Working Paper 182
September 2009
Climate Change and the Future
Impacts of Storm-Surge Disasters in
Developing Countries
Abstract
As the climate changes during the 21st century, larger cyclonic storm surges and growing populations may
collide in disasters of unprecedented size. As conditions worsen, variations in coastal morphology will magnify
the effects in some areas, while largely insulating others. In this paper, we explore the implications for 84
developing countries and 577 of their cyclone-vulnerable coastal cities with populations greater than 100,000.
Combining the most recent scientific and demographic information, we estimate the future impact of climate
change on storm surges that will strike coastal populations, economies and ecosystems. We focus on the
distribution of heightened impacts, because we believe that greater knowledge of their probable variation will
be useful for local and national planners, as well as international donors. Our results suggest gross inequality
in the heightened impact of future disasters, with the most severe effects limited to a small number of countries
and a small cluster of large cities.
Susmita Dasgupta, Benoit Laplante, Siobhan Murray,
and David Wheeler
Climate Change and the Future Impacts of Storm-Surge Disasters in
Developing Countries
Susmita Dasgupta*
Benoit Laplante
Siobhan Murray
David Wheeler
September 2009
*Authors’ names are in alphabetical order. The authors are, respectively,
Lead Economist, World Bank; Consultants, World Bank; and Senior Fellow,
Center for Global Development. Support for this study was provided by the
Research and Environment Departments of the World Bank, the Center for
Global Development, and the Economics of Adaptation to Climate Change
study, funded by the governments of the United Kingdom, the Netherlands
and Switzerland. Our thanks to Dr. James Baker, Uwe Deichmann, Zahirul
Haque Khan, Kiran Pandey, and Vijaya Ramachandran for useful comments,
and Polly Means for her help with the graphics. All remaining errors are our
own.
This paper was made possible by financial support from the UK Department
for International Development.
Center for Global Development
1800 Massachusetts Ave., NW
Washington, DC 20036
202.416.4000
(f) 202.416.4050
The Center for Global Development is an independent, nonprofit policy
research organization dedicated to reducing global poverty and inequality
and to making globalization work for the poor. Use and dissemination of
this Working Paper is encouraged; however, reproduced copies may not be
used for commercial purposes. Further usage is permitted under the terms of
the Creative Commons License.
The views expressed in this paper are those of the author and should not
be attributed to the board of directors or funders of the Center for Global
Development.
1. Introduction
Large tropical cyclones create storm surges that can strike crowded coastal regions
with devastating force.
1
During the past 200 years, 2.6 million people may have drowned
during surge events (Nicholls 2003). These disasters have continued to inflict heavy
losses on the people of developing countries. Cyclone Sidr struck Bangladesh in
November 2007, killing over 3,000 people, injuring over 50,000, damaging or destroying
over 1.5 million homes, and affecting the livelihoods of over 7 million people (UN 2007;
BDMIC 2007). Cyclone Nargis struck Myanmar’s Irrawaddy delta in May 2008,
creating the worst natural disaster in the country’s recorded history. It killed over 80,000
people and affected the livelihoods of over 7 million (UN 2009).
The scientific evidence indicates that climate change will intensify storm surges for
two reasons. First, they will be elevated by a rising sea level as thermal expansion and
ice cap melting continue. The most recent evidence suggests that sea-level rise could
reach 1 meter or more during this century (Dasgupta, et al. 2009; Rahmstorf 2007).
Second, a warmer ocean is likely to intensify cyclone activity and heighten storm surges.
2
As storm surges increase, they will create more damaging flood conditions in coastal
zones and adjoining low-lying areas. The destructive impact will generally be greater
when the surges are accompanied by strong winds and large onshore waves.
Larger storm surges threaten greater future destruction because they will move
further inland, threatening larger areas than in the past. In addition, both natural increase
and internal migration are increasing the populations of coastal areas in many developing
1
Storm surge refers to the temporary increase, at a particular locality, in the height of the sea due to
extreme meteorological conditions: low atmospheric pressure and/or strong winds (IPCC AR4 2007).
2
A sea-surface temperature of 28
o
C is considered an important threshold for the development of major
hurricanes of categories 3, 4 and 5 (Michaels et al 2005; Knutson and Tuleya 2004).
2
countries. Table 1 shows that coastal population shares increased in all developing
regions from 1980 to 2000. Population growth is particularly strong in coastal urban
areas, whose growth also reflects continued rural-urban migration in many developing
countries.
During the 21
st
century, rising storm surges and growing populations threaten to
collide in disasters of unprecedented size. And, as average effects increase, variations in
coastal morphology may magnify the effects in some areas, while largely insulating
others. In this paper, we explore the implications for 84 developing countries, along with
577 of their cyclone-vulnerable coastal cities with populations greater than 100,000.
Combining the most recent scientific and demographic information, we estimate the
future impact of climate change on storm surges that will strike coastal populations,
economies and ecosystems. We focus on the distribution of heightened impacts, because
we believe that greater knowledge of their probable variation will be useful for local and
national planners, as well as international donors. In addition, we believe that realistic
projections of the scale of these disasters will inform the current debate about the
appropriate timing and strength of carbon emissions mitigation.
The remainder of the paper is organized as follows. Section 2 reviews recent
scientific evidence on global warming and tropical cyclone intensity, and motivates the
paper. Section 3 describes our research strategy and data sources, while Section 4
describes our methodology. In Sections 5 and 6, we present our results for countries and
coastal cities. Section 7 summarizes and concludes the paper.
3
2. Global Warming, Tropical Cyclone Intensity and Disaster Preparedness
Some recent scientific studies suggest that observed increases in the frequency and
intensity of tropical cyclones in the last 35 years can be attributed in part to global
climate change (Emanuel 2005; Webster et al. 2005; Bengtsson et al. 2006). Others have
challenged this conclusion, citing problems with data reliability, regional variability, and
appropriate measurement of sea-surface temperature and other climate variables (e.g.,
Landsea et al. 2006). Although the science is not yet conclusive (IWTC 2006, Pielke et
al. 2005), the World Meteorological Organization (2006) has recently noted that “[i]f the
projected rise in sea level due to global warming occurs, then the vulnerability to tropical
cyclone storm surge flooding would increase” and “[i]t is likely that some increase in
tropical cyclone peak wind-speed and rainfall will occur if the climate continues to warm.
Model studies and theory project a 3-5% increase in wind speed per degree Celsius
increase of tropical sea surface temperatures.”
IPCC (2007) cites a trend since the mid-1970s toward longer duration and greater
intensity of storms, and a strong correlation with the upward trend in tropical sea surface
temperatures. In addition, it notes that hurricanes/cyclones are occurring in places where
they have never been observed before.
3
Overall, using a range of model projections, the
report asserts a probability greater than 66% that continued sea-surface warming will lead
to tropical cyclones that are more intense, with higher peak wind speeds and heavier
precipitation (IPCC 2007; see also Woodworth and Blackman 2004; Woth et al. 2006;
and Emanuel et al. 2008).
4
3
The first recorded tropical cyclone (Catarina) in the South Atlantic occurred in March 2004, off the coast
of Brazil. For further information on cyclone Catarina and storm surge risk, see UNISDR (2009).
4
Cyclones get their power from rising moisture which releases heat during condensation. As a result,
cyclones depend on warm sea temperatures and the difference between temperatures at the ocean surface
4
These consensus projections from the global scientific community point to the need
for greater disaster preparedness in countries that are vulnerable to storm surges. Some
adaptation has already occurred, and many lives have been saved by improvements in
disaster forecasting, evacuation and emergency shelter procedures (Shultz et al. 2005,
Keim 2006). At the same time, as the recent disasters in Bangladesh and Myanmar have
demonstrated, storm-surge losses remain huge in many areas. Such losses could be
reduced by allocating more resources to increased disaster preparedness, especially given
the likelihood that storms and storm surges will intensify. However, setting a new course
requires a better understanding of expected changes in storm surge patterns.
3. Research Strategy and Data Sources
Previous research on storm-surge impacts has been confined to relatively limited
sets of impacts
5
and locations.
6
In this paper, we broaden the assessment to 84 coastal
countries and 577 coastal cities in five developing regions.
7
We consider the potential
impact of a storm surge that is large (1 in 100 years) by contemporary standards, and then
compare it with a more intense impact later in the century. In modeling future
conditions, we take account of sea-level rise, geological uplift and subsidence along the
world’s coastlines. Our analysis includes exposure indicators for affected territory,
population, economic activity (GDP), agricultural lands, wetlands, major cities and other
urban areas. As far as we know, this is the first such exercise for developing countries.
and in the upper atmosphere. If global warming increases temperatures at the earth’s surface but not the
upper atmosphere, it is likely to produce tropical cyclones with more power (Emmanuel et al. 2008).
5
Nicholls et al. (2008) assess the impacts of climate extremes on port cities of the world.
6
The impacts of storm surges have been assessed for Copenhagen (Hallegatte et al. 2008); Southern
Australia (McInnes et al. 2008); and the Irish Sea (Wang et al. 2008).
7
We employ the five World Bank regions: East Asia & Pacific, Middle East & North Africa, Latin
America & Caribbean, South Asia, and Sub-Saharan Africa.
5
We have employed geographic information system (GIS) software to overlay the
critical impact elements (land, population, agriculture, urban extent, wetlands, GDP and
city locations) with the inundation zones projected for two cases: a current 1-in-100-year
storm surge and a 10% intensification over the next 100 years. We have used the best
available spatially-disaggregated data sets from various public sources, including the
National Aeronautics and Space Administration (NASA), US Geological Survey
(USGS), European Space Agency (ESA), Dynamic Interactive Vulnerability Assessment
(DIVA), World Wildlife Fund (WWF), Center for International Earth Science
Information Network (CIESIN), and World Bank. Table 2 summarizes our data sources
for assessments of inundation zones and impacts.
At the outset, we acknowledge several limitations in the analysis. First, we do not
assess the relative likelihoods of alternative storm surge scenarios. We follow Nicholls,
et al. (2007) in assuming an homogeneous future increase of 10% in extreme water levels
during tropical storms. In all likelihood increases will vary across regions, and better
area-specific modeling will be needed to improve regional forecasts.
8
Second, among the
84 developing countries included in this analysis, we restrict our study to coastal
segments where historical storm surges have been documented. Third, the absence of a
global database on shoreline protection prevents us from incorporating the effects of
existing protection measures (e.g., sea dikes) on exposure estimates. Fourth, we have not
been able to include small island states because the best available satellite system cannot
accurately measure ground elevation over small areas.
9
8
As noted by Emanuel et al. (2008).
9
Our SRTM data source is described on page 8.
6
In our country-level study, we assess the impacts of storm surges using existing
populations, socioeconomic conditions and patterns of land use, rather than attempting to
predict their future states. This approach is conservative, since human activity is
generally increasing more rapidly in coastal areas (Table 1). On the other hand, we do
not attempt to estimate the countervailing effects of local adaptation measures related to
infrastructure (e.g., coastal embankments) and coastal-zone management (e.g., land-use
planning, regulations, relocation). Our city-level study goes further by incorporating the
UN’s medium-term population projections for the 21
st
century. This enables us to
provide better comparative estimates of potential impacts across cities within countries,
as well as across countries.
4. Methodology
Our analysis involves a multi-step procedure. First, we employ a base
hydrologically-conditioned elevation data set to identify inundation zones and subject
them to alternative storm-surge (wave height) scenarios. Second, we construct a country
surface for each vulnerability indicator (population, GDP, urban extent, agricultural
extent,
10
wetlands, major cities). Third, we overlay these indicator surfaces with the
inundation zone layer. Then we determine the spatial exposure of each vulnerability
indicator under storm-surge conditions. More detailed descriptions of these steps are as
follows.
10
We use the Globcover database for agriculture, which includes 3 land use indicators. The first indicator
includes areas in which most of the coverage is rainfed/irrigated/post-flooding cropland. The second
includes areas for which 50-70% is mosaic cropland and the rest is grassland, shrubland and forest. The
third includes areas for which 20-50% is mosaic cropland and the rest is grassland, shrubland, and forest.
For the purpose of identifying impacted agricultural extent, we have retained only the agricultural land
identified as rainfed/irrigated/post-flooding cropland (the first indicator above). As a result, our calculations
underestimate the impacts on agricultural extent.
7
(i) For elevation, we use a recently-released hydrologically-conditioned version of
SRTM data, part of the HydroSHEDS dataset. We have downloaded all 5ºx5º coastal
tiles of 90m SRTM data from In this case, conditioning the SRTM data involves steps that alter elevation values to
produce a surface that drains to the coast (except in cases of known internal drainages).
These steps include filtering, lowering of stream courses and adjacent pixels, and carving
out barriers to stream flow.
(ii) In the calculation of storm surges (wave heights or extreme sea levels), we follow the
method outlined by Nicholls (2008), where storm surges are calculated as follows:
Current storm surge = S100
Future storm surge = S100 + SLR + (UPLIFT * 100 yr ) / 1000 + SUB + S100 * x
where:
S100
= 1-in-100-year surge height (m)
SLR
= sea-level rise (1 m)
11
UPLIFT = continental uplift/subsidence in mm/yr
SUB
= 0.5 m (applies to deltas only)
x
= 0.1, or increase of 10%, applied only in coastal areas currently prone to
cyclones/hurricanes.
We calculate surges using data associated with the coastlines.
We extract vector coastline masks from SRTM version 2, and download coastline
information from the DIVA GIS database. We use the following attributes in this
analysis:
1.
S100: 1-in-100-year surge height, based on tidal levels, barometric pressures,
wind speeds, seabed slopes and storm surge levels from monitoring stations;
2.
DELTAID: coastline segments associated with river deltas;
3.
UPLIFT: estimates of continental uplift/subsidence in mm/yr from Peltier
(2000), including a measure of natural subsidence (2 mm/yr) for deltas.
11
Nicholls (2008) assumed a SLR of 0.5 meters.
8
(iii) We compare surges (wave heights) associated with current and future storms
with the elevation values of inland pixels with respect to a coastline, to delineate potential
inundation areas.
Each inland pixel could be associated with the nearest coastline segment in a
straight-line distance. However, in order to better capture the movement of water inland,
we use hydrological drainage basins. We apply the wave height calculated for the
coastline segment closest to the basin outlet to inland areas within that basin.
As a wave moves inland its height is diminished. The rate of decay depends largely
on terrain and surface features, as well as factors specific to the storm generating the
wave. In a case study on storm surges, Nicholls (2006) uses a distance decay factor of
0.2-0.4 m per 1 km that can be applied to wave heights in relatively flat coastal plains
.
For this analysis, we use an intermediate value (0.3 m per 1 km distance from the
coastline) to estimate the wave height for each inland cell.
We delineate surge zones by comparing projected wave heights with SRTM values
in each cell. A cell is part of the surge zone if its elevation value is less than the
projected wave.
(iv) Following McGranahan et al. (2007), we delineate low-elevation coastal zones
using inland pixels with less than 10m elevation near coastlines.
Our processing uses 5º x 5º tiles, employing aml (ArcInfo Macro Language) for
automation.
We should note that our estimates are conservative because they do not take future
shoreline erosion into account. As we noted previously, the absence of a global database
on shoreline protection has prevented us from modeling likely changes in shorelines
9
associated with a 1m sea-level rise. Even a 1m rise in sea level will change shorelines
considerably in many coastal segments, if shorelines are not protected (Dasgupta et al.
2009). We present illustrative cases for Bangladesh and Viet Nam in Figure 2. Coastal
morphology will change with receding shorelines, and potential inundation areas for
storm surges will be determined by the characteristics of the changed coastlines. To
improve coastal security, future research and adaptation planning should consider such
likely shoreline changes.
(v) Calculating exposure indicators: We overlay our delineated inundation zones
with our indicators for land area, GDP, population, urban extent, agriculture extent,
wetlands and locations of cities with more than 100,000 inhabitants in 2000.
12
We have
collected exposure surface data from various public sources. Our horizontal datum is the
World Geodetic System 1984 (WGS 1984). For area calculation, we create grids
representing cell areas in square kilometers at different resolutions, using length of a
degree of latitude and longitude at cell center.
Employing the appropriate units (e.g. GDP in millions of dollars, individuals for
population), we calculate total exposure by summing over exposed units in inundation
zones. We measure exposure for land surface, urban extent, agriculture extent and
wetlands in square kilometers.
(vi) Adjusting absolute exposure indicators: For exposure indicators such as land
area, population and GDP, which have measured country coastal zone totals available, we
adjust the exposed value using the following formula:
12
The delineated surge zones and coastal zones are at a resolution of 3 arc seconds (approximately 90 m).
The resolution of indicator datasets ranges from 9 arc seconds to 30 arc seconds. Because of this difference
in resolution, a surge zone area may occupy only a portion of a single cell in an indicator dataset. In this
case, the surge zone is allocated to the appropriate proportion of the indicator cell value.
10
cal
cal
mea
adj
V
CT
CT
V
⋅
=
where
V
adj
= Adjusted exposed value;
V
cal
= Exposed value calculated from exposure grid surfaces;
CT
mea
= Country coastal zone total obtained from statistics;
CT
cal
= Country coastal zone total calculated from exposure grid surface.
5. Country Results
We summarize our results for 84 developing countries in Table 3. It indicates that
approximately 19.5% (391,812 km
2
) of their combined coastal territory is vulnerable to
inundation from a 1-in-100-year storm surge (by current reference standards). A 10%
future intensification increases the potential inundation zone to 25.7% (517,255 km
2
) of
coastal territory, taking into account sea-level rise. This translates to potential inundation
for an additional 52 million people; 29,164 km
2
of agricultural area; 14,991 km
2
of urban
area; 9% of coastal GDP and 7% of wetlands.
These exposures are far from uniformly distributed across the regions and countries
of the developing world. As Figure 1 shows, Latin America and the Caribbean have the
largest percentage increase in storm-surge zone area (35.2%), but the coastal population
exposures are largest for the Middle East and North Africa (56.2%)
,
while coastal GDP
exposures are most severe in East Asia (51.2%). Similar disparities characterize the
exposures of urban areas, agricultural lands and wetlands.
Because GDP per capita is
generally above-average for coastal populations and cities, we estimate that storm surge
intensification would cause additional GDP losses (above the current 1-in-100-year
reference standard) of $84.9 billion in East Asia and the Pacific, $12.7 billion in the
11
Middle East and North Africa, $8.4 billion in South Asia, $14.4 billion in Latin America
and the Caribbean, and $1.8 billion in Sub-Saharan Africa.
We also estimate exposures for individual countries and territories. Table 4
summarizes our results for each indicator by presenting the top-10 impacted countries (as
a percentage of their own coastal values). Our results suggest that numerous low-income
countries are susceptible to very significant damage. For land area, the most vulnerable
low-income countries are Namibia, Guinea, El Salvador and Yemen, with more than 50%
of their coastal areas at risk. For exposed populations, the top-5 low-income
countries/territories are Djibouti, Yemen, Togo, El Salvador, and Mozambique. More
than 50% of coastal urban areas lie within the potential impact zones in Guyana, Djibouti,
Togo, Yemen, Mozambique, Tanzania, Cote d’Ivoire, Equatorial Guinea and Morocco.
Coastal agriculture would be significantly affected in Guyana, Nigeria, North Korea, El
Salvador, Ghana, Togo and Equatorial Guinea. Our estimates indicate that areas prone to
storm surge in Mozambique, Togo, Morocco, Philippines, Yemen, Djibouti, El Salvador
and Ghana account for more than 50% of the GDP generated in their coastal regions.
Finally, nearly 100% of the coastal wetlands in El Salvador and more than 60% of the
wetlands in Namibia, Ecuador, Tunisia, Guinea, Yemen and Pakistan will be subject to
inundation risk. For the majority of indicators used in this research, we observe the most
consistently-severe exposure risks for El Salvador, Yemen, Djibouti, Mozambique and
Togo.
12
6. City Results
6.1 Coastal Area Vulnerability
In this section we consider two measures of coastal urban exposure. The first,
summarized in Table 5, lists cities in each developing region whose coastal areas will be
most affected by future increases in storm surges. We calculate vulnerability in three
steps. First, we rank cities in each region by percent increase in the future inundation
area relative to the current inundation area. To weight for current vulnerability, we rank
cities in each region by percent of coastal area in the current inundation zone. Then we
compute the average for the two ranks and re-order the cities by their average ranks.
Table 5 includes the 10 highest-ranking cities in each region, using future inundation
increase weighted by current vulnerability. To illustrate, Nacala, Mozambique has the
highest future vulnerability in Sub-Saharan Africa. In the 21
st
century, 25% of its coastal
area will be added to its inundation zone (Pct 2). This is a 50% increase in its current
inundation zone, which is already 50% of its coastal area (Pct 1). Using the same
calculations, we identify the top-ranked cities in the other four regions as Rach Gia, Viet
Nam; Georgetown, Guyana; Kenitra, Morocco; and Cox’s Bazar, Bangladesh. These
cities join the other top-10 cities as potentially-deadly locales, since storm water drainage
infrastructure is often outdated and inadequate in low-income urban centers.
13
The risks
may be particularly severe in poor neighborhoods and slums, where infrastructure is often
nonexistent or poorly designed and ill-maintained. Within regions, exposures are clearly
far from balanced across countries. In each region, at least half of the top-10 cities are in
only two countries: Mozambique (3) and Cote d’Ivoire (2) in Sub-Saharan Africa;
13
For port cities vulnerable to storm surges, see Nicholls et al. (2007).
13
Indonesia (3) and Viet Nam (3) in East Asia ; Brazil (3) and Mexico (2) in Latin
America; Morocco (4) and Tunisia (3) in the Middle East and North Africa; and
Bangladesh (4) and India (4) in South Asia.
6.2 Population Vulnerability
In an alternative approach, we compare cities by estimating the vulnerability of
their populations to intensified storm surges in the 21
st
century. We consider the
combined effects of projected population change, sea-level rise and storm intensification
on the distribution of exposures by the end of the century. We use the UN’s medium
population projections for 2100, as reported by IIASA (2009), and conservatively assume
that all coastal cities in each country retain their current share of the national
population.
14
In addition, we assume that coastal cities’ populations are uniformly-
distributed across their coastal and non-coastal areas. From the work reported in Section
6.1 above, we draw the percent of coastal areas in inundation zones now, and in 2100
after a 1m sea-level rise and 10% increase in the intensity of a 1-in-100-year storm.
Combining the area and demographic information, we estimate populations in the current
and future inundation zones, and the implied increase in affected population. Table 6
displays the 25 cities with the largest population exposures, expressed as changes in
affected populations and cumulative percents of the total change for all cities.
The most striking feature of our results is the extreme concentration of effects in a
handful of cities. Over 25% of the increase in developing-country urban population
affected by future storm surges is in only three cities:
15
Manila (3.4 million), Alexandria
14
As Table 1 shows, coastal cities increased their percent share of national populations during the period
1980-2000. However, we have no credible way to extrapolate this trend for the next 100 years.
15
Many large coastal cities have only small increases (or even decreases), because projected populations
and coastal inundation zones have countervailing trends during the 21
st
century. All future coastal
14
(2.7 million) and Lagos (2.1) million. The top 10 cities account for 53 % of total
exposures, and the top 25 for 72%. The 552 other coastal cities in our dataset account for
only 28% of the total. Of the top 25, 10 are in Sub-Saharan Africa (8 in coastal West
Africa alone), 3 in the Middle East and North Africa, 7 in Southeast Asia, 4 in South
Asia, and 1 in South America. We should emphasize that our results are not closely tied
to the current distribution of coastal city populations. As Table 6 shows, many of the
cities with top-25 changes in vulnerable populations are not among the world’s most
populous urban areas at present. Their future top-25 status stems from two factors:
future urban growth, and coastal characteristics that make them particularly vulnerable to
greater storm surges.
7. Conclusions
In this paper, we have assessed the vulnerability of coastal areas in developing
countries to larger storm surges associated with global warming and a 1m sea-level rise.
After identifying future inundation zones, we have overlaid them with indicators for
coastal populations, settlements, economic activity and wetlands. Our results indicate
large effects that are much more concentrated in some regions, countries and cities than
others. We have also incorporated population projections for the 21
st
century, and
computed the exposures of coastal urban populations as conditions worsen. Our results
suggest a huge asymmetry in the burden of sea-level rise and storm intensification, with
inundation zones increase at least somewhat with a 1m sea-level rise and a 10% increase in storm intensity.
However, the UN projects rapidly-declining fertility and significant population loss for many countries by
2100. In our methodology, their cities follow suit because we assume fixed city/country population ratios.
Shanghai provides a useful illustration of these countervailing forces. Our spatial analysis indicates that
Shanghai’s inundation zone will increase from 15.7% of its coastal area in 2000 to 25.8% in 2100.
However, the demographic projection indicates that Shanghai’s coastal-zone population will decline from
13.2 million in 2000 to 7.5 million in 2100. These two factors combine to produce a small decrease
between 2000 and 2100 in the population affected by severe storm-surge conditions.
15
only 3 of 577 cities accounting for 25% of the future coastal population exposure and 10
cities for 53% of the future exposure. Our results suggest that the residents of a small
number of developing-country cities will bear the additional brunt of heightened storm
surges, while many other coastal cities will experience little change in population
exposure. In light of the huge asymmetries in our country- and city-level results, we
believe that careful targeting of international assistance will be essential for the effective
and equitable allocation of resources for coastal protection and disaster prevention. In
addition, the large magnitudes of potential impacts on people, economies and ecosystems
provide strong evidence in support of rapid action to reduce global warming by
mitigating greenhouse gas emissions.
16
References
Bangladesh Disaster Management Information Centre (BDMIC). 2007. Cyclonic Storm
“SIDR”. Nov. 26.
Bengtsson, L., K.I. Hodges, and E. Roeckner. 2006. Storm tracks and climate change.
Journal of Climate, 19, 3518-3543.
Dasgupta S., B. Laplante, C. Meisner, D. Wheeler and J. Yan. 2009. The impact of sea-
level rise on developing countries: A comparative analysis. Forthcoming in Climatic
Change.
Emanuel, K., R. Sundararajan, and J. William. 2008. Hurricanes and global warming:
Results from downscaling IPCC AR4 simulations. Journal of Climate, 89, 347-367.
Emanuel, K. 2005. Increasing destructiveness of tropical cyclones over the past 30 years.
Nature, 436, 686-688.
Hallegatte, S., N. Patmore, O. Maestre, P. Dumas, J. C. Morlot, C. Herweijer, and R. M.
Wood. 2008. Assessing Climate Change Impacts, Sea Level Rise, and Storm Surge Risk
in Port Cities: A Case Study of Copenhagen. OECD Environment Directorate,
Environment Working Papers No. 3, Paris.
IIASA (International Institute for Applied Systems Analysis). 2009. CGI Scenario
Database.
International Workshop on Tropical Cyclones (IWTC). 2006. Statement on Tropical
Cyclones and Climate Change. November, 2006, 13 pp.
IPCC. 2007. Climate Change 2007: The Physical Science Basis, Summary for
Policymakers. The Intergovernmental Panel on Climate Change.
Keim, M.E. 2006. Cyclones, tsunamis and human health. Oceanography, 19(2), 40-49.
Knutson, T.R., and R.E. Tuleya. 2004. Impact of CO2-induced warming on simulated
hurricane intensity and precipitation sensitivity to the choice of climate model and
convective parameterization. Journal of Climate, 17, 3477-3495.
Landsea, C.W., B.A. Harper, K. Hoarau, and J.A. Knaff. 2006. Climate change: Can we
detect trends in extreme tropical cyclones? Science, 313(5786), 452-454.
17
McGranahan, G., D. Balk and B. Anderson. 2007. The rising tide: assessing the risks of
climate change and human settlements in low elevation coastal zones. Environment &
Urbanization 19(1): 17-37 (2007). International Institute for Environment and
Development (IIED).
McInnes, K.L., G.D. Hubbert, I. Macadam, and J.G. O’Grady. 2008. Assessing the
impact of climate change on storm surges in Southern Australia. mimeo.
Michaels, P.J., P. C. Knappenberger, and R.E. Davis. 2005. Sea-Surface Temperatures
and Tropical cyclones: Breaking the Paradigm. Presented at the 15
th
Conference of
Applied Climatology.
Nicholls, R.J. 2003. An Expert Assessment of Storm Surge “Hotspots”. Final Report
(Draft Version) to Center for Hazards and Risk Research, Lamont-Dohert Observatory,
Columbia University.
Nicholls, R.J. 2006. Storm Surges in Coastal Areas. Natural Disaster Hot Spots Case
Studies. Chapter 3. Margaret Arnold et al. 2006. The World Bank.
Nicholls, R.J., S. Hanson, C. Herweijer, N. Patmore, S. Hallegatte, J. Corfee-Morlot, J.
Chateau, R. Muir-Wood. 2007. Ranking Port Cities with High Exposure and
Vulnerability to Climate Extremes. OECD Environment Directorate, Environment
Working Papers No. 1.
Pielke, R.A., C. Landsea, M. Mayfield, J. Laver, and R. Pasch. 2005. Hurricanes and
global warming. Bulletin of American Meteorological Society, November 2005, 1571-
1575.
Rahmstorf, Stefan. 2007. A semi-empirical approach to projecting future sea-level rise.
Science, 315, 368-370.
Shultz, J.M., J. Russell, and Z. Espinel. 2005. Epidemiology of tropical cyclones: The
dynamics of disaster, disease, and development. Epidemiologic Reviews, 27, 21-35.
United Nations. 2007. Emergency Report, Sit Rep # 14. SIDR Cyclone Bangladesh
2007.
United Nations. 2009. Post-Nargis Joint Assessment.
United Nations International Strategy for Disaster Reduction Secretariat (UNISDR).
2009. Global assessment report on disaster risk reduction.
18
Wang, S., R. McGrath, J. Hanafin, P. Lynch, T. Semmler, and P. Nolan. 2008. The
impact of climate change on storm surges over Irish waters. Ocean Modeling, 25(1-2),
83-94.
Webster, P.J., G.J. Holland, J.A. Curry, and H-R. Chang. 2005. Changes in tropical
cyclone number, duration and intensity in a warming environment. Science, 309, 1844-
1846.
WMO (World Meteorological Organization). 2006. 6th International Workshop on
Tropical Cyclones of the World Meteorological Organization. November.
Woodworth, P.L. and D.L. Blackman. 2004. Evidence for systematic changes in extreme
high waters since the mid-1970s. Journal of Climate, 17(6), 1190-1197.
Woth, K., R. Weisse, and H. von Storch. 2006. Climate change and North Sea storm
surge extremes: An ensemble study of storm surge extremes expected in a changed
climate projected by four different regional climate models. Ocean Dynamics, 56(1), 3-
15.
19
Table 1
Percent of National Population in Coastal Cities,
1980 – 2000
Data Source: CIESIN, Global Rural Urban Mapping Project (GRUMP), 2005:
Population in coastal urban zone, defined as elevation < 10 m
Table 2
Summary of Data Sources
Dimension Dataset
Name
Unit
Resolution Source(s)
Coastline
SRTM v2 Surface
Water Body Data
NASA
Elevation
Hydrosheds
conditioned SRTM
90m DEM
Km
2
.
Watersheds Hydrosheds
Drainage Basins
Km
2
.
Coastline
Attributes
DIVA GIS database
Population
GRUMP 2005 (pre-
release) gridded
population dataset
Population
counts
1km CIESIN
GDP
2005 GDP Surface
Million USD
1km
World Bank , 2008
Agricultural
Land
Globcover 2.1
Km
2
Urban areas Grump, revised
Km
2
1km
CIESIN
Wetlands GLWD-3
Km
2
Cities
City Polygons with
Population Time
Series
Urban Risk Index*,
Henrike Brecht, 2007
*Urban extents from GRUMP (alpha) (
) joined with World Cities Data
(J. Vernon Henderson 2002).
World Bank Region
1980
1990
2000
Sub-Saharan Africa
7.19
9.12
11.98
East Asia and Pacific
7.09
8.55
9.36
Latin American and Caribbean
15.58
16.61
17.53
Middle East and North Africa
7.64
8.00
8.57
South Asia
4.19
4.80
5.55
20
Table 3
Impacts of Intensification of Storm Surges
Across Indicators at the Global Level
Current Storm
Surge
With
Intensification
Coastal Land Area (Total= 2,012,753 km
2
)
Exposed area
391,812
517,255
% of total coastal area
19.5
25.7
Coastal Population (Total= 707,891,627)
Exposed population
122,066,082
174,073,563
% of total coastal population
17.2
24.6
Coastal GDP (Total =1,375,030 million USD)
Exposed GDP (USD)
268,685
390,794
% of total coastal GDP
19.5
28.4
Coastal Urban area (Total=206,254 km
2
)
Exposed area
40,189
55,180
% of total coastal urban area
19.5
26.8
Coastal Agricultural area (Total = 505,265 km
2
)
Exposed area
59,336
88,500
% of total coastal agricultural area
11.7
17.5
Coastal Wetlands Area (Total = 663,930 km
2
)
Exposed area
152,767
198,508
% of total coastal wetlands area
23.0
29.9
21
Table 4
Top 10 Countries at Risk From Intensification of Storm Surges
*
(Percentage impacts in coastal zones in parentheses)
Rank Coastal
Land
Area
Coastal Population
Coastal GDP
Coastal Agricultural
Land
Coastal Urban
Areas
Coastal
Wetlands
1 Kuwait
(81.1)
Bahamas
(73.0)
Bahamas
(65.7)
Guyana
(100.0)
Bahamas
(94.1)
El Salvador
(100.0)
2 Korea
(61.7)
Kuwait
(70.0)
Kuwait
(65.3)
UAE
(100.0)
Guyana
(66.4)
Belize
(100.0)
3 Namibia
(60.2)
Djibouti
(60.1)
Belize
(61.1)
Nigeria
(100.0)
Djibouti
(60.4)
Kuwait
(95.8)
4 Guinea
(58.6)
UAE
(60.0)
UAE
(58.1)
Qatar
(85.7)
UAE
(60.2)
Taiwan
(95.2)
5 El
Salvador
(55.3)
Belize
(56.2)
Mozambique
(55.0)
Korea
(66.8)
Togo
(59.8)
Namibia
(81.6)
6 Chile
(54.7)
Yemen
(55.7)
Togo
(54.5)
El Salvador
(66.7)
Kuwait
(56.4)
Korea
(78.8)
7 Bahamas
(54.7)
Togo
(54.2)
PuertoRico
(52.7)
Ghana
(66.7)
Yemen
(55.4)
Qatar
(75.0)
8 PuertoRico
(51.8)
PuertoRico
(53.8)
Morocco
(52.6)
DPR Korea
(58.3)
Mozambique
(55.1)
Bahamas
(71.4)
9 Yemen
(50.2)
El Salvador
(53.0)
Philippines
(52.3)
Togo
(50.0)
Tanzania
(53.4)
Ecuador
(67.3)
10 Oman
(50.0)
Mozambique
(51.7)
Yemen
(52.0)
Equatorial Guinea
(50.0)
Cote d'Ivoire
(53.2)
Tunisia
(63.5)
22
Table 5: Future Storm-Surge Impacts on Coastal Areas: Regional Top-10 Cities
Pct 1 = Current Inundation Zone as Percent of Coastal Area
Pct 2 = Future Increase in Inundation Zone as Percent of Coastal Area
Ratio = 100 x [Pct 2 / Pct 1]
Region
Subregion
Country
City
Rank
Ratio
Pct 1
Pct 2
AFR Southern
Africa
Mozambique Nacala
1 50
50 25
AFR
Coastal West Africa
Benin
Cotonou
2.5
62
38
24
AFR Southern
Africa
Mozambique Quelimane
2.5 34
56 19
AFR Southern
Africa
Mozambique Beira
4 33
51 17
AFR
Coastal West Africa
Nigeria
Warri
5
50
33
17
AFR
Coastal West Africa
Côte d'Ivoire
San-Pedro
7
44
35
16
AFR
Coastal West Africa
Côte d'Ivoire
Abidjan
7
107
27
29
AFR
Coastal West Africa
Gambia
Bathurst
7
21
60
12
AFR
Southern Africa
South Africa George
9
17
67 11
AFR
Coastal West Africa
Liberia
Monrovia
10
32
38
12
EAP
Southeast Asia
Viet Nam Rach
Gia
1
46
60
27
EAP Southeast
Asia
Indonesia Tegal
2
60
48
29
EAP
Northeast Asia
Korea, Rep
Ansan
3
27
70
19
EAP Southeast
Asia
Indonesia Cirebon
4
69
35
24
EAP Southeast
Asia
Philippines Butuan
5
63
38
24
EAP China
China
Dandong
6 39
51 20
EAP
Southeast Asia
Viet Nam
Nha Trang
7
27
67
18
EAP
Southeast Asia
Viet Nam
Hue
8
26
68
18
EAP Southeast
Asia
Indonesia
Pemalang 9.5
86
31
27
EAP Southeast
Asia
Philippines
Cotabato 9.5
22
73
16
LCR
Northern South America
Guyana
Georgetown
1
37
56
21
LCR
Southern South America
Argentina
La Plata
2
93
33
31
LCR
Central America
Mexico
Ciudad del Carmen
3.5
24
73
18
LCR
Central America
Mexico
Acapulco (de Juarez)
3.5
45
44
20
LCR
Northern South America
Brazil
Aracaju
5
27
51
14
LCR
Caribbean Islands
Dominican Rep
La Romana
6
50
35
17
LCR
Northern South America
Brazil
Rio Grande
7
26
48
13
LCR
Northern South America
Brazil
Maceio
8
35
40
14
LCR
Northern South America
Venezuela
Barcelona
9.5
55
29
16
LCR
Andean South America
Ecuador
Esmeraldas
9.5
24
49
12
MNA North
Africa
Morocco
Kenitra
1 36 48 18
MNA East
Africa
Djibouti
Djibouti
2 20 50 10
MNA North
Africa
Morocco
Tetouan
3.5 35 37 13
MNA North
Africa
Tunisia
Binzart
3.5 40 33 13
MNA North
Africa
Tunisia
Susah
6 62 32 20
MNA North
Africa
Morocco
Mohammedia
6 19 51 10
MNA North
Africa
Morocco
Rabat
6 16 61 10
MNA
North Africa
Egypt
Bur Sa'id (Port Said)
8.5
13
75
9
MNA North
Africa
Tunisia
Tunis
8.5 25 46 11
SAR
Southern Asia
Bangladesh
Cox's Bazar
1
42
47
20
SAR Southern
Asia
Bangladesh Khulna
2 88
35 31
23
Region
Subregion
Country
City
Rank
Ratio
Pct 1
Pct 2
SAR Southern
Asia
Bangladesh Bakerganj
3 28
70 20
SAR Western
Asia
Pakistan
Karachi
4 30
44 13
SAR Southern
Asia
India
Jamnagar
5 32
43 13
SAR
Southern Asia
India
Vadodara (Baroda)
6
40
36
14
SAR
Southern Asia
Sri Lanka
Moratuwa
7
74
21
16
SAR Southern
Asia
India
Thane
8.5 19
43 8
SAR Southern
Asia
Bangladesh Chandpur 8.5
50
24
12
SAR Southern
Asia
India
Bhavnagar
10 14
58 8
24
Table 6
Top-25 City Population Impacts
Rank
Country
City
Change in
Affected
Population
Cumulative
%
Global
City
Population
Rank
2000
1 Philippines
Manila
3,438,334
11.8
8
2 Egypt
Al-Iskandariyah
(Alexandria)
2,723,464
21.2
18
3 Nigeria
Lagos
2,121,263
28.5
2
4 Liberia
Monrovia
1,751,428
34.5
61
5 Pakistan
Karachi
1,417,639
39.4
6
6 Yemen
Aden
1,235,473
43.6
132
7 Indonesia
Jakarta
836,130
46.5
7
8 Egypt
Bur Sa'id (Port Said)
672,210
48.8
146
9 Bangladesh
Khulna
635,950
51.0
56
10 India
Kolkata
(Calcutta)
547,004
52.9
3
11 Thailand
Krung Thep (Bangkok)
546,157
54.8
13
12 Cote
d'Ivoire Abidjan
543,928
56.6
25
13 Benin
Cotonou
491,049
58.3
100
14 Bangladesh
Chittagong
489,789
60.0
22
15 Viet Nam
Thanh Pho Ho Chi Minh
433,176
61.5
16
16 Myanmar
Yangon
384,381
62.8
17
17 Guinea
Conakry
383,551
64.1
46
18 Angola
Luanda
346,973
65.3
33
19 Brazil
Rio
de
Janeiro
344,034
66.5
9
20 Senegal
Dakar
299,405
67.5
40
21 Nigeria
Warri
266,667
68.5
140
22 Somalia
Mogadishu
235,670
69.3
71
23 Philippines
Taguig
232,703
70.1
147
24 Nigeria
Port
Harcourt
222,714
70.8
84
25 Philippines
Kalookan
212,853
71.6
74
25
Figure 1
Incremental Impact of Selected Indicators of Storm Surges (Percentage)
*
The large incremental impact of storm surges on ‘agricultural areas’ in the Middle East and
North Africa arises mostly from the estimated incremental impact in Egypt (326%) and Algeria
(143%).
26
Figure 2
Impact Zones for 1 Meter Sea-Level Rise and
Intensification of Storm Surges, and Likely Changes in Unprotected Shorelines
Illustrative Cases: Bangladesh and Viet Nam