Disaster Risk Index (DRI) for Tropical Cyclone of Bangladesh

Disaster Risk Index (DRI) for Tropical Cyclone of Bangladesh

Shamim Ara Bobby

Lecturer, Department of Civil Engineering Stamford University Bangladesh

Abstract

The risk and magnitude of damages to people and property due to natural disasters are very severe in Bangladesh. Cyclone is one of the major natural disasters in context of Bangladesh. Since it is a natural phenomenon, occurrence of cyclone cannot be prohibited. But the occurrence time of a cyclone can be predicted from statistics. So, reduction of losses and possible mitigation measures can be tak en easily if desired. Cyclone not only affects life of people but also has huge impact on social and economic losses. Geographical location of Bangladesh is primarily responsible for cyclone disasters. Although it is a physical phenomenon it affects society, community, people, institutions, environment and the overall development of the country. Due to cyclone hazard it is difficult to meet the targets of the MDGS. There is a very good relationship between development process and cyclone risk . The DRI developed here provides the national decision mak ers access the information to identify risk and propose adequate disaster risk management policies. This paper presented different perspective of cyclone risk assessment and established cyclone Disaster Risk Index (DRI) for the local level for Bangladesh. This will help to prepare and plan habitable human settlements at coastal areas.

1. Introduction

   According to Norwegian Institute for Urban and Regional Research (NIBR), Asia is the most disaster- prone continent, with Ch ina, Bangladesh, India and Iran on top of the list when measured by absolute number of a ffected people (Sø rensen et al, 2006). Impact of Natural Disasters on GDP of Bangladesh, 1990-2000 is 5.21% (DFID Report, 2005). A mong 84

   high cyclone prone countries around the world with highly populated coastal areas and deltas, Bangladesh, China, India, the Ph ilippines, and Japan are at risk (UNDP, 2004). Cyc lone is a characteristic feature of Bangladeshs physical environ ment and accepted as a norma l haza rd of social and economic life of Bangladesh. Historically, cyc lone is mentioned as a disaster in this belt in the book Ain-e-A kbari of 16th century. During the last three decades almost all of the coastal areas and offshore islands of Bangladesh faced cyclones.

   A DFID scoping study found that poverty allev iation, development and disaster risk reduction (DRR) are highly correlated (DFID report 2005). Inadequate attention to DRR can hinder progress in poverty alleviation and sustainable development. Most recently to improve understanding the relationship between development and disaster risk at the global level, UNDP has started development of a Disaster Risk Inde x (DRI) for individual country. In this paper, it is argued that less attention has been given to conducting in depth research on disasters, especially fro m risk perspective. This paper is an atte mpt to as sess the cyclone risk in diffe rent perspective fro m UNDP and attempted to calculate cyclone Disaster Risk Inde x (DRI) for local level coastal areas for Bangladesh, rather for whole country.

2. Scope of the Study

   District leve l DRI for tropica l cyclone is developed in this project. Th is provides the disaster factors through a district-by-district comparison of human vulnerability and exposure to cyclone and the identification of development factors that contribute to the risk. The method used here can also be used for developing DRI for other natural disaster such as flood, earthquake, drought etc. The results obtained here can be used to propose the most suitable location for cyclone shelter.

3. Tropical Cyclone

   A tropical cyc lone is a storm system characterized by a low pressure centre and numerous thunderstorms that produce strong winds and flooding rain. The term "tropical" refers to both the geographic origin of these systems, which form almost e xclusively in tropical regions of the globe, and their format ion in Ma rit ime Tropical air masses . The term "cyclone" refers to such storms' cyclonic nature, with counter cloc kwise rotation in the Northern He misphere and clockwise rotation in the Southern Hemisphere. Depending on their location and strength, tropical cyclones are referred to by other names, such as hurricane, typhoon, tropical storm, cyclonic storm, tropical depression and simply cyc lone.

4. Trends in Tropical Cyclones

   The Bay of Bengal is potentially energetic for cyclonic storms due to its favourable atmospheric and oceanic condition. Therefore, the highest incidence of TCs, a lmost 7% of total number of annual storms,

   located here. Amongst two storms generate winds of 55 mph or greater (as cited in Bo wditch, 1977). Ma inly, two cyclonic seasons are in the north Indian Ocean, v iz. pre-monsoon (especially May) and post-monsoon (especially October and Nove mber). A fe w cyclones form in transitional monsoon months June and September a lso. Singh et al (2001) have shown an increased TC frequency during November and May over the north Indian Ocean from 1877 to 1998.

5. Vulnerability Patte rns of Tropical Cyclones

   There are many dimensions of vulnerability. A mong these social vulnerability refers to the inability of people, organizat ions, and societies to withstand adverse impacts from mu ltiple stressors and shocks like natural disasters to which they are e xposed.

   Mainly people of Bangladesh are vulnerable to disasters for three reasons. Firstly, the important type of vulnerability- poverty. It is an indicator of lack of access to resources, income opportunities and inequitable distribution of powe r. People of Banglad esh are poor and marginalized; it is difficu lt for the m to access resources such as development loans or land. People here are simply unable to recover their losses due to their miserable poverty. In addition to the economic dimension, there are also other aspects of social positioning such as class, religion, co mmun ity structure, community decision making processes and politica l issues that determine poor peoples vulnerability. Therefore , poor are economically as well as vulnerable to social, cultural and politica l capacit ies to cope with disasters.

   Second is vulnerable livelihood of the people. According to DFID (2005), this has three components – live lihood assets (human, social, physical, natura l and financia l capital), livelihood strategies and livelihood outcomes. This sort of vulnerability arises due to shortage of jobs, low income , declining natural resources, and decreasing profitability of rice farming. Frequently cyclones adversely affect the livelihoods of people by damaging their means of earning (destruction of the factory, loss of land due to erosion in flooding, destruction of the shop) and/or tools (loss of draught animals, p loughing tools, etc) in case of Bangladesh.

   Third, patterns of natural resource use are changing, as urban development and commerc ial quarrying and environmental degradation. Localized and systemic environmental degradation is becoming highly influential as we ll, lowe ring the natural resilience to cyclones. Phenomena like El Niño/ La Niña, climate change and the potential for rising sea levels, are affecting the patterns and intensity of cyclones .

6. Disaster Risk Approaches

The DRI enables the calculation of the average risk of death per country in la rge- and mediu m-scale disasters associated with earthquakes, tropical cyclones and floods, based on data from 1980 to 000. It a lso enables the identification of a number of socio-economic and environmental variables that are corre lated with risk to death and which may point to causal processes of disaster risk.

Since 1970s, engineers, architects, social scientists, social activists all have been highlighted on the impact of natural haza rds. They find that the impacts not only depend on the physical resistance of a structure, but on the capacity of people to absorb the impact and recover fro m loss or damage. Now it is clear that development processes were not only generating different patterns of vulnerability, but were also changing and increasing patterns of hazard due to events such as impact of global climate change. Finally, d isasters are no more viewed as ext re me events created entirely by natural forces but as unresolved problems of develop ment (Yod mani, 2000). Thus the focus of interest moved to social and economic vulnerab ility, with increasing evidence that natural hazards had widely v arying impacts on different social groups and on different countries. According to Munich Re during the last four decades economic losses have increased more than ten times per decade. The causal factors of disaster thus shifted from the natural event towards the development processes that generated different levels of vulnerability. Disaster Risk has been presented by Ward, 1999 as fo llo ws

Manageability here stands for the degree to which a community can intervene and mange a ha zard in order to reduce its potential impact. Manageability is synonymous to Capacity, so this can be substituted to have the following disaster risk formu la :

7 Development of Disaste r Index

Disaster Risk can be e xpressed in different ways for e xa mple by the number o f people killed, percentage killed or percentage killed as co mpared to the e xposed population. As per the UNDP report the first two are used for DRI for cyclone calculation, and the percentage killed as compared to the exposed population for calculating relat ive vulnerability. Though exposed populations to cyclone should not be compared without standardization.

UNDP Bu reau for Crisis Prevention and Recovery made a global report Reducing Disaster Risk – A Challenge For Deve lopment (2004) that provides the Disaster Risk Inde x (DRI), as the first global assessment of disaster risk factors through country -by- country. It enables the calculation of the average risk of death per country with comparison of human vulnerability and exposure to three critical natural hazards: earthquake, tropical cyclones and flooding. It also identified of development factors that contribute to disaster risk. In the DRI, this relationship is expressed through the concept of physical e xposure, referring to the number of people located in areas where hazardous events occur combined with the frequency of hazard events. Physical e xposure is not an indicator of vulnerability, but is a condition sine qua non for disaster risk to e xist. It a lso enables the identification of a number of socio-economic and environmental variables that are corre lated with risk to death and which may point to causal processes of disaster risk. In the DRI, countries are inde xed for each ha zard type according to their degree of physical e xposure, their degree of relative vulnerab ility and their degree of risk which is calculated using follo wing fo rmula

(1)

Nu mber Where, Ph Exp is the physical e xposure, i.e. the frequency and severity multip lied by e xposed population; Vulnerability is the concept that expla ins why, with a given level of physical e xposure, people are more or less at risk. In theory, vulnerability is modified by coping capacity and adaptive capacity.

Two methods are available for calculating physical e xposure. First, by multiplying hazard frequency by the population living in each e xposed area. The frequencies of hazards were calculated for different strengths of event, and physical e xposure was computed as follo ws,

(2)

Where, PhExp is the physical exposure at national level, Fi is the annual frequency of a specific magnitude event in one spatial unit; Popi is the total population living in the spatial unit.

A second method was used when data on the annual frequency of return of a specific magnitude event was not available. In this case (Cyclone), physical exposure was computed by dividing the e xposed population by the numbers of years when a particular event had taken place as shown below

(3)

Where, Popi is the total population liv ing in a particu lar buffer, the radius of which from the eye varies according to the strength, Yn is the length of time in years PhExp is the total physical e xposure of a country (here used for second level administrative area, na mely district), in other words the sum of all physical e xposure in this country (here district).Frequency of cyclone is calculated based on the following formu la,

(4)

Where, E(x) is the statistical e xpectation, i.e . the average number of events per year, P(x) is the probability of occurrence. Here , to obtain physical e xposure, a frequency per year was derived. Then population was extracted for the selected district and mu ltip lied by frequency in order to obtain the average yearly physical e xposure (BBS, 2006). Here , equation

(2) is used to ascertain physical exposure of population of a particula r locality. Due to unavailability of data no. of killed in a district is taken as the arithmetic average of the total killed fo r simplic ity for a g iven storm that affects the different district. Affected districts for a given storm are ta ken form " EM-DAT: The OFDA/CRED International Disaster Database.

1. Cyclone DRI and Bangladesh

   DRI help to assess cyclone disaster risk for the individual areas of Bangladesh to find a holistic picture of the country. It calibrated according to the risk of death between 1980 and 2007. He re, data fro m 23 coastal districts of Bangladesh were selected for DRI calculation fro m 1980 to 2007. This is because of the data before this period is not reliable . 191 cyclones in the period of 28 years have been considered. Total population of these coastal areas is 43.7 million. The key steps follo wed in producing the DRI, according to UNDP report, were :

   1. Calculation of Physical Exposure

      The DRI ca lculated for the areas exposed to cyclones and the population living in these areas to arrive at a calculation of phys ical e xposure for each district. This is the average number of people e xposed to a hazard event in a given year. Physical e xposure varies both according to the number of people as well as to the frequency of cyclone. In the DRI, physical e xposure is e xpressed both in absolute terms (the number of people e xposed in a country) and in relative terms (the number e xposed per million people). When more people are killed with respect to the number exposed, the relative vulnerability to the hazard in question is higher. BBS (2006), population data have been used to calculate physical e xposure.

   2. Calculation of Relative Vulnerability

      Vu lnerability of locality depends on the physical e xposure of a spatial unit and no of the death for a

      particular disaster. Here, for a set of cyclones and for a district, relative vulnerability is the total number death in million to the exposed population to thos e cyclones. Total number of death for each cyclone was taken from EM-DAT database.

   3. Results and Discussion

      Figure1: revea ls that loss of human life fro m cyclone is tied to development status of a region. No developed region has recorded more than 200 deaths per million

      Table 1

      Disaster Risk for tropical Cyclone, 1980-2007

      District Name	Avg. No. of Storm per Year	No. of people killed per year	No. of people killed per million	Avg. physical exposure	physical exposure in of population	Relative Vulnerability	Risk
      Bagerhat	0.0649	23.5	15.19	100512.9	6.488762	234.16	23.5357
      Jessore	0.0105	4.5	1.82	26016.6	1.052641	172.97	4.5000
      Khulna	0.0649	22.7	9.55	154365.8	6.488762	147.15	22.7143
      Satkhira	0.0593	25.5	13.66	110607.9	5.931658	230.22	25.4643
      Shariatpur	0.0105	16.1	14.88	11392.74	1.052641	1413.81	16.1071
      Munshigonj	0.0052	0.6	0.47	6792.52	0.524936	89.38	0.6071
      Madaripur	0.0105	13.2	11.53	12066.94	1.052641	1095.08	13.2143
      Gopalgomj	0.0105	13.3	11.43	12266.14	1.052641	1086.03	13.3214
      Faridpur	0.0052	0.4	0.20	9220.336	0.524936	38.73	0.3571
      Noakhali	0.0818	906.2	351.61	210791.3	8.178943	4298.94	906.1786
      Lakshipur	0.0265	1.1	0.74	39522.24	2.652675	28.01	1.1071
      Cox's Bazar	0.1106	846.3	477.11	196189.7	11.06099	4313.43	846.2500
      Chittagong	0.1224	948.7	143.48	809158.7	12.23747	1172.43	948.6786
      Feni	0.0265	72.5	58.42	32903.36	2.652675	2202.34	72.4643
      Comilla	0.0158	1.6	0.35	72754.4	1.583147	22.09	1.6071
      Bandarban	0.0052	1.5	5.15	1564.938	0.524936	981.33	1.5357
      Patuakhali	0.1106	926.4	634.20	161576.8	11.06099	5733.67	926.4286
      Pirojpur	0.0649	19.9	17.87	72094.56	6.488762	275.43	19.8571
      Jhalakhati	0.0105	12.6	18.21	7307.762	1.052641	1730.06	12.6429
      Bhola	0.0818	927.6	544.65	139297	8.178943	6659.21	927.6071
      Barisal	0.0538	23.0	9.75	126695.4	5.37764	181.26	22.9643
      Barguna	0.0761	849.9	1001.62	64595.07	7.61237	13157.79	849.9286
      Chandpur	0.0105	0.2	0.09	23907.89	1.052641	8.96	0.2143
      Total Bagladesh	3.43	7467.62	64.02	1.5E+08	343	49.77

      The size of e xposed populations with the number of recorded deaths to cyclones is used as a measure of relative vulnerability as shown in Figure 2. It should be noted from the Figure 3 that the highest no average cyclone occurs in south eastern main ly Chittagong area. Though having highest no of death and highest no of

      population from 1980-2007.

      This observation reinforces intuitive v iews about the disaster development relat ionship. This heightens the inequality especially in rural char areas. Livelihoods there at risk due a range of factors: poverty and asset depletion, environmental degradation, ma rket pressures, isolation and remoteness, the weakness or lack of social services and climate change. Therefore, contemporary patterns of urbanization and rural live lihoods needs to be viewed alongside other critical development pressures.

      Fig. 2 Relative vulnerability at district level for 1980 to 2007 cyclone

      Fig. 3: Average storm per year for individual district

      www.ijert.org 4

      average cyclones per year this area is less vulnerable than other coastal region. Mainly, physical e xposure in million populations is very large, and the area is more developed than other areas. The analysis carried out for tropical cyclone risk showed a strong correlation between physical e xposure and human d evelopment. An area with large, predo minantly rura l populations and with a lo w development is most closely associated with cyclone risk. There are a number of reasons for this. Rural housing in many areas are tend to be more vulnerable to high winds, flooding and landslides than urban housing and are generally be associated with higher mo rtality. Conversely, the weakness or nonexistence of emergency and rescue services in rural areas and lack of access to disaster preparedness and early warning a re all other factors that would help to e xpla in mortality rates. The cyclone preparedness programme in Bangladesh is one of the few success stories. By coupling cyclone shelters and community- based preparedness measures, the programme has managed to dramat ically reduce vulnerability fro m the 1970s to the (still high) levels observed in the 1980- 2007 report ing period (Tab le 2)

2. Conclusion

One of the majo r findings of our study is the local level Disaster Risk Inde x (DRI) for cyc lone. The value of DRI indicates the vulnerable areas for cyclone; where cyclone is likely to affect the local co mmun ity severely. Therefore, fro m our study it can be concluded

Table: 2 Impacts of Cyclone

Source: DFID (2005)

that DRI can be a useful tool for s electing the suitable location for Cyc loe shelters in coastal areas of Bangladesh.

Again, it is observed that there is a good relation between cyclone risk and development level of the coastal areas of Bangladesh. DRI de monstrate the ways to assess the significance and role of cyclone hazard in the national development. The DRI developed here, with a national level of observation and a local level of

resolution that would enable the identification and e xplanation of re lative risk and vulnerability, have enormous potential to support national development planning. But without the responsibility of the Govern ment and participation of the local people, sound development cannot be achieved. One of the ma jor goals of MDGs is sustainable development. To meet the targets of MDG d isaster risk due to cyclone could be reduced significantly.

10 Recommendation

There are limitat ions of the data source used for evaluating cyclone DRI. The data used here is based on some drastic assumption wh ich may severely affect the result. The value of DRI would be more re liab le if proper investigation is carried out to accumulate the number of people killed in an area. Future wo rk can be taken to develop disaster risk for flood, drought and earthquake follo wing the procedure described in this

study. Work can be taken to build on indiv idual indices for cyclone, earthquake and flood to form a mu lti- hazard DRI. This mult i-ha zard DRI can be a sharp tool for policy advocacy.

11. References

1. Bangladesh Bureau of Statistics, 1991. Statistical Yearbook of Bangladesh, Statistical Division, M inistry of Planning, Government of the People's Republic of Bangladesh, p. 25-29.

2. Bangladesh M eteorological Department (M eteorological Headquarter) M eteorological Complex, Agargaon, Dhaka-1207.

3. BBS (2006) Bangladesh Bureau of Statistics, St atistics Bangladesh, 2006.

4. Center for Environmental and Geographic Information Services.

5. \Choudhury, A.M ., 1992. Cyclones and the coastal zones of Bangladesh, The Bangladesh Quarterly, Dept. of Films and Publications.

6. DFID (2005), Natural Disaster and Disaster Risk Reduction M easures- A Desk Review of Costs and Benefits, Report, 8 December, 2005.

7. Henderson-Sellers, H. Z.; Berz, G; Emanuel, K.; Gray, W.; HYPERLINK

   "http://en.wikipedia.org/wiki/Christopher_Landsea"Land sea, C.; Holland, G.; Lighthill, J.; Shieh, S- L.; Webster,

   P. and M cGuffie K.(2008), Tropical Cy clones and Global Climate ChangeHYPERLINK "http://www.aoml.noaa.gov/hrd/Landsea/IPCC/index.ht ml":HYPERLINK

   "http://www.aoml.noaa.gov/hrd/Landsea/IPCC/index.ht ml" A Post-IPCC Assessment.

8. Huq, S., Karim, Z., Asaduzzaman, M . and M ahtab, F. (eds.). 1999. Vulnerability and Adaptation to Climate Change for Bangladesh. Kluwer Academic Publishers, Dordrecht, the Netherlands.

9. M arcelino, E. V.; Isabela P. V. and Frederico, M .R. (2004), Cyclone HYPERLINK "http://www.dsr.inpe.br/geu/Rel_projetos/Relatorio_IAI

   _Emerson_M arcelino.pdf"CatarinaHYPERLINK

   "http://www.dsr.inpe.br/geu/Rel_projetos/Relatorio_IAI

   _Emerson_M arcelino.pdf":HYPERLINK

   "http://www.dsr.inpe.br/geu/Rel_projetos/Relatorio_IAI

   _Emerson_M arcelino.pdf" Damage and Vulnerability Assessment, Santa Catarina Federal University.

10. Shultz, J. M .; Russell, J. and Zelde E. (2005), Epidemiology of Tropical CyclonesHYPERLINK "http://epirev.oxfordjournals.org/cgi/content/full/27/1/21 ":HYPERLINK

    "http://epirev.oxfordjournals.org/cgi/content/full/27/1/21

    " The Dynamics of Disaster, Disease, and Development.

11. UNDP (2004), Reducing Disaster Risk A challenge for development, a global report. Bureau for Crisis Prevention and Recovery, New York.

12. World M eteorological Organization (2006), RSM Cs,

    Tropical Cyclone Programme (TCP).

13. Yodmani S (2000) Paper Presented on Disaster Risk M anagement and Vulnerability Reduction: Protecting the Poor at The Asia and Pacific Forum on Poverty, Organized by the Asian Development Bank.