b) in Kashmir basin (a)
[Figure omitted. See PDF]
Figure 2.
Time series plot of the annual maximum discharge (cusec) at Ram Munshi Bagh gauge station of the river Jhelum in Srinagar
[Figure omitted. See PDF]
Figure 3.
Risk triangle model
[Figure omitted. See PDF]
Figure 4.
Event specific hazard intensity and elevation-based exposure of the Srinagar city
[Figure omitted. See PDF]
Figure 5.
(a) Showing spatial disparities in the various selected social, economic and demographic indicators; (b) composite vulnerability status of the Srinagar city
[Figure omitted. See PDF]
Figure 6.
Flood risk scenario of the Srinagar city
[Figure omitted. See PDF]
Table I.Vulnerability connotations of selected variables
| Model inputs | Indicator | Code | Increases (+) or decreases (−) vulnerability | References |
|---|---|---|---|---|
| Hazard | Water depth | WD | High WD (+) | Cançado et al. (2008), Kreibich et al. (2009) |
| Exposure | Elevation | EL | High EL (−) | IPCC (2012), Sanyal and Lu (2006) |
| Vulnerability | Housing density | HD | High HD (+) | Melki et al. (2004) |
| Crowding index | CI | High CI (+) | Melki et al. (2004) | |
| Population density | PD | High PD (+) | UNEP (2002), Weber et al. (2015) | |
| Child population (age) | CP | High CP (+) | Shannon et al. (1994), Clark et al. (1998), Cutter et al. (2003), Dwyer et al. (2004), Rygel et al. (2006), Flanagan et al. (2011) | |
| Elderly population (age) | EP | High EP (+) | Clark et al. (1998), Cutter et al. (2003), Dwyer et al. (2004), Rygel et al. (2006), Evans (2010), Flanagan et al. (2011) | |
| Female population (Gender) | FP | High FP (+) | Cutter et al. (2003), Dwyer et al. (2004), Rygel et al. (2006) | |
| Working population | WP | High WP (−) | Dwyer et al. (2004) | |
| Income | IN | High IN (−) | Cutter et al. (2003), Dwyer et al. (2004), Flanagan et al. (2011) | |
| Level of education | LE | High LE (−) | Cutter et al. (2003), Flanagan et al. (2011) | |
| Critical service centres (medical, emergency financial and educational) |
CSC | High Density of CSC (−) | Cutter et al. (2003), Oven et al. (2012) | |
| Slum population | SP | High percentage of SP (+) | Present Study |
Flood risk classification
| Hazard (Water depth) |
Exposure (Elevation) |
Vulnerability (Multiple variables) |
Risk | ||||
|---|---|---|---|---|---|---|---|
| Actual value (m) | Scale value | Actual value (m) | Scale vale | Composite value | Scale value | Z-score | Class |
| 4.1-5.0 | 5 | 1478-1576 | 5 | 61-70 | 5 | 1.336 | High |
| 3.1-4.0 | 4 | 1577-1583 | 4 | 51-60 | 4 | 0.801 | |
| 2.1-3.0 | 3 | 1584-1591 | 3 | 41-50 | 3 | 0.267 | Moderate |
| 1.1-2.0 | 2 | 1592-1615 | 2 | 31-40 | 2 | −0.267 | |
| 0.1-1.0 | 1 | 1616-1671 | 1 | 21-30 | 1 | −0.801 | Low |
| 0 | 0 | >1672 | 0 | 10-20 | 0 | −1.336 | |
© Emerald Publishing Limited 2018