Page 1 of 17 Bangladesh’s Early Warning System – Normalising the Emergency? Redwan B Rokon May 2016
Page 1 of 17
Bangladesh’s Early Warning System – Normalising the Emergency?
Redwan B Rokon
May 2016
Page 2 of 17
Introduction
Many coastal communities across the globe are exposed to threats from cyclones and
storm surges. Tropical storms – like cyclones, hurricanes and typhoons – have caused more
than two million deaths and millions of injuries worldwide over the past two centuries. The
frequency of cyclones increased more than threefold between 1970 and 2006 (International
Disaster Database 2016). In South Asia, the Bay of Bengal region is considered a breeding
ground for tropical cyclones and storm surges (Haque et al. 2011).
Bangladesh is situated in the mouth of the Bay of Bengal, and is ranked sixth in the
long-term Climate Risk Index (CRI) (Kreft et al 2014). Recurring floods and cyclones, the
two most common natural disasters, have claimed hundreds of thousands of lives, and caused
severe socio-economic damages over the past few decades. Geographically, fourteen of the
nineteen coastal districts of Bangladesh are high or moderate cyclone areas, and home to
around thirty and a half million people (MWR 2005). The complex coastline and low-lying
plains with high population density make Bangladesh vulnerable to natural hazards. Due to
the commonality and high frequency of disasters in the region, the Government of
Bangladesh (GoB) gives high importance to efficient disaster management (Roy et al. 2015).
They put a strong emphasis on reducing economic, human, and environmental losses of
disasters through improved disaster management capacity (Habib et al 2012). Over the past
50 years, there have been remarkable improvements in the reduction of cyclone-related
deaths. The government and supporting institutions have developed ways to adapt to
recurrent cyclones by improving early warning systems, establishing shelter and evacuation
plans, building coastal embankments, and raising community level awareness (Haque et al.
2011).
Although these new developments are adding value to the disaster management
system, there is still an increased need for higher number of shelters in the areas that are most
vulnerable, especially those disconnected from the relief and response centres (Haque et al.
2011). The modernisation of the early warning system, compared to what it was a decade
ago, has been vital in reducing casualties. The numbers of deaths from the cyclones of 1970,
1991, and 2007 were 500300, 138958, and 4234, respectively (Dasgupta et al. 2010). It is
important to note that the recent numbers are still high, and thousands of people live in the
vulnerable coastal regions.
Page 3 of 17
Despite the loss of lives and infrastructure from disasters, Bangladesh’s disaster
management is highly commended nationally and internationally for its improved success in
the past three decades. However, the prevailing high numbers of deaths from cyclones raise
several questions. Behind their narrative of progress and success, is Bangladesh legitimising a
‘normality’ that obscures the extent of the lingering sufferings? Does the behaviour of the key
players demonstrate a notion of guilt or innocence? Is the vulnerability caused by prejudice
against certain groups, or is it a result of personal choices? Is the early warning system not
effective enough to evacuate people during the lead time of a cyclone? To answer these
questions, this paper investigates the early warning system of Bangladesh, using Mark
Bradbury’s (2002) theory of normalisation of crisis, Cannon’s (2008) definition of guilty and
innocent disasters, and Mary Douglas and Aaron Wildavsky’s (1982) cultural theory of risk.
Through a qualitative approach of extensive literature review and few semi-structured
interviews, this research finds elements of guilt and innocence within the system, and an
evidence of normalisation of the protracted vulnerability. The United Nation International
Strategy for Disaster Reduction (UNISDR) early warning framework is employed as the
foundation for the interviews.
The organisation of this paper is as follows. The next section analyses existing
literature to establish the background and theories that motivate this research. The following
section gives an overview and summary of the interviews conducted with officials in
Bangladesh’s disaster management system. The last two sections connect the dots between
theory, existing literature, and interviews to answer the research question, and to suggest
future research on this topic.
Literature Review
Vulnerability and Hazards
The increasing prominence of disasters emphasises the need for better disaster
preparedness and response mechanisms that can efficiently overcome vulnerabilities and
reduce the consequent losses. Vulnerability includes political, socio-economic, environmental
and/or physical factors that aggravate the capability of communities to respond to hazards
(Jegillos 1999). Pelling (1999) defines vulnerability in two ways – a focus on risk exposure,
and a coping and adaptation mechanism. Overall, when a community is vulnerable to a
hazard and cannot cope with its repercussions, it becomes a disaster (ISDR 2006a).
Page 4 of 17
Early Warning System
To prepare for effective response, countries need to focus on the establishment of
well-functioning and efficient Early Warning System, EWS (Deltares, 2015). The UNISDR
(2007) defines EWS as “the set of capacities needed to generate and disseminate timely and
meaningful warning to enable individuals, communities and organizations (threatened by
hazards) prepare, and act appropriately and in sufficient time to reduce the possibility of harm
or loss.” It allows people to prepare for an upcoming hazard, and permits associated
institutions to arrange and relocate vulnerable people to shelters. An effective EWS has four
components – detecting, monitoring and forecasting a hazard; analysing the risks involved;
disseminating timely and effective warnings; and, executing emergency preparedness and
response plans (ISDR 2006b).
EWS helps both developed and developing countries. In the United States, tornado
mortality reduced by 45 per cent and injuries by 40 per cent in between 1986 to 1999 due to
timely early warnings (Teisberg and Weiher 2009). Similarly, Bangladesh lowered mortality
by developing an efficient EWS. In 2007 before the Cyclone Sidr, Bangladeshi authorities
evacuated 3 million people from their homes and accommodated most of them in shelters.
Although there were about 55,000 injured and 4,234 dead or missing, the numbers could have
been significantly higher without an effective EWS (Golnaraghi, 2012). On the other hand,
inefficient early warning with poor dissemination of information and lack of government
capacity can lead to catastrophic outcomes. For example, when Cyclone Nargis hit Myanmar
in early May 2008, it caused around 140, 000 deaths and severely affected around 2.4 million
people. Poor warning system and lack of capacity led to inadequate preparations for timely
evacuation and shelter during Nargis.
A widely accepted agenda for early warning system is the UNISDR’s framework. It is
formulated according to the second priority of the Hyogo Framework for Action (HFA) 2005,
which is a blueprint that is agreed and endorsed by many governments, international agencies
and non-governmental organisations (NGOs) (IFRC 2008). This priority focuses on
enhancing people centred early warning, and formulating systems that allow timely
dissemination of information that is comprehensible to those at risk, policy makers, and the
associated actors in the system. The HFA calls for a paradigm shift in the disaster
management – from emergency response and vulnerability mitigation to a focus on response
and resilience. With an attempt to have a more proactive, holistic and systematic approach to
risk reduction and response, it stresses on creating opportunities for the vulnerable population
Page 5 of 17
to spring back to their normal livelihoods after a disaster. Further, it acknowledges the
importance of possessing knowledge of the hazards and the physical, social, economic and
environmental vulnerabilities to societies (ISDR 2005).
The UNISDR’s new approach to disaster management includes a four-step early
warning framework (ISDR 2006a). First, “risk knowledge” indicates systematic risk
assessment and mapping that helps in setting priorities and guides the expected response
activities. Second, “monitoring and warning” refer to a continuous monitoring of climate
change and meteorological factors to generate accurate and timely warnings. Third,
“dissemination and communication” denote collaboration with people and local communities
who are at risk, giving them comprehensible information for successful and effective
response. Fourth, “response capability” requires national and local communities to accurately
understand risk and response accordingly. This requires the formal and informal education
sectors to teach locals about the overall concepts of risk and vulnerability of their
communities. It is important for all these elements to work in coherence with each other;
failure of one can lead to the failure of the entire system.
Understanding Bangladesh’s Disaster Management System
The human settlements in the coastal regions of Bangladesh are formed in
unorganised and isolated manner, making the task of government initiatives nontrivial.
Geographical setting becomes an important part in disasters. The Bay of Bengal, enveloping
the south of the country, is one of the most dangerous cyclone basins in the world (Bern et al.
1993). With a steep growth in the population rate, more and more people live in flood plains,
and as a result, continue to be highly susceptible to disasters (Hewitt 1983; Paul and Rahman
2006). Due to their long history of living in these areas, coastal population are reluctant to
migrate despite their high vulnerability to natural hazards. Unfortunately, thousands from
these communities lose their lives and property when a natural disaster hits.
In an attempt to reduce this vulnerability, the Government of Bangladesh (GoB),
International Federation of Red Cross and Red Crescent Societies (IFRC), and
nongovernmental organisations put an emphasis on improving the early warning system.
Primary focus of disaster management in Bangladesh used to be centred on rehabilitation and
relief activities. But, there has been a significant shift towards the adoption of a holistic
approach of hazard identification and mitigation. With suggestions from the United Nations
(UN), the GoB together with the IFRC established a Cyclone Preparedness Programme
Page 6 of 17
(CPP). Their mandate is to disseminate cyclone warning to the local communities using
IFRC’s volunteer network. The CPP has been a successful initiative because of its simplicity,
focus on socio-cultural aspects, and an extensive community based volunteer network (Habib
et al. 2012). Along with the CPP, the government has a framework for disaster management
and risk reduction. This framework includes the Disaster Management Act 2012, National
Plan for Disaster Management, National Disaster Management Policy, and the Standing
Order on Disasters (SOD), among others (Habib et al. 2012). In addition, the Bangladesh
Meteorological Department (BMD) plays the leading role in forecasting and creating early
warnings before a cyclone.
Bangladesh has a two phase risk reduction framework. The first phase involves
forming the foundations for long-term risk reduction, and the second phase focuses on
building holistic partnerships. The former strengthens the policy and planning frameworks
within the Ministry of Disaster Management and Relief. The latter is an extension of the first
phase. It comprises of the management of risks and results of disasters, with an emphasis on
the involvement of local communities to increase the protection of lives and assets (Khan and
Rahman 2007).
Beside the early warning system, the coastal population depends largely on their
indigenous warning indicators. The knowledge of such indicators is rooted to a locality and
can vary from one village to the other. Howell (2003) identifies the local indicators as
weather patterns (like the sky turning gloomy and overcast, east wind blowing at full moon,
and so on), sea/river patterns (like big waves/dark rolls of water, cloudy shapes in the sea),
animal behaviour (like cattle becoming restless, bees moving in clusters, dogs wailing
continuously at night, and so on), and others (like new leaves falling to the ground, muddy
smell in the wind). Many of these indicators alone can be termed as normal weather
conditions, but a combination of two or three can be a warning.
Normalisation of Crisis
Normalisation of crisis is a ‘creeping acceptance of higher levels of vulnerability,
malnutrition and morbidity’ (Bradbury 1998, P 330). State of severe instability and associated
humanitarian crises can occasionally be redefined as opportunities for development. This
changes the perceptions of what constitutes an emergency (Macrae and Bradbury 1998). In
Somalia, around 3000 people were dying on average every day from starvation during the
famine in 1992 (Bradbury 1997). Even though it was considered a severe epidemic, by the
Page 7 of 17
end of 1993, the US troops were asked to pull out, and the emergency was deemed to have
ended. Somalia remained in a state of a chronic disaster under most standards. Interestingly,
the UN declared in early 1997 that the major humanitarian crisis was over, and that daily life
of Somalis was just ‘very difficult’ (UNSC 1997). Accepting protracted difficult conditions
instead of classifying them as humanitarian crisis is a process of normalisation. Likewise, in
2003, over 50 per cent of the population in Malawi did not have access to essential drugs.
Although this was unaccepted in humanitarian standards, the authorities normalised the
situation. Griekspoor et al. (2004) say that unacceptable situations can often be accepted
because the actors on the ground – humanitarian or development – cannot cope with the
overwhelming needs of a particular disaster.
Innocent and Guilty Disasters
Economic, political and social constructions define the magnitude of disaster on a
population. Vulnerability does not only refer to people being susceptible to a natural hazard
or impact. Rather, social characteristics can determine the chances for people to be harmed to
a greater or lesser extent (Cannon 2006). Cannon (2008) identifies social constructions of
disasters in two ways – innocent and guilty. The former is when vulnerability is shaped by
social forces and choices, and does not imply exploitation. This is when people situate
themselves in vulnerable places to pursue their choice of livelihood, without any outside
force. No class, institution, or actors can be held responsible for their situations. Since there is
no socially-differentiated system that forces them to live in vulnerable places, such
constructions can be labelled as innocent. Even in developed countries like the United States,
people move to naturally susceptible and dangerous areas (like regions in California or
Florida) with the knowledge that they are at a risk. While on the other hand, guilty is when
specific economic and social factors decide how some groups of a population will be more
affected than others. This happens when people are not usually at risk of hazards but are
compelled to be victims by dominant, enforcing factors. In other words, their opportunity to
live safely is restricted by factors like class, ethnicity, gender, and as a result, they are placed
in an area that is susceptible to a hazard. The Izmit earthquake in Turkey 1999 is an example
where the disaster to the community was largely determined by inequality, exploitation and
corruption.
A crisis cannot always be categorised solely as innocent or guilty (Cannon 2006,
2008). There are complex political, social and economic issues that aggravate the impacts of
Page 8 of 17
cyclones, and increase vulnerability. These social structures can lead to varying perceptions,
and consequently, varied responses to risk and hazards. Mary Douglas (1978), and Douglas
and Wildavsky (1982) view risk perception and interpretation with a sociological approach.
This perspective, known as the cultural theory of risk, looks at risk not with a technical
objective (where risk is defined as the product of ‘extent of damage’ and ‘the probability of
occurrence’), but with a social construction process. Central to their theory is the idea that
risk perception is closely tied to cultural adherence and social learning. People’s social
participation and their attachment with certain types of groups determine their classification
of risk. A group of people can choose what to fear and how much to fear. The social relations
of people are organised with shared self-confidence and shared fears that determine their risk
adoption and aversion strategies (Douglas and Wildavsky, 1982). This theory can explain the
differing attitude towards risk and risk reduction of the Bangladesh government and the
coastal population.
Interviews
The UNISDR framework is used to conduct three semi-structured interviews with
officials from different sections of Bangladesh’s disaster management system.
The interviewees were Major General (Retd.) A. R. Rokonuddawla, former
Bangladesh Ambassador, who worked as an Advisor to the Tripartite Group in Myanmar,
Staff Officer to the Force Commander in the Operation Sea Angels after the devastating
cyclone in Bangladesh (1991), and participated in various disaster relief operations with the
Bangladesh Army; Mr Khaled Masud Ahmed, Disaster Management Delegate of the IFRC,
who was actively involved in the formation of the CPP in Bangladesh; and, Mr Khalid
Mahmood, Joint Secretary in the Ministry of Disaster Management and Relief of the
Bangladesh Government. A summary of these interviews is given in Table A.
Page 9 of 17
Table A: Summary of Interviews
Name Risk Knowledge Monitoring and Warning Dissemination Response
Major General A. R. Rokonuddawla
The Ministry of Disaster Management and Relief is very active. There is smooth coordination, but a lot of room for improvement. There is knowledge of risk, but often it is not widely circulated across the institutions, mainly the local NGOs.
Right parameters are monitored with the advanced scientific technology available in the Bangladesh Meteorological Department (BMD) and Space Research and Remote Sensing Organisation (SPARRSO). Cyclone warnings are formulated timely before the actual calamity.
Elaborate and timely warnings do not reach everyone at risk. Lack of education and exposure make understanding difficult.
The mechanism is updated and reformulated annually. There are no 'fire-fighting' drills.' Lack of manpower and resources act as drawbacks. Local knowledge is used for easier implementation. People are never ready and not everyone, especially elders, at risk evacuate to shelters
Khaled Masud Khan
Government has institutional arrangement at different levels of the state, determined by the Standing Orders of Disasters (SOD). The national revenue structure includes disaster management. There are hazard maps that can track cyclones for the last 50 years.
Cyclones give a lead time of 4 to 5 days, which allows the Bangladesh Meteorological Department (BMD) to forecast accurately.
The CPP disseminates using its volunteers. Messages are translated via megaphones, sirens and flags. People receive warnings, but at different times due to differences in proximity from the volunteers' offices/hubs.
Annual reviews are done by CPP, ICRC and the government. The volunteers are trained regularly. Knowledge of disasters is now embedded in the education system. Many people do not evacuate to shelters with the fear of looting, false warnings, and losing their capital/assets.
Khalid Mahmood
The government's Standing Order 2010, and various Disaster Management Acts allow the creation of maps and modelling. These are circulated to the concerned institutions and players to spread the risk knowledge.
The parameters of monitoring are developed in line with the HFA priority action. BMD and Flood Forecasting and Warning Centre (FFWC) are vital at developing the forecast. BMD and FFWC generate warnings with the formation of depression in the Bay of Bengal.
BMD forecasts the cyclones, and CPP disseminates. The warnings by BMD and FFWC are clear and useable. All the locals know the risk scenario; hence they understand the vulnerability.
Disaster Management Plan according to the Disaster Management Act 2012 allows national and local development of the response plans. People understand the risks, and evacuate to shelters, resulting in tje reduction of casualties.
Page 10 of 17
Discussion
The GoB’s disaster management mechanism is structured by the Standing Order on
Disasters (SOD), and includes various government ministries, nongovernment organisations,
the CPP, and local authorities, among others. The meteorological department works well to
scientifically collect and formulate risk assessments, and to monitor hazards for early
warning. In addition, the CPP uses the wide volunteer network of the IFRC to disseminate the
warnings. Overall, the annual review under the Disaster Management Act 2012 demonstrates
that Bangladesh has the institutional arrangement to continuously improve the early warning
system (Mahmood 2016). They proactively upgrade it, adhering to the UNISDR early
warning framework. The presence of disaster management fund in the national revenue
structure is an added positive (Khan 2016). Further, by including disaster in the national
education system, and also regularly training volunteers and locals, the system is building a
more resilient response capability. Institutional and technological improvements make
Bangladesh better alert to natural disasters, especially cyclones, and more capable to reduce
the number of casualties. As such, Bangladesh is doing well compared to a couple of decades
ago (Rokonuddawla 2016).
Despite the improvements, the interviews and literature show gaps that fit the notion
of guilt in the ‘dissemination and communication’ process. The UNIDSR framework requires
warnings to reach all those at risk, and an assurance that there is complete understanding.
Due to the complex geography of vulnerable areas, it is difficult to confirm that the warnings
reach on time. In Bangladesh, warnings do not reach the vulnerable people at the same time
as many people live in remote areas, which are far-off from the CPP volunteer offices/hubs
(Khan 2016; Rokonuddawla 2016). Among those who receive the warnings, many lack the
education and awareness to understand the grave consequences of such disasters
(Rokonuddawla 2016). Often the rural women and dependant members of the households do
not understand the warnings entirely (Howell 2003). The selective nature of the
dissemination and communication indicate guilt. Many of these vulnerable people become
victims as the system fails to transcend the geographical and social obstacles, and as a result,
only certain people are informed on time to be safely evacuated to shelters.
Additionally, the interview with Mahmood (2016) also indicates elements of guilt
from the government’s perspective. He claims that the locals know the risk scenario to
understand their vulnerability, and therefore, evacuate to shelters as a response to warnings.
Page 11 of 17
Although the reduction of the numbers of casualties between comparable cyclones proves
some validity of his claims, it is incorrect to perceive that all the local people understand the
warning and evacuate to shelters. The existing literature and other interviews verify that
many people who receive warnings on time do not evacuate to shelters because, either their
livelihood restricts them to be ‘prepared’ to move, or they do not trust the warnings. In the
first case, Rokonuddawla (2016) says that the coastal people are mostly farmers, who rely
heavily on seasonal cropping. During the time of harvest, they are reluctant to evacuate,
regardless of the warnings. The cost of missing the harvest can lead to severe consequences –
both financial and personal. Many of these farmers do subsistence farming, relying on
sharecropping methods. They often lack the collateral to fall back to in the case of an
emergency. Hence, they perceive the cost of losing their harvest higher than the cost of losing
their lives. This is the reason why they are never prepared to move. In the second case, Khan
(2016) says that false warnings in the past discourage many locals to trust warnings. Even
when they do, the high rates of theft and looting in the coastal areas discourage them to leave
their households and productive assets. Similar to the first case, the cost of losing their assets
and belongings discourage people, especially the male household members, to evacuate.
Instead, they send the women and dependants to the shelters, and purely depend on their luck
and God’s mercy to battle the repercussions of a cyclone. In Cannon’s (2008) words, these
vulnerable people ‘get in the way’ of a hazard, and therefore, become victims of disasters.
The livelihood of the coastal people is innocent as their settlement is not forced by
any type of prejudice or administration. People choose to live and work in susceptible areas
on their own will due to the nature of their livelihoods. The issue becomes severe when these
people become reluctant to evacuate after receiving an early warning. Rokonuddawla (2016)
also says that lactating women and elderly people are usually the most vulnerable. They are
affected due to their immobility, and the lack of facilities offered in shelters. This can be
termed as guilty because the existing shelters do not have the infrastructure to meet
everyone’s needs. It rather provides a place for a selected, healthy group. In response to this,
Khan (2016) says that the GoB is taking measures to increase the number of cyclone shelters,
and equipping them with adequate facilities for special needs in the future.
Although the GoB is improving their EWS annually, the coastal regions remain
highly susceptible to disasters. With reference to Bradbury’s definition, this paper questions
if the lingering vulnerability and sufferings of the people are normalised by the actors
involved in the EWS in Bangladesh.
Page 12 of 17
From the government’s point of view, there is an acceptance that Bangladesh suffers
from high risks and vulnerability, and that there is a need for continuous efforts to improve
the EWS. However, there is a general consensus among the interviewees and existing
literature that the EWS system in Bangladesh is successful. The achievement of reducing the
number of casualties overrides or obscures the extent of the protracted vulnerability of the
coastal population. From the locals’ point of view, there is either a distrust of the warnings or
a reluctance to obey them. In either case, the possibility of economic or social loss
discourages their response to warnings. In both the instances, governments and local people
are normalising the crisis due to their differing perceptions of disaster. This can be explained
further by the cultural theory of risk (Douglas and Wildavsky 1982).
The commonality of cyclones shapes the perception of people. Government defines
success through the decreasing number of deaths, but fails to acknowledge that every death is
an indication of their failure. Like the cases of Malawi and Somalia, the unacceptable number
of casualties and difficult circumstances are perceived to be normal behind the curtains of
their comparative success over the past few decades. For the local communities, their long
history in the coastal regions and their protracted sufferings impact how they perceive and
respond to fear. Hence, the locals’ shared measure of vulnerability and risks is different from
those living inlands. According to Cannon (2008), these local people can be identified as
insiders (who are affected by disasters), while the GoB as the outsiders (who go to ‘help’). He
emphasises on the gaps between them to claim that the insiders do not usually differentiate
between their normal everyday lives and the shocks. The GoB’s early warning system
addresses risk reduction with a rational and methodological approach that is quantifiable to
evaluate performance. The local communities perceive risk reduction as an opportunity cost
for their assets and livelihoods. Often, these costs are deemed higher than the possible gains
from moving to shelters. Their personal will of choosing to be vulnerable is an example of
innocent disaster, however this innocence leads them to normalise their vulnerability.
Conclusion
Overall, this paper evaluates the early warning system of Bangladesh to find if the
protracted vulnerability of the coastal population is normalised by the key actors within the
mechanism. The query extends to investigate if the vulnerability is due to economic, social or
political prejudice (defined as guilty disaster), or is it a result of the normal livelihood and
social forces (defined as innocent disaster).
Page 13 of 17
The early warning system (EWS) of Bangladesh encompasses several key actors like
the different ministries of the GoB, IFRC, CPP, Meteorological Department, and the local
communities, among others. Using the UNISDR framework for EWS, three officials were
interviewed to supplement the literature review. This essay focuses on the coastal regions of
Bangladesh due to their high vulnerability to natural disasters, mainly cyclones.
The analysis proves the existence of both guilt and innocent disasters in Bangladesh’s
EWS. The institutional structure of the EWS adheres to the priorities set by the HFA 2005,
and is upgraded annually to meet the international standards. Their approach to risk reduction
is encouraged by the elements outlined by the UNISDR framework, and their success is
determined by the reduction of the number of casualties between comparable cyclones.
However, there are gaps in the dissemination, communication and response processes that
result in selective information sharing of early warnings. This can be termed as guilty. Due to
living in areas that are distant from CPP warning dissemination offices, lack of education and
awareness, and physical and medical difficulties, many people are selected to be the victims
by the current EWS.
Although Mahmood (2016) claims that vulnerable people get warnings and are
evacuated to shelters, interviews with Khan (2016) and Rokonuddawla (2016) indicate that
many people are either uninformed or are not ready to evacuate. The fear of loss of harvest
and assets, and the high rates of theft and looting discourage people to evacuate to shelters
despite timely warnings.
Finally, differing perceptions of vulnerability and risk reduction of cyclones lead to a
process of normalisation of crisis by both the government and local communities. The former
views success through reduction of the number of casualties, but in the process obscures the
protracted vulnerability of the coastal population. The latter perceives risk reduction and
evacuation as an opportunity cost of their assets and daily livelihoods. Due to the
commonality of cyclones in these areas and the differing perceptions of crisis, both the
government and local people normalise the protracted vulnerability.
Despite the conclusions mentioned in the preceding paragraphs, there are few
limitations of this study that can be addressed through future research. The analysis here is
based on interviews with officials who are ‘outsiders’ of the EWS. Including the ‘insiders’
like representatives of the local communities will make the study more comprehensive. It will
also be interesting to compare the EWS of Bangladesh to another country with similar
Page 14 of 17
vulnerabilities to understand guilt and innocent factors. Lastly, investigating the power
structure of the EWS in Bangladesh through the lens of accountability measures like the
rights-based approach can be beneficial.
Page 15 of 17
References
Bern, C., J. Sniezek, G. M. Mathbor, Mizzan S. Siddiqi, Caroline Ronsmans, A. M.
Chowdhury, A. E. Choudhury, Kamal Islam, M. Bennish, and E. Noji. 1993. "Risk
factors for mortality in the Bangladesh cyclone of 1991." Bulletin of the World Health
Organization 71(1): 73.
Bisson, Tiffany. 2012. “An Assessment of Cyclone Mitigation and Management Policies of
Bangladesh: A Focus on Early Warning Systems.” Thesis for Graduate Studies
University of Manitoba.
Bradbury, Mark. 1997. "A Review of Oxfam Somalia Programme (1995-1997)."Reports from
Oxfam UK/I, Oxford.
Bradbury, Mark. 1998. "Normalising the Crisis in Africa." Disasters 22(4): 328-338.
Cannon, Terry. 2006. “Vulnerability analysis, livelihoods and disasters.” Risk 21 (2006): 41-49
Cannon, Terry. 2008. "Vulnerability,“innocent” disasters and the imperative of cultural
understanding." Disaster Prevention and Management: An International Journal 17(3):
350-357.
Dasgupta, Susmita, Mainul Huq, Zahirul Huq Khan, Manjur Murshed Zahid Ahmed, Nandan
Mukherjee, Malik Fida Khan, and Kiran Pandey. 2014. "Cyclones in a changing
climate: the case of Bangladesh." Climate and Development 6 (2): 96-110.
Deltares. 2015. “Mobile Services for Flood Early Warning in Bangladesh: Final Report.”
Deltares 2015
Douglas, Mary. 1978. "Cultural Bias.” London: Royal Anthropological Institute .
Douglas, M. & Aaron Wildavsky. 1982. “Risk and Culture.” Berkeley; Los Angeles;
London: University of California Press.
Golnaraghi, Maryam. 2012. Institutional Partnerships in Multi-Hazard Early Warning
Systems: A Compilation of Seven National Good Practices and Guiding Principles.
Springer. doi:10.1007/978-3-642-25373-7.
Griekspoor, Andre, Paul Spiegel, William Aldis, and Paul Harvey. 2004. "The health sector
gap in the southern Africa crisis in 2002/2003." Disasters 28 (4): 388-404.
Habib, Arjumand, Md. Shahidullah, and Dilder Ahmed. 2012. “The Bangladesh Cyclone
Preparedness Program. A Vital Component of the Nation’s Multi-Hazard Early
Warning System.” In , edited by Maryam Golnaraghi, 29–62. Berlin, Heidelberg:
Springer Berlin Heidelberg. doi:10.1007/978-3-642-25373-7_3.
Haque, Ubydul, Masahiro Hashizume, Korine N Kolivras, Hans J Overgaard, Bivash Das,
and Taro Yamamoto. 2011. “Reduced Death Rates from Cyclones in Bangladesh: What
More Needs to Be Done?”Bulletin of the World Health Organization 90 (2). WHO
Press: 150–56. doi:10.2471/blt.11.088302.
Page 16 of 17
Hewitt, K. 1983. Interpretations of Calamities. London: Alan and Unwin.
Howell, P. 2003. “Indigenous Early Warning Indicators of Cyclones: Potentail Application in
Coastal Bangladesh.” Disaster Studies Working Paper 6, Benfield Hazard Research
Centre.
IFRC. 2008. “The Red Cross Red Crescent and the Hyogo Framework for
Action.” International Federation of Red Cross and Red Crescent Societies.
ISDR, UN. 2005. "Hyogo framework for action 2005-2015: building the resilience of nations
and communities to disasters." Final report of the World Conference on Disaster
Reduction (A/CONF. 206/6).
ISDR. 2006a. “ISDR: Platform for the Promotion of Early Warning.” Accessed February 25,
2016. http://www.unisdr.org/2006/ppew/whats-ew/basics-ew.htm.
ISDR. 2006b. “Developing Early Warning Systems: A Checklist.” Third International
Conference on Early Warning: From Concept to Action.
Jegillos, Sanny. 1999. "Fundamentals of Disaster Risk Management: How Are Southeast
Asian Countries Addressing These?." Risk, Sustainable Development and Disasters:
Southern Perspectives: 7-16.
Khan, M., and M Rahman. 2007. “Partnership approach to disaster management in
Bangladesh: A critical policy assessment.” Natural Hazards 41: 359-378.
Khan, Khaled Masud. 2016. Interview about Bangladesh’s Early Warning System.
Interviewed February 26, 2016.
Kreft, S, D Eckstein, L Junghans, C Kerestan, and U Hagen. 2014. “GLOBAL CLIMATE
RISK INDEX 2015. Who Suffers Most from Extreme Weather Events? Weather-
Related Loss Events in 2013 and 1994 to 2013.” Think Tank & Research.
https://germanwatch.org/en/download/10333.pdf.
Lateef, Fatimah. 2009. “Cyclone Nargis and Myanmar: A Wake up Call.” Journal of
Emergencies, Trauma and Shock 2 (2). Medknow: 106. doi:10.4103/0974-2700.50745.
Macrae, Joanna, and Mark Bradbury. 1998. "Aid in the Twilight Zone: A Critical Analysis of
Humanitarian-Development Aid Linkages in."
Mahmud, Khalid. 2016. Interview about Bangladesh’s Early Warning System. Interviewed
February 8, 2016
MWR. (2005). Coastal zone policy. Dhaka: Ministry of Water Resources.
Paul, A., and M Rahman. 2006. “Cyclone mitigation perspectives in the Islands of
Bangladesh: A case of Sandwip and Hatia Islands.” Coastal Management 34, 2: 199-
215.
Page 17 of 17
Pelling, Mark. 1999. "The political ecology of flood hazard in urban
Guyana."Geoforum 30(3): 249-261.
Rokonuddawla, Major General A. R. 2016. Interview about Bangladesh’s Early Warning
System. Interviewed March 29, 2016.
Roy, Chandan, Saroje Kumar Sarkar, Johan Åberg, and Rita Kovordanyi. 2015. “The Current
Cyclone Early Warning System in Bangladesh: Providers’ and Receivers'
Views.” International Journal of Disaster Risk Reduction 12 (June). Elsevier BV: 285–
99. doi:10.1016/j.ijdrr.2015.02.004.
Shultz, J. M., J. Russell, and Z. Espinel. 2005. “Epidemiology of Tropical Cyclones: The
Dynamics of Disaster, Disease, and Development.” Epidemiologic Reviews 27 (1).
Oxford University Press (OUP): 21–35. doi:10.1093/epirev/mxi011.
Stone, R. 2009. “One Year after a Devastating Cyclone, a Bitter Harvest.” Science 324
(5928). American Association for the Advancement of Science (AAAS): 715–715.
doi:10.1126/science.324_715.
Teisberg, Thomas J., and Rodney F. Weiher. 2009. "Background paper on the benefits and
costs of early warning systems for major natural hazards."World Bank, Washington, DC
“TERMINOLOGY ON DRR.” 2007. UNISDR. August 30. Accessed January 15, 2016.
https://www.unisdr.org/we/inform/terminology.
The International Disaster Database. 2016. “Centre for Research on the Epidemiology of
Disasters.” Accessed January 25. http://www.emdat.be/result-country-profile.
UNSC (United Nations Security Council). 1997. Report of the Secretary General on the
Situation in Somalia. February, S/1997/135, New York.