d i s c u s s i o n pa p e r n u m B e r 1 0 octoBer 2010 d e v e l o p m e n t a n d c l i m at e c h a n g e d i s c u s s i o n pa p e r s 1 57475 D E V E L O P M E N T A N D C L I M A T E C H A N G E Economics of Coastal Zone Adaptation to Climate Change D I S C u S S I O N PA P E R N u M B E R 1 0 OCTOBER 2010 D E V E L O P M E N T A N D C L I M A T E C H A N G E Economics of Coastal Zone Adaptation to Climate Change Robert Nicholls, Sally Brown, Susan Hanson School of Civil Engineering and the Environment University of Southampton and Jochen Hinkel Potsdam Institute for Climate Impact Research (PIK) Research Domain Transdisciplinary Concepts and Methods, Pappelallee Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use and cita- tion of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the World Bank. Copies are available from the Environment Department of the World Bank by calling 202-473-3641. © 2010 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, NW Washington, DC 20433, U.S.A. Telephone: 202-473-1000 Internet: www.worldbank.org/climatechange E-mail: feedback@worldbank.org All rights reserved. October 2010 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of the World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. RIGHTS AND PERMISSIONS The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applica- ble law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone 978-750-8400; fax 978-750-4470; Internet: www.copyright.com. Cover images: © Shutterstock Images, LLC. All dollars are U.S. dollars unless otherwise indicated. III Contents Acknowledgments vii Executive Summary ix 1. Context 1 1.1 What are the potential impacts of climate change, including extreme weather events, on the sector? 1 1.2 Who (across and within countries) is likely to be most affected? 4 1.2.1 Geographically 4 1.2.2 By income or vulnerability class 4 1.3 What experience is there with adaptation in the sector? 4 1.3.1 Autonomous adaptation 5 1.3.2 Public sector investment 5 1.3.3 "Soft" adaptation--policies and regulations 7 1.3.4 Reactive (and proactive) adaptation 7 1.4 What is the nature and extent of adaptation/development deficit in this sector? 7 1.5 How will emerging changes in development and demographics influence adaptation? 8 2. Literature Review 9 2.1 Previous studies relevant to the sector 9 2.1.1 Nature and extent of damages 9 2.1.2 Nature of adaptation and its cost, private and public 9 2.1.3 Strategic conclusion (timing, sequencing, policy, etc.). 10 2.2 How this study complements existing work 11 3. Methodology 12 3.1 How we represent the future--2010 to 2050 12 3.1.1 The baseline 12 3.1.2 Climate change scenarios 13 3.2 How climate change impacts are calculated 15 3.3 How costs of adaptation are defined and calculated 16 3.3.1 Land Use Planning 17 3.3.2 Dike Maintenance and Operation 17 3.3.3 Port Upgrade 18 3.4 Data (Sources, Assumptions, and Simplifications) 18 IV e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e 4. Results 20 4.1 Investment costs (upfront and maintenance) in the baseline scenario 20 4.2 Investment costs (upfront and maintenance) due to climate change 21 4.2.1 High, medium, and low scenarios: Global adaptation costs 21 4.2.2 Medium scenario: Adaptation costs in World Bank Regions 21 4.2.3 Effect of cyclone activity and of no population growth in the coastal zone on adaptation costs 23 4.2.4 Adaptation costs: Synthesis 23 5. Limitations 25 5.1 Treatment of adaptation/development deficit 25 5.2 Treatment of extreme events 25 5.3 Treatment of technological change 26 5.4 Treatment of inter-temporal choice 26 5.5 Treatment of "soft" adaptation measures 26 5.6 Treatment of cross-sector measures 26 5.7 Areas for follow-up work and research advances 27 Appendix 1. EACC Population and GDP Projections 28 A1.1 Population Projections 28 A1.2 GDP Projections 28 Appendix 2. Adaptation Improvements 31 A2.1 Dike Maintenance and Operation 31 A2.2 Port Upgrade 32 A2.2.1 Methodology 33 A2.2.2 Source data 33 A2.2.3 Port selection criteria 33 A2.2.4 Traffic to area calculations 33 A2.2.5 Tonnage and containers 34 A2.2.6 Oil and petroleum products 34 A2.2.7 Amalgamating data 34 A2.2.8 Costs of upgrade 34 A2.2.9 Results/Discussion 35 Appendix 3. Results by World Bank Region (excluding high-income countries) 36 References 43 Tables 1 Main Climate Drivers for Coastal Systems, Trends due to Climate Change, and Main Physical and Ecosystem Effects (adapted from Nicholls and others 2007a) 2 2 Main Effects of Relative Sea-Level Rise 3 3 Summary of Climate-Related Impacts on Socioeconomic Sectors in Coastal Zones 3 4 Climate-Induced Global Mean SLR Scenarios used in EACC study 14 5 Sea-Level Rise and Impact/Adaptation Assessment Decisions 14 D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S V 6 Coastal Scenario Combinations used in EACC Study 15 7 Sea-Level Rise Effects, Impacts, and Adaptation Options Considered 17 8 Incremental Average Annual Costs (2010s­2040s) of Adaptation for Coastal Protection and Residual Damages by Scenario under the No-Rise SLR Scenario 20 9 Incremental Average Annual Costs (2010s­2040s) of Adaptation for Coastal Protection and Residual Damages by Scenario 22 A1.1. Aggregate GDP Projections from Economics Models 29 A1.2. Aggregate GDP Projections from SRES Scenarios 30 A2.1. Estimated Regional Costs of Port upgrade Costs Based on Elevating Port Area for High, Medium, and Low Sea-Level Scenarios to 2050 35 A3.1. Incremental Annual Costs of Adaptation for Coastal Protection and Residual Damages by Region and Decade for the No-Rise SLR Scenario 37 A3.2. Incremental Costs of Adaptation for Coastal Protection and Residual Damages by Region and Decade for the Low SLR Scenario 38 A3.3. Incremental Annual Costs of Adaptation for Coastal Protection and Residual Damages by Region and Decade for the Medium SLR Scenario 39 A3.4. Incremental Annual Costs of Adaptation for Coastal Protection and Residual Damages by Region and Decade for the High SLR Scenario 40 A3.5. Incremental Annual Costs of Adaptation for Coastal Protection and Residual Damages by Region and Decade for the High SLR Scenario with Cyclones 41 A3.6. Incremental Annual Costs of Adaptation for Coastal Protection and Residual Damages by Region and Decade for the High SLR Scenario with No Population Growth 42 Figures 1 Potential Responses to Coastal Hazards 5 2 Schematic of Module Linkages in the DIVA Model 12 3 Global Incremental Adaptation Costs for the High, Medium, and Low SLR Scenarios 21 4 Beach Nourishment Costs in the World Bank Regions for the Medium SLR Scenario 23 5 Percentage of Adaptation Costs from Sea Dikes, River Dikes, Maintenance Costs, and Beach Nourishment for Six World Bank Regions 23 A2-1 World Tonnage for Goods Loaded and unloaded, 1970­2007 32 VII aCknowledgments the socioeconomic scenarios that were used, and to Siobhan Murray, who provided data on the present occurrence and absence of tropical storms in coastal We are indebted the DINAS-COAST consortium who areas. Abiy Kebede assisted with the preparation of the developed the DIVA model that made this work possi- final report. ble. Thanks also go to Gordon Hughes, who provided Ix exeCutive summary on current investment in coastal adaptation and expert knowledge on the level of preparation for sea-level rise and climate change. Selected residual impacts that This report explores the answer to a difficult question: remain even with adaptation are also reported (e.g., land What are the potential costs for coastal adaptation from loss costs, coastal flood costs, and the number of people 2010 until 2050 in response to human-induced climate flooded), stressing that larger investments would be change? The work reported here builds on the earlier required to avoid all impacts of sea-level rise, if this is estimate of the United Nations Framework Convention even possible or desirable. on Climate Change (Nicholls 2007) of incremental protection costs in 2030. While these have been Four scenarios of global sea-level rise are considered: a improved in a number of aspects, the results remain a no-rise in sea level and temperature (the reference case preliminary first estimate of the possible adaptation of no climate change) and low, middle, and high needs and they show that significant further analysis of scenarios embracing a rise to 2100 of between 40 and the topic is necessary. 126 cm. These scenarios were selected to represent interesting, useful, and plausible scenarios to adopt for In terms of climate change, sea-level rise is the climate the exercise of adaptation planning under uncertainty. driver that is analyzed; the possibility of enhanced They were informed by the Intergovernmental Panel storm impacts due to higher water levels in areas subject on Climate Change's Fourth Assessment Report to tropical storms and cyclones is also considered as a (IPCC 2007) and the subsequent debate about the sensitivity analysis with the high sea-level-rise scenario. possibility of higher rises in sea level during the The analysis uses the framework of the Dynamic twenty-first century, and they should not be interpreted Interactive Vulnerability Assessment (DIVA) model to as predictions. Note that they are not specifically linked explore the costs of three main protection responses to to temperature rise. The impacts are considered in rela- climate change: tion to an Economics of Adaptation to Climate Change socioeconomic scenario that is quite similar to · Sea and river1 dike construction and maintenance the Special Report on Emission Scenarios A2 scenario. costs Following best engineering practice for sea and river · Beach nourishment dikes, sea-level rise is anticipated in terms of additional · Port upgrade. height for 50 years into the future (i.e., expected extreme sea levels in 2100 determine the dike heights These adaptation methods are applied using a standard built in 2050). For other adaptation measures, there is methodology around all the world's coasts using criteria no anticipation of future conditions, again reflecting that select optimum or quazi-optimum rule-based adap- best engineering practice. Ports are treated separately, as tation strategies. If we protect following the DIVA approach, the actual damages of sea-level rise will be much lower than the potential damages of sea-level rise 1 The impact of sea-level rise on rivers concerns the incremental costs of upgrading river dikes across coastal lowlands where sea-level rise will if protection is ignored. The resulting adaptation costs raise extreme water levels. Additional upgrades may be required if are interpreted in a broad sense based on information extreme river flows are increased, but this factor is not investigated here. x e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e they are only upgraded at the end of their design life- for climate and development policy. Clearly, a wider time (i.e., estimated here at 2050). range of adaptation options than considered in DIVA are available, and this may lead to successful adaptation Even without climate change, there are adaptation strategies of lower cost than estimated here. However, needs and residual impacts: DIVA provides a minimum realizing these benefits will require long-term strategic estimate of these costs, but some aspects of these costs planning and more integration across coastal planning are not considered. Assuming sea-level rise, global adap- and management on a sub-global scale. Few if any tation costs are in the range $26­89 billion a year by the countries have this capacity today and an enhancement 2040s: the cost depends on the magnitude of sea-level of institutional capacity for integrated coastal manage- rise. Most of these investments would be sea dike ment would seem a prudent response to climate change construction, and their maintenance costs would rise (as well as realizing benefits for non-climate issues). with time. Beach nourishment costs are also significant While all countries need to develop and enhance such and would also increase with time. Other adaptations, capacity, the need is greater in poorer countries--with such as river dikes in coastal lowlands and port small islands, populated deltaic areas, and Africa's coast upgrades, are almost negligible at a global level. presenting some of the greatest challenges. In these areas, the need for capacity development of coastal Putting these results in context, it is not clear that all management institutions linked to disaster preparedness the investments that DIVA suggests are prudent are is largest, and this is an important issue for being made, even under today's conditions: this could be development. considered an "adaptation deficit" that might usefully be assessed. If there is a large adaptation deficit, then the These global studies need to be reinforced by national investment levels estimated here will be insufficient to case studies to better understand how adaptation might adapt to climate change and the residual impacts will be operate on the ground, including the relationship with much larger than estimated here. Policymakers need to wider coastal management and non-climate-change be aware of the adaptation deficit and its implications issues. 1 1. Context expanding exposure to coastal hazards associated with climate variability such as storms (as well as non-climate events such as tsunamis). As an example, about 120 This study estimates the costs of adaptation to climate million people are on average exposed every year to tropi- change in coastal areas and is a background paper for the cal cyclone hazard (UNDP 2004). At least 300,000 people World Bank Economics of Adaptation to Climate were killed in Bangladesh in 1970 by a single cyclone. Change (EACC) study. Sea-level rise is one of the issues Worldwide, from 1980 to 2000 a total of more than that brought human-induced climate change to the fore 250,000 deaths have been associated with tropical due to the large concentration of settlements and cyclones, of which 60 percent occurred in Bangladesh. economic activity in low-lying coastal areas. The issue has Most recently, in 2008, Cyclone Nagris in Myanmar been extensively assessed since the 1980s (e.g., Barth and caused at least 138,000 fatalities. Exposure and asset loss Titus 1986; Milliman and others 1989; Warrick and is also significant, especially in the industrial world, and others 1993), with the specter of millions of environmen- there has been significant growth in losses, driven largely tal refugees as a worst-case impact. Adaptation needs and by the increase in exposure (e.g., Pielke and others 2008). costs were considered from the beginning, drawing on the The growth of population and especially asset exposure is extensive experience of flood and erosion management, expected to continue to grow, with the developing world including on subsiding coasts. The global costs of protect- contributing the most change (Nicholls and others 2008a; ing developed coasts against sea-level rise (SLR) were Hanson and others 2009). Without appropriate adapta- first estimated by the Intergovernmental Panel on tion, this will translate into growing losses. Climate Change (IPCC) in 1990, with improvement by Hoozemans and others (1993) (see also Nicholls and Climate change will exacerbate these hazards and Hoozemans 2005). There have been updates of these threaten much greater losses in the future, as summa- costs based on several different methodologies, as outlined rized in Table 1. Rising sea levels due to global warming below. However, other dimensions of climate change in have received most attention to date, with thermal coastal areas have received less quantitative assessment expansion and the melting/disintegration of the small and could raise damage and adaptation costs, most espe- glaciers and the large ice sheets of Greenland and cially more-intense hurricanes and tropical storms, which Antarctica being the underlying cause. Changing water are investigated here (Nicholls and others 2007a). levels are already an issue, and in the twentieth century global mean sea levels rose an estimated 17­19 cm 1.1 w hat are the potential impaC ts (Bindoff and others 2007; Jevrejeva and others 2008). of C limate Change, inCl uding This was primarily due to thermal expansion and the extreme weather events, on t h e melting of the small land-based glaciers. se Ctor? Human-induced global warming is expected to cause a Coasts contain high and growing concentrations of people significant acceleration in sea-level rise throughout the and economic activity (Sachs and others 2001; Small and twenty-first century due to continued thermal expan- Nicholls 2003; Nicholls and others 2007a; McGranahan sion and the melting of land-based ice. There is some and others 2007). Hence, there is a significant and debate about the potential magnitude of these changes, 2 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e table 1. main Climate drivers for Coastal systems, trends due to Climate Change, and main physiCal and eCosystem effeCts (adapted from Nicholls aNd others 2007a) Climate driver (trend) Main physical and ecosystem effects on coastal systems CO2 concentration () Increased CO2 fertilization; decreased seawater pH (or "ocean acidification") negatively impacting coral reefs and other pH-sensitive organisms Sea surface temperature (, R) Increased stratification/changed circulation; reduced incidence of sea ice at higher latitudes; increased coral bleaching and mortality; poleward species migration; increased algal blooms Sea level (, R) Inundation, flood and storm damage; erosion; saltwater intrusion; rising water tables/ impeded drain- age; wetland loss (and change) Intensity (, R) Increased extreme water levels and wave heights; increased episodic erosion, storm damage, risk of flooding, and defense failure Storm Frequency (?, R) Altered surges and storm waves and hence risk of storm damage and flooding Track (?, R) Wave climate (?, R) Altered wave conditions, including swell; altered patterns of erosion and accretion; re-orientation of beach orientation Runoff (R) Altered flood risk in coastal lowlands; altered water quality/salinity; altered fluvial sediment supply; altered circulation and nutrient supply key: = increase; ? = uncertain; r = regional variability with the possible contribution of Greenland and The other climate factors shown in Table 1 are all Antarctica being important: the range produced by the potentially important. Of particular significance are IPCC's Fourth Assessment Report (AR4) (Meehl and changes in storms. It has been suggested that tropical others 2007) quantified rises of up to 59 cm,2 but the storms may increase in intensity as the world warms report is clear that the upper bound of SLR rise remains (Meehl and others 2007), and the possibility of more- uncertain and unquantified due to the uncertainty about intense storms in the coastal areas experiencing them is the response of the large ice sheets (see IPCC 2007). analyzed with sea-level rise. More recent studies have emphasized a range of rises, with the upper limit exceeding the quantified AR4 Collectively, the climate effects shown in Tables 1 and 2 range (e.g., Rahmstorf 2007; Grinsted and others 2009; can have a range of negative socioeconomic impacts, as Vermeer and Rahmstorf 2009). Hence, it is clear that at summarized in Table 3. This shows that the impacts of present a rise of 1 m or more through this century climate change on coasts are quite varied. In this analy- cannot be excluded (Lowe and others 2009) and needs sis, we consider the sea-level rise and the possible to be evaluated in impact and adaptation assessments. increases in the intensity of tropical storms, as they are some of the largest impacts. The impacts of sea-level rise are produced by relative (or local) SLR, which includes regional sea-level varia- Hence, climate change and sea-level rise will have adverse tion and geological uplift/subsidence (Nicholls in press). impacts and costs on coastal areas around the world Subsidence exacerbates climate change, as observed in through the twenty-first century and beyond (Nicholls many subsiding deltas and coastal cities, while uplift and others 2007a). The impacts of SLR also depend counters sea-level rise to some degree, such as observed upon future socioeconomic change (e.g., Nicholls 2004). in parts of Scandinavia (e.g., Helsinki). In this study, Regardless of climate change, socioeconomic change will global mean SLR is downscaled using local estimates of result in profound changes in the coastal zone, such as a uplift/subsidence, as explained later. Human-induced growth in population and coastal infrastructure (e.g., subsidence is not considered, but this will lead to local increased values of relative sea-level rise. The physical impacts of SLR are varied and summarized in Table 2. 2 76 cm if scaled-up increased ice sheet discharge is included. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 3 table 2. main effeCts of relative sea-level rise this includes relevant climate and non-climate factors that interact with the physical effects; some factors (e.g., sediment supply) appear twice, as they may be influenced by both climate and non-climate factors (adapted from nicholls 2002). Other relevant factors Physical effect Climate Non-climate Inundation, flood Surge Wave and storm climate, mor- Sediment supply, flood management, morphological and storm damage phological change, sediment change, land claim supply Backwater effect Runoff Catchment management and land use (river) Wetland loss (and change) CO2 fertilization Sediment supply, migration space, direct destruction Sediment supply (Long-term) erosion Sediment supply, wave and Sediment supply storm climate Saltwater intrusion Surface waters Runoff Catchment management and land use Groundwater Rainfall Land use, aquifer use Rising water tables/ impeded drainage Rainfall Land use, aquifer use table 3. summary of Climate-related impaCts on soCioeConomiC seCtors in Coastal Zones most are linked to mean or extreme sea level, as indicated (adapted from nicholls and others, 2007a). Climate-related impacts Temperature Extreme Floods Erosion Saltwater Coastal rise events (sea Rising water (sea level, intrusion socioeconomic (air and (storms, sea level, tables storms, (sea level, Biological effects sector seawater) level, waves) runoff) (sea level) waves) runoff) (all climate drivers) Freshwater x x x x -- x x resources Agriculture x x x x -- x x and forestry Fisheries and x x x -- x x x aquaculture Health x x x x -- x x Recreation and x x x -- x -- x tourism Biodiversity x x x x x x x Settlements/ x x x x x x -- infrastructure key: x = strong impacts; x = weak impacts; -- = negligible impacts or not established Nicholls and others 2008b). These baseline changes due rational response on populated coasts (although other to non-climate factors need to be considered in addition adaptation strategies might be considered). to the effects of climate change. Investment in adaptation allows these damage costs to be substantially reduced, This document explains the methods that are being and all the available analyses suggest that protection is a used within the World Bank study to estimate potential 4 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e protection costs from 2010 to 2050. It explores the costs 1.2.2 by income or vulnerability class of protecting the world's coast against sea-level rise using the Dynamic Interactive Vulnerability Assessment The issue of the distribution of the impacts of sea-level (DIVA) model, assuming dike construction and upgrade rise has been less considered. Anthoff and others (2006) and beach nourishment where this is optimal or quazi- applied equity weighting to the damages of SLR. optimal. Residual impacts after protection are also Taking these distributional issues into account increased reported. In the analysis, DIVA has been extended to the damage estimates by a factor of three, reflecting the also consider the costs of port upgrade and the mainte- fact that the costs of sea-level rise fall disproportionately nance and operational costs for dikes. In the following on poorer developing countries. This is consistent with treatment, the main focus is the impacts and responses the fact that Africa is consistently identified as being to sea-level rise, with some consideration of more- highly vulnerable to sea-level rise (and other aspects of intense tropical storms in those areas already so affected. climate change). 1.2 w ho (a Cross and within 1 . 3 wh at e x p e r i e nC e i s t h e r e w i t h Co untries) is likely to be mos t a d a p tat i o n i n t h e s eC to r ? affe Ct ed? Coastal areas have a long and established tradition of Both direct and indirect effects are possible due to adaptation. While there has often been a focus on protec- climate change. Here the direct effects are emphasized. tion, the available adaptation measures can be placed in a wider context as one of three generic options (IPCC 1.2.1 g eogr a p h i c a l l y 1990; Bijlsma and others 1996; Klein and others 2001): In general, all people in low-lying coastal areas are threat- · (Planned) Retreat ­ The impacts of sea-level rise ened to varying degrees. A range of analyses have consis- are allowed to occur, and human impacts are mini- tently found that deltaic areas and small islands are the mized by pulling back from the coast via land use most threatened coastal settings (e.g., Nicholls and others planning, development control, set-back zones, etc. 2007a). Deltas are by definition at an elevation related to · Accommodation ­ The impacts of sea-level rise are present sea level; many of them are densely populated and allowed to occur and human impacts are minimized are subsiding due to both natural and human causes by adjusting human use of the coastal zone to the (Ericson and others 2006; Syvitski and others 2009). hazard via increasing flood resilience (e.g., raising Small islands are also threatened (Mimura and others homes on pilings), early warning and evacuation 2007). Atolls, like deltas, are low-lying areas threatened by systems, risk-based hazard insurance, etc. submergence, with the Maldives being an excellent exam- · Protection ­ The impacts of sea-level rise are con- ple of a nation of atolls. However, all islands are threat- trolled by soft or hard engineering (e.g., nourished ened, as economic activity is concentrated around the beaches and dunes or seawalls), reducing human coast and the capacity to respond is nearly always much impacts in the zone that would be affected without lower than in continental countries. Lastly, poor regions protection. However, a residual risk always are problematic as they have a low capacity to adapt. remains, and complete protection cannot be achieved. Managing residual risk is a key element Geographically, regions with large densely populated of a protection strategy that has often been over- deltas in South, Southeast, and East Asia contain the looked in the past. largest concentrations of people threatened by sea-level rise. All small island regions are threatened, including The three approaches are illustrated in Figure 1. the Caribbean and the Indian and Pacific Oceans. While the absolute impacts in small islands are quite Throughout human history, improving technology has small at a global scale, in relative terms the impacts are increased the range of adaptation options in the face of highest (Nicholls 2004; Nicholls and Tol 2006). Lastly, coastal hazards, and there has been a move from retreat Africa is threatened due to its relative poverty, rapid and accommodation approaches to hard protection and demographic growth, and limited capacity to respond. active seaward advance via land claim. This is illustrated D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 5 Hence, the cost estimates are grounded in coastal engi- figure 1. potential responses to neering experience and are reasonably robust. Coastal haZards The United Nations Framework Convention on Retreat Climate Change (UNFCCC) assessment (Nicholls 2007) used the DIVA database and focused on dike construction and upgrade and on beach nourishment. The dike costs are derived from Hoozemans and others Accomodatiuon (1993), while nourishment costs are derived from the recent experience of Deltares, among others, in beach Protect nourishment projects around world. Residual damages ­ Soft are also estimated in terms of land values, depth- ­ Hard damage curves, and the costs of relocating people. The computations are conducted on 12,148 coastal segments source: van koningsveld and others, 2008. (average length of about 70 km) that collectively make up the world's coast, except for Antarctica (McFadden and others 2007; Vafeidis and others 2008). The DIVA by the changing approaches to managing coastal flood- database is based on extensive experience and a realistic ing and erosion in the Netherlands (van Koningsveld description of the adaptation measures, which are and others 2008). More recently, there has been a move informed by empirical experience. It is much more from hard to softer protection in the Netherlands, based detailed than any earlier assessment tool in terms of the on large-scale beach nourishment with sea-dredged impacts and adaptation responses considered as well as sand--now amounting to 12 million m3/yr. There are the spatial resolution of the computations: the nearest also concerns about making adaptation multifunctional assessment tool is the FUND model, which has a such that environmental impacts are minimized while national resolution and considers both impacts and ensuring human safety: this suggests, for example, protection costs (e.g., Nicholls and Tol 2006). moves from fixed to mobile surge barriers to allow water and biotic exchanges. Looking to the future, the 1 . 3 . 1 a ut o n o m o u s a d a p t a t i o n Deltacommissie (2008) has considered the national response of the Netherlands to sea-level rise over the Autonomous adaptation describes the spontaneous twenty-first century. Hence, the Netherlands illustrates adjustments that occur in response to climate (or other) how thinking about adaptation is rapidly evolving. change without any active policy intervention. Hence Looking more widely, there is an important debate autonomous adaptation has negligible cost. There is concerning the appropriate mixture of hard and soft some autonomous adaptation in response to climate protection, accommodation, and retreat. change in coastal areas, such as increased accretion of salt marshes or market adjustment to the price of land In terms of costing, most experience is available or properties after a coastal disaster. However, human concerning traditional hard engineering approaches and impacts in terms of flooding and erosion are little protection. There is much less understanding of retreat reduced by autonomous adaptation in coastal areas. and accommodation costs, reflecting the much more Hence, it has been concluded for the last decade that limited experience of these measures. Most of the avail- significant planned adaptation is essential to manage able cost estimates are bottom-up ones based on a long the growing risks from sea-level rise (e.g., Klein and history of coastal management and engineering experi- Nicholls 1999). ence. This mainly assumes protection via dikes (for flood management) and nourishment (to preserve 1.3.2 public sector investment beaches). The costs of these measures were documented globally using a series of country cost factors by IPCC There is considerable experience of adaptation in coastal (1990) and Hoozemans and others (1993), based on the zones to a range of drivers, of which climate change is global experience of Delft Hydraulics (now Deltares). only one. Unlike adaptation in many other sectors, 6 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e coastal adaptation measures usually represent a collec- in 2035, with the increase being primarily due to tive government-led activity, reflecting that the coast is climate change (Environment Agency 2009). a shared resource (Klein and others 2000). Hence while · Japan. 120 to 150 billion yen per year from 2003 to some adaptation will need to be funded by private 2006. investment (e.g., port and harbor upgrade), much of the · Netherlands. $600­1,200 million (in 2006 prices), or cost falls on government finances. However, individual 0.1­0.2 percent of gross domestic product (GDP). adaptation measures are also apparent. Insurance is a This is expected to double or triple from 2020 to mechanism that helps private individuals gain resources 2050 as the recent recommendations of the to recover from disasters such as coastal flooding and is Deltacommissie (2008) are implemented. This rep- potentially an important response mechanism (Clarke resents a combination of climate change adaptation, 1998; Grossi and Muir-Wood 2006). The availability of looking 100­ 200 years into the future, and increas- appropriate insurance varies greatly between coastal ing safety to much higher levels (risk of failure will countries; it is unavailable in many developing countries be < 1 in 100,000 in any year). and in mainland Europe (as the government is the insurer of last resort), while in the United Kingdom and For individual projects, Nicholls (2007) identified the the United States it is the norm. following costs: While there is significant interest in elaborating coastal · The Maldives. "Safe Island" Projects for tsunamis-- adaptation measures and understanding their costs (e.g., the cost of reclamation and coastal protection UNFCCC 1999; Klein and others 2001; Bosello and including harbor works for the Vilifushi project was others 2007), hard numbers on investment in coastal about $23 million. adaptation are difficult to identify as there is never a · Venice, Italy. The MoSE Project to manage flooding single "Ministry for Coastal Adaptation" with published of Venice cost roughly 4,000 million euro's. The accounts in any country. The reality is that coastal adap- project is mainly addressed to solve current flood tation costs fall between government and the private problems. sector, and different ministries are responsible for differ- · St Petersburg, Russia. The Flood Protection Barrier ent aspects of the process. For instance, in England and was started in the 1980s, and was 65 percent com- Wales, the major investment in coastal adaptation is in pleted, when construction was halted until about flood and erosion management, but the budget covers 2002. Then completion was funded by the all flood and erosion management--that is, manage- European Bank for Reconstruction and ment of all flood mechanisms, including inland flood- Development, costing about 440 million euros. ing. Integrated coastal management in England and Again, the Barrier is mainly designed to solve cur- Wales is covered by a separated budget, and this invest- rent flood problems. ment is quite small compared with that in flood and · London, UK. The Thames Estuary 2100 Project is erosion management. investing £15 million on appraising the flood man- agement options for London for the twenty-first Nicholls (2007) identified the following national/regional century and beyond, including building a completely estimates of current investment in coastal adaptation new downstream barrier. Unlike the previous two (reflecting many drivers, including climate change): cases, this is mainly a response to climate change. While nothing has been decided, costs of £4­6 bil- · European Union. The total annual cost of coastal lion for this century have been mentioned for adaptation for erosion and flooding across the upgrade, while £10­20 billion has been mentioned European Union was an estimated 3,200 million for a new downstream barrier, which would be the euro's (in 2001). response to a large rise in sea level (several meters). · England. The flood and coastal management bud- get for coasts is roughly £250 million per annum In conclusion, the present investment in coastal engi- and growing. New estimates show expenditure on neering is significant, and any investment in adapting to all flood defense (rivers and coasts) rising from £575 climate change will be building on a portfolio of exist- million per annum in 2011 to more than £1 billion ing activity in many parts of the world. In many parts of D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 7 the developing world, however, the major investment to rising sea levels or model outputs of future impacts). We date in coastal engineering are port and harbor assets, mainly observe reactive adaptation in coastal zones at and the types of investments considered in the analyses present (e.g., Tol and others 2008; Moser and Tribbia in this report will represent a significant departure from 2008), with the history of New Orleans illustrating this established practice. well. Each major flood there triggered major investment in better defenses, including after Hurricane Betsy in 1.3.3 "s oft" a d a p t a t i o n -- p o l i c i e s a n d 1965 and Hurricane Katrina in 2005. Roughly $10 regulations billion is being spent to upgrade the defenses post- Katrina, but this will only achieve the design standards As well as the hard infrastructure considered here, and thought to exist before that storm. Substantial addi- this includes soft engineering such as beach nourish- tional investment would be required to achieve ment, there is much "soft" infrastructure that constitutes Category 5 hurricane projection. Anecdotally, numbers an important component of the adaptive capacity that is as high as $50 billion have been suggested. essential for coastal adaptation to take place (Smit and others 2001; Adger and others 2007). Institutions are The dynamic nature of the risks due to changing fundamental to manage the coast, including addressing climate means that a more proactive approach to assess- the challenges raised by climate change and other activ- ment and adaptation planning is essential for coastal ities such as warning services. For instance, storm tide areas; otherwise these risks will reach unacceptable warning services are an important component to an levels (Nicholls and others 2007a). Even without integrated management response to potential flood climate change, growing populations and economic events in low-lying coastal areas, as demonstrated in wealth in coastal areas suggests that substantial invest- areas as diverse as the United States, the southern ment in coastal adaptation would be required through- North Sea, and Bangladesh. out the twenty-first century (e.g., Nicholls and others 2008a), again demanding proactive assessment and There is also the issue of the context in which adapta- responses. In a few limited cases, present adaptation tion occurs. Traditionally, coastal management has been investment includes anticipating climate change (e.g., in sectoral in nature, and the focus of management has the United Kingdom and the Netherlands). Some of been a single goal rather than addressing multiple issues. the limited cases of anticipatory adaptation have been Integrated coastal management is an attempt to address highlighted in Section 1.3.2. this problem that is receiving widespread support in coastal areas both academically (e.g., Cicin-Sain and 1 . 4 wh at i s t h e n at u r e a n d e x t e n t Knecht 1998; Brown and others 2002; Kay and Alder o f a d a p tat i o n / d e v e l o p m e n t d e f iCi t 2005; Williams and Micallef 2009) and in policy terms i n t h i s s eC to r ? (e.g., European Union 2010). However, the application and success of this approach remains uncertain. In Analysis of climate change often implicitly assumes that general, all of these "soft" measures are low cost in terms the current state is optimal, while the current state is of application compared with hard protection measures, often far from optimal, as shown by Hurricane Katrina although there are many other barriers to application. in 2005 and Cyclone Nagris in 2008. This gap has been However, the difficulty in developing these capacities termed the adaptation deficit (Burton 2004; Parry and where they do not exist should not be underestimated, others 2009). In coastal areas, the adaptation deficit is and this is an important issue for the wider development an important issue due in part to a reactive approach to agenda and sustainable development in general. adaptation, a general under-recognition of the risks in many coastal areas, and the rapid expansion of the 1.3.4 r eact i v e ( a n d p r o a c t i v e ) a d a p t a t i o n population and economy in many areas, which means that historic hazard events are little guide to the level of Reactive adaptation is adaptation that occurs in contemporary (or future) exposure or risks from hazard response to actual (or observed) change and impacts, as events (e.g., Nicholls and others 2008a). As just noted, opposed to proactive adaptation that takes place in New Orleans is spending $10 billion post-Katrina, anticipation of expected change (such as projections of while the actual investment required to make New 8 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e Orleans' defenses sufficient to survive a Category 5 1 . 5 h o w w i l l e m e r g i n g C ha n g e s i n hurricane is on the order of $50 billion. Defense stan- d e v e l o p m e n t a n d d e m o g r a p h iC s dards also give an indication of the adaptation deficit. in f l u e nC e a d a p tat i o n ? For instance, New York City has much lower standards of protection by one to two orders of magnitude than Coastal populations and economies are presently grow- European cities with a similar or lower exposure to ing rapidly with little regard to the growing risks of coastal flooding (Nicholls and others 2008a). It can be coastal locations. For instance, population is increasing argued that this represents an adaptation deficit as was at rates that often double global trends. As such, the seen in New Orleans, although others may interpret it exposure of coastal areas is growing rapidly. This is as differing attitudes to risk. illustrated in the coastal scenarios of Nicholls (2004), Nicholls and Lowe (2004), and Nicholls and others Apart from these industrial world examples, the adapta- (2007b), where population growth of up to fourfold tion deficit in coastal areas is poorly quantified and our may occur in the coastal zone. If this development understanding of it is essentially qualitative. However, it continues in a business-as-usual manner, this growth is a significant issue, as many developing countries have will strongly reinforce the need for protection, as few organized defenses or flood management systems demonstrated by Anthoff and others (2010). The use of comparable to those in the developed world. This is an retreat and accommodation options could reduce the important deficiency that future assessments need to need for protection, but these policies have a long lead address. time, and they require proactive implementation to be fully effective. 9 2. literature review growth and a strong urbanizing trend in many parts of the world (e.g., Small and Nicholls 2003; McGranahan and others 2007); see Section 1.5. The threat is particu- 2.1 p revious studies relevant to larly strong in populated deltas and on small islands. the seCtor The abandonment of coastal islands due to sea-level rise appears a quite plausible outcome unless appropriate 2.1.1 n atur e a n d e x t e n t o f d a m a g e s adaptation can be mobilized. Hence, there is a strong consensus that the potential impacts of SLR are large. The focus in the literature is overwhelmingly on sea-level rise impacts and adaptation costs as summarized in a Until recently, no study has addressed the impacts of series of IPCC assessments (Bijlsma and others 1996; changing storms, in part due to the lack of credible McLean and others 2001; Nicholls and others 2007a). scenarios. Nicholls and others(2008a) did consider Actual impacts in coastal zones are a product of relative more-intense tropical and extra-tropical storms (follow- SLR: this is the sum of climate-induced changes and ing the regions identified by Meehl and others 2007) as non-climate effects causing land uplift/subsidence due to one factor in the potential increase in exposure of natural and human processes (Nicholls in press). Uplift/ coastal cities to coastal flooding: the effect was signifi- subsidence processes include tectonics and glacial- cant and comparable in magnitude to changes due to isostatic adjustment as well as human-induced processes, climate-induced sea-level rise, and human-induced such as subsidence due to fluid withdrawal, and drainage subsidence, but much smaller than socioeconomic of coastal soils susceptible to subsidence and oxidation. changes. Narita and others (2009) examined historical Hence, relative sea-level rise varies from place to place. It damages due to tropical storms and concluded that is generally higher than the global mean in areas that are future changes are likely to be small. Dasgupta and subsiding, which includes many populated deltas (e.g., the others (2009a) also investigated the impacts of more- Mississippi delta) (Ericson and others 2006; Syvitski and severe tropical storms and sea-level rise and found that others 2009), while many coastal cities have also subsided. severe impacts are likely to be limited to a relatively small number of countries and a cluster of large cities at The major impacts of sea-level rise have already been the low end of the international income distribution. summarized in Table 2. No published impact analysis Hence, sea-level rise appears a bigger threat globally considers all of these impacts (Nicholls in press). than more-intense storms, although in certain regions Historically, analyses have either focused on flooding or the impacts may be more comparable. land loss and have not considered both issues together, while salinization has received the least investigation. 2 . 1 . 2 n at u r e o f a d a p t a t i o n a n d i t s c o s t , Synthesis across the available literature suggests that private and public "inundation, flood and storm damage" has the largest impact potential. All coastal lowlands are threatened to Adaptation has been a feature of assessments of sea- varying degrees, and hence hundreds of millions of level rise since the 1980s (e.g., Barth and Titus 1986; people are threatened around the world today. Further, IPCC 1990). Initially, this largely built on the experi- these areas are the nexus for population and economic ence of coastal engineering, but as the need for 10 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e adaptation and interest in it increased, it has broadened and actual impacts that are much less than potential to the protect, accommodate, and retreat options impacts. Both views tend to be caricatures of real defined in Figure 1. However, most policy analyses responses, but they do stress that the inevitability of consider a choice of protect versus retreat to examine worst-case impacts should not be accepted, and they the economics of SLR, and accommodation has not point to the importance of studies like the EACC Project been considered as extensively as yet. Most analyses that to better understand adaptation and its costs. One impor- have considered protection at the global scale have tant message is the importance of continued economic considered one of two distinct approaches: growth to support the investment in adaptation. Protection is much harder to justify if a no-economic- (1) Arbitrary protection of all "developed areas,"3 as in growth scenario is considered (Anthoff and others 2010). IPCC (1990), the Global Vulnerability Assessment This shows that coastal adaptation is strongly linked to (Hoozemans and others 1993), and the Fast Track wider development goals and issues, and it has been Analyses (Nicholls 2004) argued that assistance for adaptation is critical in the developing world in the coming decades, while they (2) An optimization approach in which "economically develop the capacity to adapt (Patt and others 2010). worthwhile areas" are defended, as in Fankhauser (1995), Tol (2007), Sugiyama and others (2008), and The UNFCCC conducted the most recent assessment Anthoff and others (2010); this is normally based on of the adaptation costs for sea-level rise (Nicholls 2007; comparing avoided damage and protection costs. Parry and others 2009). This was based on assumptions of protection using dike construction and beach nour- In both cases, the costs of the required protection and ishment. The investment costs were estimated for 2030, the residual impacts in areas that are not protected can assuming a range of AR4 sea-level rise scenarios from be determined. This is not always done in economic Meehl and others (2007). As the rate of sea-level rise is terms. similar between scenarios, the range of uncertainty for adaptation costs for reactive adaptation measures (beach A fundamental result is that protection based on bene- nourishment) is small, but it is larger for dikes that fit-cost approaches greatly reduces the impacts of sea- anticipated future sea level. The UNFCCC estimated level rise, at least for people and assets, and the residual additional costs in 2030 of $4­11 billion a year, assum- damage is as much as two orders of magnitude lower ing a 50-year planning horizon and no adaptation defi- than the potential impacts (e.g., Nicholls and Tol 2006; cient. However, the costs may be underestimates if we Nicholls and others 2007b). This reflects that most consider responses to high-end SLR and other climate coasts remain undeveloped, and hence coastal infra- changes such as more-intense storms. The additional structure and people are concentrated in smaller areas residual damage attributed to sea-level rise in terms of that are more easily protected. Hence, a greater rise in sea flood and land loss is estimated at $1­2 billion a sea level translates into greater protection costs, and year. Environmental damages such as loss of coastal while residual impacts also increase, they remain a small wetlands would be in addition to this, and the costs and fraction of the potential impacts. methods of adaptation are less certain. It should be noted that the success or failure of protection 2.1.3 strategic conclusion (timing, is highly controversial, and the different views concerning s e q u e n c i n g , p o l i c y, e t c . ) . this aspect of adaptation explain much of the differences between different estimates of actual (as opposed to The results of these studies suggest that impacts could potential) impacts of sea-level rise (Nicholls and Tol be disastrous for coastal areas unless there is significant 2006; Nicholls in press). Pessimists expect protection to adaptation. The available literature also demonstrates a either be unavailable or to fail, and hence potential significant debate about adaptation and its likely impacts translate into actual impacts and the world faces success. These differing views can be seen as caricatures, tens of millions of environmental refugees due to sea-level rise alone (e.g., Myers 2001; Dasgupta and others 2009b). 3 Usually based on an arbitrary definition, such as all areas with a popula- Optimists expect widespread protection for sea-level rise tion exceeding 10 persons/km2. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 11 but it appears that protection is a rational response on · A time series of costs from 2010 to 2050, rather most developed coasts around the world, especially than a single snapshot under scenarios of greater economic growth. This is · Consideration of a wider range of sea-level rise sce- counter to many people's intuition about the response to narios, reflecting the post-AR4 debate on this issues sea-level rise, which is often seen as a widespread · Inclusion of more-intense tropical cyclones as a sen- retreat. However, this view fails to appreciate the devel- sitivity analysis opment of coastal engineering technology and the rela- · Improved estimates of protection costs, including tively low cost of these responses compared with what is maintenance costs for dikes and port upgrade threatened. To better understand the issue of adaptation, · Consideration of the consequences of avoiding research such as the EACC Project is fundamental. future coastal population growth, reflecting strin- gent land use planning 2.2 ho w this study Complemen t s · More explicit consideration of the adaptation existing work deficit. This study builds on all the earlier assessments, includ- ing the recent UNFCCC assessment of adaptation costs in 2030 (Nicholls 2007). A number of significant improvements have been made compared with previous studies: 12 3. methodology studies hold developing countries at their current level of development when estimating adaptation costs for both the near and medium term. Over the medium Following existing practice, the EACC study is focused term these countries will, however, become more devel- on preserving the human uses. The methodology is oped and wealthy and this, in turn, will change the based on the DIVA model, including some new exten- impact of climate change on their economies and the sions. DIVA is an integrated model that estimates type and extent of adaptation that is required, as well as impacts for given climate and socioeconomic scenarios their capacity to adapt. The EACC study accounts for and for stated adaptation options (Figure 2). Given that the impact of development on estimates of adaptation it can provide adaptation costs in coastal areas, it is well costs by establishing its own development baselines suited to the EACC Project. First the socioeconomic sector by sector (see Appendix 1). These baseline and climate change scenarios are considered, followed scenarios establish a population and GDP growth path by the different impacts. The adaptation option choices are then considered, followed by their implementation in the EACC project. figure 2. sChematiC of module linkages The DIVA model is an integrated model of coastal in the diva model systems that assesses biophysical and socioeconomic adaptation decisions are implemented at the next time impacts of sea-level rise and socioeconomic develop- step (after five years). ment. One important innovation introduced by DIVA User Selection is the explicit incorporation of a range of adaptation Socio-economic Climate change/ Adaptation options; impacts depend not only on the selected scenarios sea-level rise scenarios options climatic and socioeconomic scenarios but also on the selected adaptation strategy. DIVA is driven by climatic Initialization and socioeconomic scenarios. The climatic scenarios (for 1995) consist of the variables temperature change and sea- level rise. The socioeconomic scenarios consist of the Impact Assessment (5 year time steps to 2100) variables land-use class, coastal population growth, and Storm Backwater Erosion Salinization Wetland loss/ GDP growth. surge effect change 3.1 ho w we represent Flood risk Socio-economic impact Wetland valuation the future--2010 to 2050 Adaptation assessment 3.1.1 th e b a s e l i n e Impact/adaptation metrics In the baseline, climate is constant (i.e., maintained at (from 2000 to 2100) 1995 levels), but non-climate changes do occur, most source: authors data utilizing the diva model. especially population and GDP growth. Most existing D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 13 in the absence of climate change that determines sector- Several approaches were considered to analyze the level performance indicators such as the stock of infra- impacts of global SLR, and initially it was proposed structure assets, level of nutrition, water supply that we would construct response surfaces across a wide availability, etc. Climate change impacts and costs of range of SLR scenarios. However, this led to difficulties adaptation are then examined relative to this evolving when we considered different time frames for adapta- baseline, with no climate change. tion: as explained later, beach nourishment and port upgrade costs are based on the actual sea-level rise (to Baselines, in turn, are established across sectors using a 2050), while dike upgrade anticipates sea-level rise 50 consistent set of future population and GDP projec- years into the future (to 2100) (i.e., proactive adapta- tions. The population trajectory has been developed to tion). Therefore we need self-consistent scenarios that be consistent with United Nations Population Division evolve over time to 2050 (for socioeconomic change) middle fertility population projections for 2006 (UNPD and to 2100 (for proactive adaptation to climate 2006). In order to ensure consistency with emissions change). Hence, after a debate within the wider project, projections, the GDP trajectory is based on the average the three SLR scenarios proposed by Neumann (2009) of the GDP growth projections from five sources that were adopted, in addition to a no-SLR scenario as a provide growth estimates at a reasonable regionally reference case. These scenarios assume that sea-level rise disaggregated level: three main Integrated Assessment is effectively independent of temperature, precipitation, Models of global emissions growth--FUND (Tol and other climate parameters of interest to the EACC 2008), PAGE2002 (Hope 2006), and RICE99 study over the timescale of interest, in the sense that the (Nordhaus 2001)--and the growth projections used in scenarios not derived directly from specific Global the energy demand forecasts by the International Circulation Model runs.5 Because the main temperature Energy Agency and the Energy Information scenarios for the EACC Project are roughly consistent Administration at the U.S. Department of Energy4. with the IPCC SRES A2 scenario, the SLR projections considered here are also consistent with the A2 emis- The resulting global average real GDP per capita sions trajectory. growth rate is 2.1 percent per year, which is similar to global growth rates assumed in the Special Report on The four SLR scenarios are defined as follows: Emission Scenarios (SRES) A2 emissions scenario used in the IPCC AR4 (Nakicenovic and Swart 2000). The · No-rise scenario ­ no climate change, so sea-level study chose not to use the regionally downscaled GDP rise only results from vertical land movements; only projections from the different IPCC scenarios (available the vertical movements considered in the DIVA from the Center for International Earth Science database are considered (see Vafeidis and others Information Network, Columbia University) because 2008) and the potential for human-induced subsid- these are based on data that do not include recent ence is not considered changes, such as the continued rapid growth of China. · Low scenario ­ based on the midpoint of the IPCC AR4 A2 range in 2090­99 (Meehl and others 3.1.2 C lima t e c h a n g e s c e n a r i o s 2007); it is consistent with a MAGICC TAR A2 mid-melt 3oC sensitivity run The main climate factor considered here is climate- · Mid scenario ­ based on the Rahmstorf (2007) A2 induced sea-level rise, with some consideration of changes trajectory)6 in tropical storms. The SLR scenarios published by the IPCC AR4 Report (Meehl and others 2007) have been 4 http://www.eia.doe.gov/ widely contested since they were published: many papers have indicated the potential for larger rises than included 5 Of course, higher future temperature outcomes are correlated with higher future sea-level rise outcomes. Temperature is the main driver of in the AR4 range (e.g., Rahmstorf 2007; Vermeer and both thermal expansion of the oceans and melting of land-based ice, Rahmstorf 2009), and this has been included in some which in turn drive sea-level projections. However, there is a decoupling over decades due to the uncertainty of the response of the two major national SLR scenarios (e.g., Lowe and others 2009). continental ice sheets: Greenland and West Antarctica. These insights are acknowledged here, and a high scenario 6 This is similar to the DEFRA (2006) SLR scenario used for planning to describe sea-level rise is included for this purpose. and design of flood defenses in Great Britain. 14 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e · High scenario ­ based on the "maximum" trajectory of Rahmstorf (2007). table 4. Climate-induCed global mean slr sCenarios used in eaCC study It is important to note that these are scenarios (or plau- in cm above 1990 levels sible futures), and they do not represent our judgment of the most likely global SLR outcomes. Rather they Sea-level rise (SLR) scenario represent interesting, useful, and plausible scenarios to Year No rise Low Medium High adopt for the exercise of adaptation planning in coastal 2010 0.0 4.0 6.6 7.1 zones under uncertainty. 2020 0.0 6.5 10.7 12.3 2030 0.0 9.2 15.5 18.9 The SLR scenarios give a climate-induced global-mean 2040 0.0 12.2 21.4 27.1 rise in sea level of 16­38 cm by 2050 and 40­126 cm by 2050 0.0 15.6 28.5 37.8 2100, respectively (Table 4). The scenarios as used in 2060 0.0 19.4 37.0 50.9 the impacts and adaptation analysis are defined in 2070 0.0 23.4 47.1 66.4 Table 5: after 2050, sea-level rise only influences dike 2080 0.0 28.1 58.8 84.4 costs, as from 2010 to 2050 all the dikes are proactively 2090 0.0 33.8 72.2 104.4 upgraded to anticipate sea levels in 2100. The "no-rise" 2100 0.0 40.2 87.2 126.3 scenario allows us to explore the evolving baseline of no climate change combined with socioeconomic change. Air temperature rise is also required for the Hamburg addition to the high SLR scenario in these areas by Tourism Module (HTM) (Hamilton and others 2005a, 21007 (see Section 3.2 for more details). This again 2005b), which is used with DIVA to simulate tourist influences dike costs and residual flood damage (assum- demand for beaches: an A2 temperature scenario was ing a 50-year anticipation of future conditions). This used. aspect of the analysis constitutes a sensitivity analysis. Last, a scenario of no population growth in the coastal Intensification of tropical cyclones (or storms) in areas zone is considered. While rather an artificial scenario, it that currently experience them is of widespread concern illustrates the implications of a land use policy where all (Meehl and others 2007; Nicholls and others 2007a). As there is no scientific consensus as to whether storms will or will not intensify, we consider an arbitrary 10 percent 7 Note that other impacts of more-intense storms, especially increased increase in extreme water levels for the 100-year event in wind damage, are not considered. table 5. sea-level rise and impaCt/adaptation assessment deCisions based on the slr scenarios (in cm above 1990 levels) being used in the eaCC study for flooding and erosion impacts and beach erosion/nourishment and port upgrade (no proactive adaptation) and for dike costs (proactive adaptation over 50 years). (see table 4.) Impact/adaptation assessment Flooding, beach erosion, nourishment, port upgrade costs Sea and river dike costs Year No rise Low Medium High No rise Low Medium High 2010 0.0 4.0 6.6 7.1 0.0 4.0 6.6 7.1 2020 0.0 6.5 10.7 12.3 0.0 14.9 29.8 40.8 2030 0.0 9.2 15.5 18.9 0.0 24.7 51.4 72.6 2040 0.0 12.2 21.4 27.1 0.0 33.2 70.8 101.5 2050 0.0 15.6 28.5 37.8 0.0 40.2 87.2 126.3 D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 15 development is prohibited in areas vulnerable to erosion and inundation/flooding and steered instead to less table 6. Coastal sCenario Combinations vulnerable areas. Again, this should be considered a used in eaCC study sensitivity analysis. The scenario combinations that are being considered can be summarized in Table 6. Sea-level rise scenarios No rise Low Medium High High 3.2 ho w C limate C hange impaCt s a r e EACC Tropical Ca lC u lated socioeconomic scenarios Tropical Cyclones Constant cyclone Intensification Population and x x x x x The impacts in terms of both physical change (and GDP growth adaptation) are calculated using the DIVA model. GDP only growth x -- -- x x (i.e., constant DIVA first downscales to relative sea-level rise by coastal population) combining the SLR scenarios due to global warming with the vertical land movement. The latter is a combi- nation of glacial-isostatic adjustment according to the geo-physical model of Peltier (2000a, 2000b) and an assumed uniform natural subsidence in deltas of 2 mm/ DIVA includes beach/shore nourishment--i.e., the yr. Human-induced subsidence (due to ground fluid replacement of eroded sand--as an adaptation option abstraction or drainage) is not considered due to the for coastal erosion. In beach nourishment, the sand is lack of consistent data or scenarios. Based on the rela- placed directly on the intertidal beach, while in shore tive sea-level rise (and the influence of the selected nourishment the sand is placed below low tide, where it adaptation option), several types of bio-physical impacts will progressively feed onshore due to wave action, are assessed, including long-term coastal erosion8 and following current Dutch practice (van Koningsveld and damage from inundation, floods, and storms.9 The others 2008). Shore nourishment is substantially following impacts are evaluated in the EACC study cheaper than beach nourishment, but the benefits are (with units in parenthesis): not felt immediately. The way these options are applied is discussed in more detail in Section 3.3. For a more detailed account of the erosion impact and adaptation · Land loss due to erosion (km 2/yr) methods see Nicholls and others (in prep). · Land loss due to submergence (km 2/yr) · Forced migration (thousands/year) Inundation and flooding of the coastal zone caused by · People actually flooded (thousands/year) mean SLR and associated storm surges is assessed for · Land loss costs (million dollars/year) both sea and river floods. Large parts of the coastal zone · Forced migration costs (million dollars/year) are already threatened by extreme sea levels produced · Sea-flood costs (million dollars/year) during storms, such as shown by Hurricane Katrina in · River flood costs (million dollars/year) 2005 and Cyclone Nagris in Myanmar in 2008 (Nicholls For long-term coastal erosion due to sea-level rise, the in press). Extreme sea-level events produced by a combi- impacts of both direct and indirect effects are assessed. nation of storm surges and astronomical tides will be The direct effect of sea-level rise on coastal erosion is raised by mean sea level: the return period of extreme sea estimated using the Bruun Rule (e.g., Zhang and others levels is reduced by higher mean sea levels (e.g., Haigh 2004; Nicholls in press). Sea-level rise also affects coastal and others in press). The magnitude of this effect erosion indirectly as tidal basins become sediment sinks depends on the slope of the exceedance curve. Sea-level under rising sea level, trapping sediments from the nearby open coast into tidal basins. This indirect erosion 8 Only erosion due to sea-level rise is considered. Short-term erosion due is calculated using a simplified version of the Aggregated to individual storms when the beach is expected to largely recover is not Scale Morphological Interaction between a Tidal basin considered. Autonomous adaptation is not relevant to considerations of beach erosion. and the Adjacent coast model (Stive and others 1998; 9 Extreme events are an explicit part of this analysis. They are considered Van Goor and others 2003). About 200 tidal basins via the return periods of extreme events as explained below. around the world are considered in DIVA. Autonomous adaptation is not relevant to considerations of flooding. 16 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e rise also raises water levels in the coastal parts of rivers river depth width and slope. For a more detailed presen- (via the backwater effect), increasing the probability of tation of the flooding model, see Tol (2006) and Tol and extreme water levels. DIVA considers both these flooding others (in prep). mechanisms. Due to the difficulties of predicting changes in storm surge characteristics (e.g., von Storch and Woth DIVA translates these physical changes into social and 2008) in the standard DIVA method, the present storm economic consequences. Social consequences are surge characteristics are simply displaced upward with expressed in terms of various indicators. The indicator the rising sea level, following twentieth century observa- "people actually flooded" gives the expected number of tions (e.g., Zhang and others 2000; Woodworth and people subject to annual flooding. The indicator "forced Blackman 2004; Haigh and others in press). In the migration" gives the number of people who have to EACC Project, increased extreme water levels due to the migrate from the dry land permanently lost due to possibility of more-intense tropical cyclones are consid- erosion and areas submerged by sea level. For inunda- ered. We assume that the 100 year event increases by 10 tion it is assumed that all areas subject to flooding more percent in 2100 due to climate change, and to simplify often than once per year are abandoned by people. For the analysis we assume a linear change with time. Based the base calculation of these population numbers (in on this assumption, we rotate the existing exceedance 1995), the Gridded Population of the World dataset, curve upwards by 1 percent per decade. This method version 3 was used (CIESIN and CIAT, 2004). means that the lower the probability of the event, the greater the increase in water level. This differs from the The economic consequences are expressed in terms of effect of sea-level rise, which is uniform. damage costs (and adaptation costs as outlined in Section 3.3). The cost of (dry) land loss is estimated DIVA assumes the construction and upgrade of dikes as based on the land use scenarios and the assumption that the adaptation option for inundation and flooding, only agricultural land is lost. Agricultural land has the drawing on the experience of Deltares, including its lowest value, and it is assumed that if land used for application in the global analysis of Hoozemans and other purposes (e.g., industry or housing) is lost, then others (1993). Since there are no empirical data on those usages would move and displace agricultural land. actual dike heights available at a global level, a demand The value of agricultural land is a function of income for safety is computed and assumed to be provided by density. The cost of floods is calculated as the expected dikes (Tol 2006; Tol and Yohe 2007). DIVA is not able value of damage caused by sea and river floods based on to apply benefit-cost analysis as it was too computation- a damage function logistic in flood depth. The costs of ally expensive. Hence, dikes consistent with the demand migration are calculated on the basis of loss of GDP per for safety are applied based on population density. There capita. For a more detailed account of the valuation of are no dikes where there is very low population density impacts, see Tol (2006) and Tol and others (in prep). (< 1 person/km2). Above this population threshold, an increasing proportion of the demand for safety is 3 . 3 h o w Co s t s o f a d a p tat i o n a r e provided. Half of the demand for safety is applied at a d e f i n e d a n d C al C ul at e d population density of 20 persons/km2, and 90 percent at a population density of 200 persons/km2. This is akin to We are addressing the adaptation options defined in providing isolated dikes around individual settlements at Table 7. This includes land use planning where we limit lower population densities and more-continuous dikes at the coastal population to current levels to illustrate what higher population densities. Based on the selected dikes, an extreme land use planning policy might accomplish. land elevations, and relative sea level (including more- All these results are developed with the global DIVA extreme sea levels if appropriate), the frequency of flood- model of impacts and adaptation to sea-level rise, except ing is estimated over time. This is further converted into for the costs of port upgrade and dike maintenance, flooded people and economic flood damages based on which are developed offline in new extensions of the population density and GDP (see below). River flooding DIVA method. is evaluated in a similar fashion along approximately 115 major global rivers. The distance that requires dike is Beaches and shores are nourished according to a cost- determined by the backwater effect, which relates to the benefit analysis that balances costs and benefits (in terms D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 17 table 7. sea-level rise effeCts, impaCts, and adaptation options Considered Sea-level rise effect Impacts considered Adaptation Response considered (Long-term) beach erosion Land loss and its costs; forced migration Beach/shore nourishment and its costs Land use planning Increased flooding due to storm Expected flood damage costs; expected Sea and river dikes, including maintenance surges and the backwater effect people flooded Port upgrade (raising elevation) Land use planning Submergence Land loss and its costs; forced migration Sea and river dikes, including maintenance and its costs Land use planning of avoided damages) of protection. Shore nourishment subsidence. The unit costs for dikes are provided by has a lower unit cost than beach nourishment, but it is Hoozemans and others (1993). It is assumed that river not widely practiced at present and has the disadvantage dikes are on average half the cost of sea dikes, and the of not immediately maintaining the intertidal beach. distance inland that they need to be constructed is Beach nourishment is therefore chosen as the better determined by the backwater effect, which includes the adaptation option, but only if the tourism revenue is effect of sea-level rise. This provides an estimate of the sufficient to justify the additional costs. Tourism reve- annual capital investment. It also develops a stock of nues are derived from the Hamburg Tourism Model dikes that require maintenance and operation, but the (HTM), an econometric model of tourism flows standard DIVA does not consider these costs. (Hamilton and others, 2005a; 2005b). In HTM, tourism numbers increase with population and income. Rising We have made three extensions to the DIVA method temperatures pushes tourists toward the poles and the for the EACC research compared with Nicholls (2007): tops of mountains in search of the optimum tempera- land use planning, dike maintenance and operation tures. Hence, there is a change in the spatial pattern of costs, and port upgrade. These are outlined in Sections tourism. However, while some present tourist hotspots 3.3.1 to 3.3.3. such as the Mediterranean countries might see their market share fall as a result of climate change, there is a 3 . 3 . 1 l an d u s e p l a n n i n g significant increase in absolute tourism numbers driven by the population and GDP scenarios. Land use planning to limit growth in vulnerable coastal areas can be simulated simply by holding population For adaptation to flooding/inundation, the changing constant over time. The individual wealth would still demand function for safety is computed over time (Tol follow the GDP scenario. It is almost inconceivable that 2006; Tol and Yohe 2007). This increases with per capita we could achieve such a population trajectory based on income and population density and decreases with the current trends, even with stringent and persistent costs of dike building. As with the initial case outlined government action. Hence, this is an extreme best case in Section 3.2, dikes are not applied where there is very and is mainly illustrative of sensitivity analysis of what low population density (< 1 person/km2), and above this such a policy might accomplish. population threshold an increasing proportion of the demand for safety is applied. Half of the demand for 3 . 3 . 2 d ik e m a i n t e n a n c e a n d o p e r a t i o n safety is applied at a population density of 20 persons/ km2 and 90 percent at a population density of 200 The capital costs of building and upgrading dikes as sea persons/km2. It is assumed that any increase in demand level rises is calculated within DIVA. There are addi- for safety is provided by a new or increased dike height, tional costs required to take account of dike mainte- and the incremental costs of dike construction are deter- nance and operation throughout the lifetime of the mined. Explicit in these calculations is the assumption dike, as outlined in more detail in Appendix 2. that all existing dikes can be raised incrementally, which Operational costs reflect the costs of drainage landward is increasingly the norm due to sea-level rise and of the dike, such as drain clearance and pumping costs: 18 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e without drainage, this land would often become water- 1990 report, primary data on port areas were found to logged or flooded due to rainfall and rising water tables, be limited, and a methodology based on statistics of the combined with the lack of natural drainage. tonnage moved was developed to estimate port areas that would require raising. As primary data on port area Most of the data that were found came from the have not significantly improved, this statistical approach Netherlands (IPCC 1990; Verhagen 1998; Kok and others has been adapted here based on Lloyds List (2009) 2008), with some additional data from Canada (RSBC Ports of the World 2009 directory. This contains informa- 1996; Dike and Channel Maintenance and Habitat tion on 1,220 ports located in the regions of interest, of Subcommittee 2001). A range of estimates of maintenance which 501 in 85 countries reported usable data. Where and operational values were identified, with river dikes countries had ports recorded in Lloyds List (2009) but being consistently lower in cost, reflecting the lack of wave no usable data, assumptions based on the IPCC (1990) loadings. Maintenance costs as high as 2 percent were study were made. This increased the number of coun- identified in some cases (UNCTAD 1985; Smedema and tries included in the analysis to 108. others 2004). Taking a conservative view, maintenance and operation costs were assumed as follows: river dikes at 0.5 For the purposes of this study, no change in port area by percent and sea dikes at 1 percent. These costs could be in 2050 is included. The methodology was developed to error by a factor of 100 percent. Full details of the meth- cost the upgrading of existing areas, preserving current odology and data sources can be found in Appendix 2. risk levels for inundation; any new development is assumed to be designed for future changes in sea level Note that beach nourishment requires no consideration to 2050. The unit cost estimates are based on those used of maintenance, as these costs are built into the ones in the IPCC (1990) report, translated into current produced by DIVA. monetary value. 3.3.3 p ort u p g r a d e Full details of the methodology and data sources can be found in Appendix 2. Data from the World Bank10 show that there has been a 6 percent growth per year between 1990 and 2007 in 3 . 4 d ata (s ou rC es , a ss u m p t i o n s , a n d total global exports, a trend also shown by the volume s im p l i f iCat i o n s ) of seaborne trade, which has tripled globally over the past 30 years (UNCTAD 2008). The ability of ports to The analysis is mainly based on the DIVA database, maintain their future role in the supply chain requires which was developed specifically for the DIVA model that port infrastructure is adapted to changes in sea (McFadden and others 2007; Vafeidis and others 2008). levels. The goal of this investigation was to estimate the This is a one-dimensional database that divides the costs associated with port adaptation to sea-level rise at world's coasts (excluding Antarctica) into 12,148 linear World Bank regional levels--with adaptation being the segments and associates about 100 pieces of data with raising of existing port ground level to offset the future each segment concerning the physical, ecological, and effect of sea-level rise. Based on discussions with port socioeconomic characteristics of the coast. The operators, this is a reasonable approach that ports are segments have a variable length, with an average length likely to adopt. The estimated costs do not include of 70 km. Hence the spatial resolution is two orders of explicit cost/benefit considerations--it is assumed that magnitude higher than any other integrated assessment these strategic and valuable areas will need to be main- models that can conduct coastal analyze. As an example, tained to 2050 (and beyond). The costs that will be esti- FUND operates at national scales, so resolves approxi- mated are those associated with maintaining current mately 200 coastal units (Tol 2007). port areas in response to a total SLR projected to 2050. While some data in DIVA are stored in other forms, The methodology to estimate the costs of upgrade is such as those associated with rivers, lagoons/basins, based on that used in the IPCC (1990) report administrative units, and countries, and on a raster grid (produced by Delft Hydraulics), which identified global costs of protecting against a 1-m rise in sea level. In the 10 http://econ.worldbank.org D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 19 (Vafeidis and others 2008), the segment is the funda- The offline calculations for dike maintenance and oper- mental spatial unit of DIVA. Most calculations operate ation use the DIVA results directly, while the estimate at the segment scale, and this defines the fundamental of the costs of port upgrade is based on a range of data, resolution of the model. as explained in Appendix 2. 20 4. results Yohe (2007) will produce a higher demand for safety due to growing wealth and population density without any rise in sea level. This is consistent with changing attitudes The results are all given in 2005 U.S. dollars with no to risk during the twentieth century as living standards discounting. Appendix 3 gives the results by World rose substantially. In some locations, especially deltas, sea Bank Region for each of the scenarios that were levels would be expected to rise due to natural subsidence, considered. and this also drives adaptation needs. In DIVA, the total global adaptation costs for a scenario of no sea-level rise Globally and regionally, with the high, medium, and is estimated at from $10.4 billion/yr in the 2010s to $9.5 low sea-level rise scenarios, there is a wide range of billion/yr in the 2040s. World Bank regions account for results for all parameters considered. Protection dramat- approximately 60 percent of these costs. Two-thirds of the ically reduces land loss, the expected number of people total adaptation cost comes from sea dikes, increasing to flooded and those forced to migrate, and their associ- over 90 percent when maintenance costs are considered. ated costs compared with a scenario of no protection. Hence the focus of these results will be on the associ- Out of the World Bank regions (excluding high-income ated costs of adaptation: the construction and mainte- countries), Latin America and the Caribbean and East nance of sea and river dikes, the costs of beach nourishment, and the costs of port upgrade. In Section 4.1 (the baseline scenario) global adaptation costs are reported, and the World Bank regions that have table 8. inCremental average annual the highest costs are discussed. Section 4.2 provides Costs (2010s­2040s) of adaptation for discussion of global costs across the three sea-level rise Coastal proteCtion and residual scenarios, the adaptation costs (including ports) of the damages by sCenario under the no-rise medium SLR scenario across World Bank regions, the slr sCenario adaptation costs associated with an increase in surge billion dollars per year at 2005 prices, no discounting; heights due to more-intense tropical cyclones and with high-income countries are excluded. no population growth in the coastal zone, and a synthesis. Adaptation/Damage measures Costs 4.1 i nvestment C osts (upfron t a n d Beach nourishment 0.2 maintenanCe) in the baseline Port upgrades 0.0 s Ce nario River dikes Capital 0.1 Maintenance 0.1 Under a scenario of no climate change, DIVA still esti- Sea dikes Capital 3.6 mates that there are adaptation investment costs, most Maintenance 2.0 especially for improved dikes and to a lesser extent for Total adaptation costs 6.0 beach nourishment (Table 8). This reflects that the Total residual damage costs 8.3 demand for safety function of Tol (2006) and Tol and D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 21 Asia and the Pacific have the greatest percentage of adaptation costs at 40 percent and 30 percent, respec- figure 3. global inCremental tively. Thus on a global level, more investment is adaptation Costs for the high, medium, required in these regions compared with other World and low slr sCenarios Bank regions regardless of climate change. In contrast, 100 the Europe and Central Asia region and the Middle (billions US$/year at 2005 prices) 90 East and North Africa region have the lowest adapta- 80 Total adaptation costs 70 tion costs, at approximately 5 percent of the World 60 Bank total. Thus less investment would be required in 50 40 these two regions if sea levels do not rise. 30 20 10 However, DIVA is only designed to examine climate 0 change impacts, and there will be adaptation costs in 2010 2015 2020 2025 2030 2035 2040 2045 2050 Year coastal areas that are not linked to climate change. In High scenario Medium scenario Low scenario Section 1.3.2, a number of current adaptation invest- ments are listed; only part of the investment is linked to source: authors' data. climate change, as opposed to climate variability, which has been a major driver of coastal investment. The incremental costs of adapting to climate variability post- decreasing to 61 percent by the 2040s as the costs of 2010 are included in the DIVA costs as investments in beach nourishment grow with time. dikes, but the maintenance costs of the dikes built before 2010 are not included, and this cost would be The distribution of the adaptation costs can be seen in substantial. Investments to address the adaptation defi- Table 9, which summarizes the costs for the World cit discussed in Section 5.1 are not considered, and Bank Regions. investments due to non-climatic problems such as subsiding cities and deltas are not considered either. Sea dikes are the dominant adaptation costs, followed Hence, the DIVA figures are minimum estimates of by beach nourishment, while river dikes and port future adaptation without climate change and should upgrade are relatively minor costs. In terms of capital not be over interpreted. and maintenance costs, the latter grow rapidly with time as the stock of dikes to maintain increases. These main- 4.2 investment Costs (upfront tenance costs only consider the maintenance of dikes a n d m a i n t e n a nC e ) d u e to C l i m at e required to adapt to climate change; substantial addi- Change tional investment in maintenance would be required to maintain the overall dike system. 4.2.1 h igh, m e d i u m , a n d l o w s c e n a r i o s : g lobal adap t a t i o n c o s t s 4 . 2 . 2 m ed i u m s c e n a r i o : ad a p t a t i o n c o s t s in wor l d b a n k r e g i o n s Total incremental global adaptation costs are shown in Figure 3 for sea and river dike construction and mainte- The capital costs for sea dikes under the medium nance costs, plus beach nourishment. The Figure illus- scenario (from the 2010s through to the 2040s) due to trates that adaptation costs increase linearly with time, climate change are as follows: with the medium scenario increasing in total adaptation cost from $43.4 billion/yr in the 2010s to $59.5 billion/ East Asia and Pacific $6.4 billion/year yr in the 2040s. Europe and Central Asia $1.9 billion/year Latin America and the Caribbean $7.1 billion/year Sea dike costs are the main contributor to the global Middle East & North Africa $0.8 billion/year adaptation costs, accounting for $36 billion/yr from the South Asia $1.2 billion/year 2010s to the 2040s under the medium scenario. This Sub-Saharan Africa $2.5 billion/year accounts for 82 percent of the total costs in the 2010s, TOTAL $20.0 billion/year 22 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e table 9. inCremental average annual Costs (2010s­2040s) of adaptation for Coastal proteCtion and residual damages by sCenario billion dollars per year at 2005 prices, no discounting; high-income countries are excluded. Sea-Level Rise Scenarios High SLR with High SLR with no Costs and damages Low SLR Medium SLR High SLR cyclones population growth adaptation costs: Beach nourishment 1.7 3.3 4.5 4.5 4.5 Port upgrades 0.2 0.4 0.5 0.5 0.5 River dikes Capital 0.2 0.4 0.6 0.6 0.6 Maintenance 0.0 0.0 0.1 0.1 0.1 Sea dikes Capital 8.7 20.0 29.9 31.8 30.0 Maintenance 2.2 4.9 7.2 7.7 7.2 Total 13.0 29.0 42.8 45.2 42.9 Total residual damage costs: 0.7 1.5 2.1 2.1 1.5 This accounts for 55 percent of the total global costs of increasing to $0.44 billion/yr in the 2040s (Appendix dike construction. Sea dike costs are assumed to be 3). Maintenance costs increase from $0.02 billion/yr to roughly uniform over time. The regions with the high- $0.07 billion/yr over the same time period. Under a est cost are Latin America and the Caribbean, followed high SLR scenario, these costs would be anticipated to by East Asia and the Pacific Region, while Middle East increase 1.5 times, whereas for the low SLR scenario and North Africa has the lowest costs. costs would expect to decrease by one-third. Maintenance costs become more expensive as time Sea dike maintenance costs increase approximately progresses. Some 70 percent of the costs in World Bank linearly with time as the stock of dikes that require Regions occur in Latin America and the Caribbean. maintenance increases 4.2 times from the 2010s to a total of $7.9 billion/yr by the 2040s. It is important to Throughout the regions, beach nourishment is the note that this maintenance cost would continue to second highest component of total adaptation costs grow with time beyond the period of analysis--some- after sea dikes and their maintenance. The percentage thing that has not been considered in previous analyses increase in costs is 1.5 to 2.5 times (from the low to the and that has important implications for adaptation high scenario) from the 2010s to the 2040s. In absolute costs based on hard defenses. For the high scenario, sea terms for the medium scenario, the costs of beach nour- dike costs are approximately 1.5 times higher than the ishment increase from $2.2 billion/yr in the 2010s to medium scenario, whereas in the low SLR scenario, $4.6 billion/yr by the 2040s. dike costs are 2.3 times lower than in the medium scenario. Under the medium scenario (Figure 4), Latin America and the Caribbean and Sub-Saharan Africa account for In all regions, the cost of construction and maintenance 60 percent of the total cost for the World Bank regions. of river dikes is small in comparison to sea dikes at 1.4 The region with the lowest beach nourishment costs is percent of the total adaptation cost, except for Latin the Middle East and North Africa ($0.11 billion/yr in America and the Caribbean at 3.5 percent of the total the 2010s rising to $0.2 billion/yr in the 2040s). adaptation cost, as there is a large value in investment in initial defenses. Total river dike costs for the World Port upgrade costs are relatively minor, as detailed in Bank Regions in the 2010s are $0.37 billion/yr, Appendix 3. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 23 4 . 2 . 4 a da p t a t i o n c o s t s : sy n t h e s i s figure 4. beaCh nourishment Costs in the world bank regions for the In terms of the adaptation measures considered here, medium slr sCenario protection by sea dikes is the major contribution to defense costs, followed by beach nourishment. River dike 1.6 (billions US$/year at 2005 prices) 1.4 costs and port upgrade costs are relatively small in Beach nourishment coasts 1.2 comparison (Figure 5). For the World Bank Regions, sea 1.0 dike costs remain similar throughout the study period at 0.8 $19.8 billion/yr. They are 9.1 times more costly than 0.6 0.4 beach nourishment in the 2010s, decreasingly to 4.4 times 0.2 more costly in the 2040s (as there is greater investment in 0.0 beach nourishment). In the East Asia and Pacific and 2010 2015 2020 2025 2030 2035 2040 2045 2050 Europe and Central Asia Regions, sea dike costs can be Year East Asia and Pacific Europe and Central Asia South Asia up to 14.1 times more costly than beach nourishment. Middle East and Latin America and Sub-Saharan North Africa the Caribbean Africa In the 2010s, sea dike costs in Europe and Central Asia source: authors' data. account for 86 percent of the total adaptation cost, decreasing to 63 percent by the 2040s. In Sub-Saharan Africa they are responsible for 73 percent of total adap- tation cost in the 2010s and decrease to 53 percent in 4.2.3 e ffect o f c y c l o n e a c t i v i t y a n d o f n o the 2040s. This is because beach nourishment costs in population g r o w t h i n t h e c o a s t a l z o n e o n both regions increase over time, particularly in adaptation c o s t s Sub-Saharan Africa, where they increase from 19 percent to 26 percent of the total adaptation cost. As The effect of a 10 percent increase in tropical cyclones sea dikes require maintenance, the regions that have an during the twenty-first century is relatively small compared initial high sea dike cost (such as Europe and Central with the high SLR scenario. Globally, an increase in tropi- Asia) also have a high maintenance cost as time cal cyclones increases sea dike costs by 8 percent in the progresses. Therefore constructing sea dikes demands 2010s and by 9 percent in the 2040s compared with no continued investment in the future. tropical cyclones. The effects are felt most in the East Asia and Pacific Region and the South Asia Region, where dike costs could be rise by 13 percent compared to the high scenario in the 2040s. Consequently, dike maintenance figure 5. perCentage of adaptation costs also increase at a slightly higher rate. Costs from sea dikes, river dikes, maintenanCe Costs, and beaCh The absolute increase in global adaptation costs due to nourishment for six world bank cyclones becomes greater with time, from $5.1 billion/yr in the 2010s to $7.3 billion/yr in the 2040s. regions 100 adaptation costs (%) 80 Percentage of total Limiting population growth in the coastal zone has 60 only a small effect on adaptation costs (around 1 40 percent). This shows that the existing coastal develop- 20 ment leaves a large legacy in terms of the demand for 0 coastal protection. In practice, adaptation costs would be 2015 2025 2035 2045 reduced more than calculated here, but the DIVA data- Year base cannot resolve such details. Sea dike costs Maintenance of sea dike costs River dike costs Maintenance of sea dike costs Beach nourishment costs Port Upgrade In addition to adaptation costs, damage costs are reduced source: authors' data. as fewer people and assets are present to be affected. 24 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e With a 10 percent increase in surge levels due to has minimal impact on adaptation costs, as a large cyclones by 2100, sea dike costs increase by 8­9 percent, amount of wealth is still generated in that area. while limiting population growth in the coastal zone 25 5. limitations just one coastal city. As the adaptation costs for the medium SLR scenario (in the 2040s) were only esti- mated to be about $33 billion/yr for World Bank A study of this type inevitably has a number of limita- Regions alone (and $60 billion/yr globally, to include tions, as already indicated in the text. This section the United States), this single estimate suggests that the reflects on the limitations and the next steps to inves- adaptation deficit could exceed the incremental costs of tigate adaptation in coastal areas under climate sea-level rise estimated here. This issue clearly requires change. further more comprehensive assessment. This demon- strates that the issues of development and the cost of 5.1 t reat ment of adaptation/ successful adaptation to climate change are intimately development defiCit linked. As with all previous studies, these results only consider 5 . 2 t re at m e n t o f e x t r e m e e v e n t s the incremental cost of climate change as defined in the UNFCCC. In other words, they assume that there is a Extreme events are an explicit part of the DIVA analy- good existing adaptation system to upgrade for climate sis. The flood analysis explicitly considers extreme change. This gap between actual and desired adaptation water level events and how they might change due to systems has been termed the adaptation deficit (Burton sea-level rise--and for the first time due to increasing 2004; Parry and others 2009). If there is an adaptation intensity of tropical cyclones. However, the treatment deficit, the total cost of adaptation to achieve the resid- of stronger storms is only based on a sensitivity analy- ual impacts presented in this study will be more costly sis. It should be recognized that coastal storms cause than reported here. damage by multiple mechanisms, and wind damage, which can be widespread during storm landfall, is not The magnitude of the adaptation deficit for coastal considered here. areas has not been systematically investigated, so it is difficult to quantify. Experience shows that current There is also always a residual risk for infrastructure adaptation systems are often inadequate even in the behind defenses, and hence in the future we should industrial world (e.g., Hurricane Katrina in the United expect occasional coastal disasters, even if we protect in States in 2005 and Storm Xynthia in France in 2010), an efficient manner. Of course, this is a product of and analysis points to other cases (Nicholls and others, climate variability and hence this would be true with- 2008a), while in developing countries this is a much out climate change--the rise in the mean sea level and bigger, although poorly quantified problem. possibly more-intense storms will exacerbate this issue and mean that when floods occur they will be deeper, There are limited cost estimates, which may suggest the the flow will be faster, and they will be more likely to order of this adaptation deficit. An example is the $50 cause significant infrastructure and loss of life. billion price tag to upgrade New Orleans for Category Infrastructure losses can be reduced by flood-proofing, 5 hurricanes after the Katrina disaster: given a 30-year while loss of life can be minimized by good warning period of upgrade, this translates into $2 billion/yr for and evacuation systems. They can also be reduced by 26 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e land use planning to encourage development away from 5 . 5 t re at m e n t o f " s o f t " a d a p tat i o n vulnerable areas. Hence, even in city areas a portfolio of measures adaptation measures is likely to be optimum in response to sea-level rise rather than just depending on Delivering the "hard" adaptation measures considered in defenses. this analysis will require significant institutional and technical capacity in terms of coastal management insti- 5.3 t reat ment of teCh nologi Ca l tutions and coastal engineering expertise. In monetary Ch ange terms, these institutional costs are relatively small compared with the adaptation measures. However, It is generally assumed that all the protection technolo- money alone is not enough to develop the required gies considered in the protection analysis using DIVA capacity, and it requires enhancements across the wider are mature (cf. Tol 2007), and hence technological issue of adaptive capacity (Smit and others 2001). This change is unlikely to have much influence on the adap- is an issue that should be addressed within the develop- tation costs considered here. Important innovations are ment agenda, as the issues are broader than adapting to likely in adaptation technology, especially concerning climate change. information technology aspects such as storm forecasts and warnings. It is difficult to forecast how these might Protection will tend to degrade coastal ecosystems via change adaptation approaches in coastal areas and hence coastal squeeze, so this adaptation approach produces this is not considered. secondary impacts that are evaluated in some cases. Hence, protection does not preserve the status quo, but 5.4 t reat ment of inter-tempor a l it does preserve the valuable dryland. Coastal ecosystem Ch oi Ce degradation will be significant due to sea-level rise, and this will be reinforced by protection based on dikes. Inter-temporal choice raises the question of the role of Nicholls and Klein (2005) identified the twin challenges reactive versus proactive adaptation. The methods used of maintaining human safety and sustaining coastal in this analysis reflect these choices. Beach nourishment ecosystems as a major challenge in a European context. and port upgrade are shorter-term decisions where This is also true more widely and constitutes a major investment can be linked more closely to need, while challenge for coastal management. flood management has a longer lead time. Hence, a more reactive approach was adopted for the first two 5 . 6 t re at m e n t o f Cro s s - s eC to r adaptation measures, and a more proactive approach for measures the flood protection via dikes. In the United Kingdom and the Netherlands, both land use planning and flood The coastal system is cross-sectoral by definition. In this defense design is thinking 100 years (rather than 50 report, where we can identify significant human activi- years) into the future (e.g., DEFRA 2006; ties, they have been protected from the threats raised by Deltacommissie 2008; Kabat and others 2009; sea-level rise and increased storms. Some human Environment Agency 2009). This planning is based on impacts have not been considered, however, most rele- a precautionary SLR scenario on the order of a 1-m rise vantly changes in water supply due to salinization through the twenty-first century and consideration of (Ward and others 2010). There are also natural system much larger sea-level rise over the twenty-second impacts such as coastal wetland loss and change. These century. But these countries are exceptions, and in most natural systems are certainly valued in richer societies, locations in Europe and globally there is no proactive as illustrated by the EU Habitats Directive. adaptation to climate change, even in other industrial countries (Tol and others 2008; Moser and Tribbia In addition, coastal areas experience significant changes 2008). Hence under current behavior there is much less due to non-climate factors (Nicholls and others 2007a, proactive adaptation than considered in the EACC 2009). A major example is subsidence and failure of analysis presented here. More proactive preparation for sediment supply in many deltas, which may locally have the effects of climate change in coastal areas should be consequences as large as climate change, if not larger encouraged and supported. (e.g., Ericson and others 2006; Woodroffe and others D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 27 2006; Syvitski and others 2009). This stresses the need · Better understanding and accounting of local fac- for integrated responses to climate change that address tors that influence coastal impacts and adaptation all the issues facing coastal zones, including climate and needs, such as subsiding deltas and cities non-climate drivers (Klein and others 2001). · Improved integrated cross-sectoral assessment · Improved methods to recognize and prioritize 5.7 a r eas for follow-up work a n d potential coastal adaptation options researC h advanC es · Improved understanding of the best way to adapt to sea-level rise (e.g., soft versus hard options, retreat/ · The work raises a number of issues for further realignment) to minimize other impacts research as follows: · National adaptation case studies with more local · Improved understanding the adaptation deficit for details to reinforce this global analysis. coastal areas · Better predictive capacity of the future--for coastal changes associated with both climate and non-cli- mate drivers 28 appendix 1. eaCC population and measures of GDP per person in real terms: GDP in dollars at 2000 constant prices and market exchange gdp projeCtions rates and GDP at purchasing power parity (PPP) in dollars at 2005 prices. Some sectors, such as infrastruc- a 1.1 po pulation p roje Ctions ture, use GDP per person at PPP as the income vari- able, as this is standard in work on cross-country The choice of projections of population out to 2050 comparisons. However, others, in particular the presents few difficulties. The UN Population Division IMPACT model by the International Food Policy (UNPD) publishes updated population projections to Research Institute used for the analysis of agriculture, 2050 every two years using four sets of assumptions are calibrated to country incomes measured in dollars at about future fertility: constant fertility (CF), based on market exchange rates, and they therefore use GDP at fertility rates at the date of the projection; low fertility constant prices and market exchange rates. (L), reflecting an assumption that fertility rates will either fall rapidly (in developing countries) or remain Most of the economic models used to project future low (in industrial countries); medium fertility (M); and emissions and analyze the impacts of climate change high fertility (H). The CF projections largely provide a rely on economic projections in terms of GDP per reference point for measuring the impact of the antici- person--the International Energy Agency (IEA), the pated profile of future fertility on world/national popu- Energy Information Administration (EIA) of the U.S. lations. UNPD consistently uses the M projections as Department of Energy, Hope, and Tol. Nordhaus's its main projections, while the L and F projections RICE model, on the other hand, seems to be consistent provide a range for plausible outcomes. Further, the M PPP incomes rather than market exchange rate figures. projections are used as the foundation for other UNPD The SRES aggregates and the downscaled estimates forecasts, such as levels of urbanization, which the study from CIESIN use projections starting from 1990 are also uses. now very out of date and are therefore not used. Thus, the starting point for the projections is the World The most recent version of the UNPD projections is Bank's estimates of GDP per person in 2005 at PPP the 2008 Revision published in March 2009. Derived and 2005 prices. Aggregate GDP at PPP is calculated projections for urbanization and other variables are, by multiplying total population and is then projected at however, still based upon the 2006 Revision published five-year intervals using the real GDP growth rates for in March 2007. For this reason, the EACC Project used the country/region in each of the economic models. the 2006 Revision population projections together with Note that each model relies on different definitions of the associated derived figures (UNPD 2006). regions, so that, for example, India is treated as a sepa- rate projection unit by the IEA but is included in South a 1.2 gdp p roje Ctions Asia by Tol. Thus, it is necessary to map regions to countries separately for each model. One can get back There are different requirements for GDP projections to the implied PPP level of GDP per person by divid- in the various elements of the global analysis. The World ing aggregate GDP for the country by its projected Development Indicators (WDI) provides two main population. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 29 Two minor points concern the sample of countries and Table A1.2, on the other hand, shows the results of the treatment of missing values: applying the growth rates implied by the downscaled SRES scenarios to the 2005 starting point used for the · The full set of countries included in the WDI data- economic models. Note that there is a very small base includes many small countries and territories. discrepancy in the aggregates for 2005 because the Often the data available for such countries are very downscaled SRES projections exclude some small coun- limited. For this reason, the sample of countries tries that are included in the sample used for the analy- included in the analysis excludes all countries and ter- sis. With the exception of the A2 scenario, the SRES ritories with populations of less than 400,000 in 2005. projections show much higher rates of growth of aggre- · Data on key variables are also missing for a number gate GDP than the average of the economic models. of larger countries. For example, the WDI database The A2 scenario is quite similar to the economic aver- does not have information on either population or age, while the B2 scenario is close to the highest of the income per capita (on any basis) for Afghanistan, economic models (EIA). At the other end of the scale, North Korea, or Iraq. Since it would be very unde- the A1 scenario shows total GDP in 2050 that is more sirable to exclude such countries, the WDI data than double the economic average. have been supplemented with information from a variety of sources, including the United Nations, To run the IMPACT model, it is necessary to convert the EIA, and non-official sources. the PPP estimates of GDP per person into GDP at market exchange rates and constant 2000 prices. A The various models diverge significantly in their projec- standard method of imputing PPP values for countries tions of world GDP at PPP. Table A1.1 shows the for which the necessary price comparison data have aggregate GDP projections (in billion dollars at 2005 not been collected is to regress calculated values PPP) for the five economic models as well as the log(GDP per person at PPP) for the International equally weighted average of these for each projection Comparisons Project countries on the equivalent period. The rankings are not entirely consistent, but values of log (GDP per person at market exchange broadly the EIA generates the highest figures while rates) and use the resulting equation for imputation Nordhaus's model generates the lowest. Overall, the purposes. It is perfectly straightforward to reverse that average is closest to the figures generated by Hope's process--that is, to use log (GDP per person at model. In the absence of good reasons to give some market exchange rates and constant 2000 prices) as projections more weight than others, it seems reasonable the dependent variable and log (GDP per person at to use the equally weighted average. PPP) as the primary independent variable. There is table a1.1. aggregate gdp projeCtions from eConomiCs models in billion dollars at 2005 ppp Tol Hope Nordhaus IEA EIA Average 2005 55,303 55,303 55,303 55,303 55,303 55,303 2010 64,401 63,213 61,772 67,295 69,199 65,176 2015 73,893 72,140 69,226 83,109 83,422 76,358 2020 84,949 82,683 76,025 94,981 98,769 87,481 2025 96,362 96,514 83,712 108,915 115,915 100,284 2030 109,490 113,125 92,457 125,330 135,357 115,152 2035 123,085 133,139 102,387 144,743 158,474 132,366 2040 138,604 157,325 113,268 167,793 186,013 152,601 2045 154,492 172,379 125,567 195,268 218,884 173,318 2050 172,519 189,024 136,221 228,144 258,190 196,820 30 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e table a1.2. aggregate gdp projeCtions from sres sCenarios in billion dollars at 2005 ppp SRES A1 SRES A2 SRES B1 SRES B2 Average 2005 55,294 55,294 55,294 55,294 55,294 2010 69,225 63,641 67,413 66,189 65,176 2015 91,142 72,993 82,557 77,802 76,358 2020 113,058 84,196 101,860 92,474 87,481 2025 154,133 97,403 124,837 109,106 100,284 2030 195,208 115,475 153,185 130,138 115,152 2035 252,194 133,548 186,518 153,340 132,366 2040 309,180 151,620 226,901 182,486 152,601 2045 366,166 169,692 271,825 212,840 173,318 2050 423,152 187,764 323,069 250,066 196,820 evidence that the relationship is not linear (see Deaton PPP). However, one modification to a simple linear and Heston 2009), so the fitted equation includes regression was adopted. We have data on the depen- linear and quadratic terms. It is sometimes argued that dent and independent variables for a large sample of other variables may influence the relationship between countries from 1980 to 2005, though with some miss- PPP and market exchange rates, but these can only be ing values. Thus, it is possible to use panel methods of used in this context if either the variables are constant estimation, which can take account of systematic over time or it is possible to obtain independent differences across countries over time. This approach projections of the values of the variables up to 2050. produces much better econometric results than stan- None of the possible candidates--country size, popu- dard ordinary least squares and has been used to lation, etc.--yielded a significant coefficient, so the impute the projected values of GDP per person at model is a simple quadratic in log(GDP per person at market exchange rates. 31 appendix 2. adaptation the length of the dike type to give maintenance cost in guilders/yr. improvements Percentage of maintenance cost per year = a 2.1 di ke maintenanCe and op e r at i o n Maintenance cost x 100 The capital costs of building and upgrading dikes as sea Total cost level rises is calculated within DIVA. Additional expen- ditures are required for maintenance and operation throughout the lifetime of the dike. Operational costs Hence based on this Dutch data, the maintenance costs reflect the costs of drainage landward of the dike, such are estimated as 0.3 percent of the initial construction as drain clearance and pumping costs. costs for river dikes and as 1.1 percent for sea dikes. Some of the best cost estimates are available from the There are other sources of these numbers, again mainly Netherlands. Verhagen (1998) estimated that the from the Netherlands. Delft Hydraulics (1990) esti- 3,600 km Dutch primary dike system was worth 26 mated that a 1-m high sea dike with regular mainte- billion guilders (at 1991 values). This cost was assumed nance will be maintained at 50 percent of the to cover the cost of a total dike rebuild. Unit mainte- construction cost. UNCTAD (1985) estimates that nance costs per km were 25,000 guilders/yr for river breakwaters have a 50-year design life, so combining dikes and 85,000 guilders/yr for sea dikes. The higher these values suggests maintenance costs are about 1 costs for sea dikes reflect the fact that they are subject percent per year. UNCTAD (1985) also estimates the to wave loading in addition to high still-water levels. maintenance cost as a percentage of the initial cost of port structures: quay steel piling with reinforced Kok and others (2008) reports the national length of river concrete deck (1 percent), reinforced concrete piles and and lake dikes as 1,441 km and the length of seawalls and deck (0.75 percent), including rock-filled embayment's delta dikes as 1,880 km. Sandy coasts (which are main- (0.75 percent) and breakwaters (2 percent). tained by beach nourishment) account for the remaining 268 km of defense length. For DIVA calculations, river Prof. Marcel Stive (Professor of Coastal Engineering, and lake dikes are jointly classed as river dikes (as they do Delft Technical University) reports that over the last 10 not take the direct force of waves), while seawalls and years Rijkswaterstaat spent approximately 250million delta dikes as jointly classified as sea dikes (as they have on 2,875 km of primary defenses. Estimating construc- to withstand the greater wave impacts). tion costs from IPCC (1990), maintenance cost equates to 1.7 percent of the capital cost. A proportional relationship was assumed between the value of dikes and the length of each dike type (approx- British Columbia's (Canada) Drainage, Ditch and Dike imately 40 percent and 60 percent for river and sea Act (RSBC 1996) states that an annual levy for drainage, dikes respectively). Maintenance cost was multiplied by ditch, and dike maintenance fund cannot exceed 32 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e 5 percent of the cost of the original works. It is antici- pated that this is an upper limit of the funding required. figure a2-1. world tonnage for goods Furthermore, the Dike and Channel Maintenance and loaded and unloaded, 1970­2007 Habitat Subcommittee (2001) of Fraser Basin in British 18,000 Columbia estimates $5.5million/year for routine mainte- 16,000 nance and an additional $4.0million/year for major river 14,000 dike repairs. Assuming construction costs from IPCC 12,000 Million tonnes (1990), maintenance cost is at 0.8 percent for minor 10,000 maintenance and 0.6 percent for major works, combining 8,000 maintenance for a total of 1.4 percent of capital costs. 6,000 4,000 For drainage projects, a rule of thumb in that operation 2,000 and maintenance costs (such as weed clearance and 0 1970 1980 1990 2000 2002 2003 2004 2006 2007 desilting) account for 2 percent of the construction costs. source: unCtad 2005, 2008. Hence, a range of estimates of maintenance and opera- tional costs were identified, with river dikes being consis- tently lower in cost, reflecting the lack of wave loadings. 1,091.5 million tons, representing an increase from 2.06 Locally, maintenance and operational costs can be as percent to 8.06 percent of world trade (UNCTAD high as 5 percent, but this is atypical, and a maximum of 2005). More recent figures, which include Hong Kong, 2 percent appears a reasonable generic maximum case for Macau, Taiwan, Mongolia, and South Korea sea dikes. Taking a conservative view, maintenance and (UNCTAD 2008), show a further increase of trade in operation costs were assumed as follows: the region to 19.63 percent of world trade (3,151.3 million tons) by 2007. · River dikes: 0.5 percent · Sea dikes: 1 percent This growth in trade has fuelled the development of significant new port areas and highlights the impor- These costs could be in error by a factor of 100 percent, tance of the world ports for current and future trade. and further investigation of these costs in more-diverse The nature of this trade necessitates the location of settings is recommended. This should include how they ports in coastal areas, which makes them particularly might be expected to evolve under a scenario of rising vulnerable to the potential impacts of climate change sea level. and sea-level rise. This has been recognized recently with the C40 Climate Leadership Group's World Ports a 2.2 po rt up grade Climate Conference held in Rotterdam in 2008 (http:// wpccrotterdam.com/) and studies such as Herberger In 2007, the volume of international seaborne trade and others (2009), which noted that the Port of Los reached 8.02 x 109 tons--with an estimated annual Angeles, which handles 40­50 percent of the containers average growth rate of 3.1 percent over the past three that enter the United States, will be subject to increas- decades--representing 80 percent of world trade ing flood risk due to sea-level rise over this century. (UNCTAD 2008). Total amounts for goods loaded and unloaded have increased steadily (Figure A2.1), with Only a limited number of previous assessments of port trade through the worlds container ports reaching 485 adaptation to climate change have been carried out at million TEUs11 in 2007, of which China is estimated to regional and global scales. This is largely because data account for approximately 28.4 percent. China has on ports are fragmented and inconsistent, and the exhibited the largest growth in seaborne trade, notably necessary physical parameters are not systematically after joining the World Trade Organization in 2001. reported. This lack of information has led to the use of Between 1990 and 2004, for example, the annual amount of goods handled for China and North Korea increased almost six times, from 167.7 million to 11 TEU is twenty-foot equivalent unit, or a 6.1 m x 2.4 m container. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 33 indicators in order to assess potential impacts and asso- · Port facilities (including which type of cargo the ciated costs. port is able to accommodate) · Traffic (tonnage and number of containers (TEUs) The aim of this work is to estimate the costs associated for a given year) with the adaptation of ports to sea-level rise at country, · Tides (e.g., tidal range/tidal levels) World Bank regional and global levels. As the main form of adaptation to sea-level change is assumed to be Traffic data were supplemented, where appropriate, raising land levels, adaptation costs are assumed to be from the Containerisation International Yearbook (Fossey only those associated with raising current port areas to 2009). maintain their elevation relative to sea level and preserv- ing current risk levels for inundation. a2.2.3 port selection criteria a 2.2.1 m eth o d o l o g y Countries included in the analysis were only those clas- sified as upper middle income, lower middle income, or The methodology adopted for this study is based on low income in the World Bank list of economies. For that developed by Delft Hydraulics (now Deltares) in China, this meant ports located in Taiwan were its report Sea-Level Rise: A World-wide Cost Estimate of included, but those in Hong Kong and Macau were Basic Coastal Defence Measures (IPCC 1990). This report excluded. on global costs of protecting against a 1-m rise in sea level used statistics of maritime tonnage at country level Individual port data for the selected countries were to estimate the port area that would require ground recorded on a spread sheet and then transferred into levels to be raised. GIS software (ArcGIS), which showed the geographic position of individual ports. This allowed easy classifi- In this study, due to the lack of comprehensive data at cation of the ports as either river, coastal, or offshore. country level, port-level information was sourced that This excluded from the analysis ports located inland on could then be aggregated to country level. In addition, major rivers, such as the Yangtze in China, which may estimated costs are calculated in response to the sea- be subject to changes in future water levels due to the level rise scenarios to 2050 (Section 3.1). The costs for "backwater" effect as river waters meet rising sea levels raising port areas by 1-m are scaled to determine unit but which are not considered a direct influence of sea- costs for other increments of sea-level rise. It is assumed level change in this analysis. To allow the inclusion in that all port areas need to be raised, and no formal cost/ the analysis of ports located in deltas or estuaries that benefit analysis is conducted; it is assumed that these are subject to the direct effects of sea-level change, an strategic and valuable areas will need to be maintained "up river" limit was established for those classified as to 2050 (and beyond). Only existing port areas are coastal using the description of tidal influence in the considered and they are assumed to be flat. Any new original data. Ports upstream of this limit were developments between 2010 and 2050 are presumed to excluded. Ports located on the Caspian Sea were also be constructed with an appropriate allowance for sea- excluded, as change in global sea level will not have level rise. direct impacts. a 2.2.2 s o ur c e d a t a a 2 . 2 . 4 tr a ff i c t o a r e a c a l c u l a t i o n s The most comprehensive data source on ports was The methodology for translating reported traffic used found to be Lloyds List's Ports of the World 2009 direc- values from the IPCC (1990) report. Based on informa- tory. This provides port data listed by country, tion from the Port of Rotterdam, the IPCC determined including: tonnage-space ratios for a range of cargo types and an amalgamated ratio that represents the quays, storage · Port name areas, roads, general areas (offices, etc.), and industrial · Port activity (if the port is still commercially active) areas within the port area. In this report three ratios · Port location (degrees and minutes) were used: 34 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e · Mixed cargo: 3* 10 6 tons handled per and was translated into area (km2) using the following km 2 of port area equation: · Bulk/oil: 30 * 10 6 tons handled per km 2 of port area Y = (bbl/day * 50)/ 30 * 106 Eq.2 · Containers (TEUs): 16 * 10 6 tons handled Y= area (km2) per km 2 of port area bbl/day = Number of barrels per day as reported by the CIA World Factbook13 a 2.2.5 tonn a g e a n d c o n t a i n e r s 50 = conversion factor of 1 barrel a day is 50 tons a year14 Two categories of traffic are recorded in Lloyds List 30 * 106 = Tonnage equivalent for bulk liquids (2009): tonnage and TEUs. The tonnage and TEUs and petroleum that can be handled in handled reported in the Lloyds List are assumed to 1km2 (IPCC 1990) represent the current capacity of the port, although it is recognized that this may change over time. Potential a2.2.7 amalgamating data changes in traffic to 2050, such as a potential decrease in oil transport, are not considered. Areas were derived for individual port areas by adding the tonnage and TEU area equivalent values calculated Separate area calculations were undertaken for the two using the method described in Section A2.2. These were values. Tonnage was transferred directly onto the then aggregated to country level and added to the area spreadsheet, and if a range of port facilities was indi- equivalent for oil exports to provide a total area estimate cated (e.g., general cargo, dry bulk, Ro/Ro, liquid bulk) per country. Small island counties with no area calculated that was divided by the mixed cargo rate above to by these methods but that had listed ports were given a generate area for the port. nominal area of 0.1 km2; larger countries were assigned the areas from the IPCC (1990) report to represent a As TEU is a volume based unit, translation into other minimum area. Country-level data were then summed units is necessarily imprecise. With reference to the according to the World Bank geographic regions. design and specification of containers, the most common dry cargo maximum gross weight is approxi- a2.2.8 Costs of upgrade mately 24tonnes.12 This value was therefore used in the equation below to determine the area required for the The cost of the upgrade to port ground levels is based number of reported TEUs: on that reported in IPCC (1990) of $15 million per km2 to raise ground levels by 1m. This was based on Y = (N *24)/(16* 106) Eq.1 Dutch procedures including design, execution, taxes, Y= area (km 2) levies and fees and the assumption that the operation N= Number of TEUs handled would take place as one event. It is likely that these 24 = tonnage equivalent for a container costs are overestimated for some countries and underes- 16* 106 = Tonnage equivalent for containers which timated for others. At the scale of this investigation, can be handled in 1km2 (IPCC 1990) these discrepancies will largely balance out. To stan- dardize the results, costs were inflated from 1990 to a 2.2.6 o il a n d p e t r o l e u m p r o d u c t s 1995 using the U.S. Retail Price Inflation (Annual Average), which shows a cumulative inflation increase It was noted that ports identified in Lloyds List (2009) of 22.89 percent over this period. This inflation as being oil terminals or only having facilities for bulk liquid and petroleum did not report any traffic values. 12 See http://www.emase.co.uk/data/cont.html, http://www.freightraders. As oil forms an important aspect of maritime trade, co.nz/containerspecs.html,and http://www.bslcontainers.com/prod- data from the CIA's World Factbook website were used ucts41.php). as a basis for calculating area. These data represent the 13 https://www.cia.gov/library/publications/the-world-factbook/ total oil exported in barrels per day (bbl/day) at rankorder/2176rank.html#. national level, including both crude oil and oil products, 14 http://www.eppo.go.th/ref/UNIT-OIL.html. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 35 table a2.1. estimated regional Costs of port upgrade Costs based on elevating port area for high, medium, and low sea-level sCenarios to 2050 (high-income countries are excluded) Costs of port raising, by Sea-level rise scenario (billion dollars, at 2005 prices) Total port area Low Medium High Region (km2) (15.6cm) (28.5cm) (37.8cm) East Asia & Pacific 1,124 4.4 7.3 9.7 Latin America & 385 1.4 2.5 3.3 Caribbean Europe & Central Asia 251 0.9 1.6 2.2 South Asia 243 0.9 1.6 2.1 Middle East & North Africa 134 0.5 0.9 1.2 Sub-Saharan Africa 127 0.5 0.8 1.1 Total 2,263 8.5 14.7 19.6 increased costs from $15 million to $18.5 million per · Traffic values in Lloyds List (2009) and oil values km2 to raise ground levels by 1m, or $0.185 million per from the CIA World Factbook are reported for a sin- km2 to raise ground levels by 1 cm. gle year. These values vary annually, and the num- ber reported may not represent the actual capacity a 2.2.9 r esu l t s / di s c u s s i o n of the port. However, as global maritime trade is expected to continue to increase, any future calcula- The results shown in Table A2.1 are for the scenarios in tions would be expected to increase port area. Section 3.1.2. They show a total cost for adaptation to projected sea-level rise by 2050 of between $6,855 Comparison of port areas with the IPCC (1990) report million and $15,822 million for countries of low, lower is difficult due to changes in both the reporting and middle, and upper middle income as defined by the nature of maritime traffic. However, if China (which World Bank. The region with the largest area is East shows atypical growth) is excluded, on average port Asia and the Pacific, largely due to the rapid growth in areas are 2.6 times larger; this is commensurate with the China's maritime trade, as evidenced by rapid port increase in the amount of goods handled (see expansion in this country over recent years. Figure A1-1). On an individual country basis, estimation errors can It is important to remember that this report does not have a significant effect on adaptation costs, but the include any cost/benefit considerations. It is calculated regionalized figure gives a valid indication of the scale on maintaining current risk levels and therefore gives no of potential costs. However, the values in the table indication of vulnerability. should be regarded as minimums for several reasons: · The database contains 1,220 identified ports, of which 1,135 are located either on the coast or off- shore; only 501 of these reported either tonnage or TEU data. These data are derived largely from imports and exports, excluding port areas mostly 15 Comorros, Fiji, Kiribati, Marshall Islands, Mayotte, Micronesia, Palau, linked to domestic trade. São Tomé and Principe, Solomon Islands, St. Kitts and Nevis, St. Vincent · Port areas for 13 small island states15 and eight and the Grenadines, Tonga, and Vanuatu. mainland countries16 were approximated using val- 16 Gabon, Guinea-Bissau, Guyana, Iraq, Democratic Republic of Korea, ues from the IPCC (1990) report. Liberia, Myanmar, and Suriname. 36 appendix 3. results by world · EAP ­ East Asia and Pacific bank region (exCluding high- · ECA ­ Europe and Central Asia · LAC ­ Latin America and Caribbean inCome Countries) · MNA ­ Middle East and North Africa · SAS ­ South Asia Tables are presented for each scenario, including · SSA ­ Sub-Saharan Africa (1) no-rise SLR, (2) low SLR, (3) medium SLR, (4) high SLR, (5) high SLR with increased tropical cyclones, and (6) high SLR with constant coastal popu- lations. The following abbreviations are used for the regions: D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 37 table a3.1. inCremental annual Costs of adaptation for Coastal proteCtion and residual damages by region and deCade for the no-rise slr sCenario (billion dollars per year at 2005 prices, no discounting; high-income countries are excluded.) World Bank Regions EAP ECA LAC MNA SAS SSA Total Total adaptation costs: Beach nourishment 2010s 0.03 0.02 0.04 0.01 0.02 0.05 0.17 2020s 0.03 0.03 0.05 0.02 0.02 0.05 0.2 2030s 0.03 0.03 0.05 0.02 0.02 0.06 0.21 2040s 0.04 0.04 0.05 0.02 0.02 0.06 0.23 Port upgrades 2010s 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2020s 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2030s 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2040s 0.00 0.00 0.00 0.00 0.00 0.00 0.00 CC MC CC MC CC MC CC MC CC MC CC MC River dikes 2010s 0.07 0.01 0.00 0.00 0.08 0.08 0.00 0.00 0.00 0.00 0.00 0.00 0.24 2020s 0.02 0.01 0.00 0.00 0.07 0.08 0.00 0.00 0.00 0.00 0.00 0.00 0.18 2030s 0.02 0.01 0.00 0.00 0.06 0.09 0.00 0.00 0.00 0.00 0.00 0.00 0.18 2040s 0.02 0.01 0.00 0.00 0.05 0.09 0.00 0.00 0.00 0.00 0.00 0.00 0.17 Sea dikes 2010s 1.79 0.37 0.22 0.08 1.54 0.74 0.22 0.04 0.34 0.08 0.62 0.22 6.26 2020s 1.12 0.49 0.15 0.10 1.34 0.87 0.19 0.06 0.29 0.11 0.45 0.26 5.43 2030s 1.07 0.60 0.13 0.11 1.28 1.00 0.13 0.07 0.29 0.14 0.45 0.31 5.58 2040s 0.91 0.70 0.10 0.12 1.03 1.11 0.10 0.09 0.27 0.17 0.39 0.35 5.34 Total residual damage costs: Land loss, migration, 2010s 3.11 0.13 0.71 0.11 1.18 0.08 5.3 and sea and river flood costs 2020s 4.01 0.15 0.87 0.15 1.90 0.12 7.2 2030s 4.89 0.17 1.08 0.21 2.80 0.18 9.3 2040s 6.23 0.18 1.08 0.28 3.15 0.26 11.1 Note 38 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e table a3.2. inCremental Costs of adaptation for Coastal proteCtion and residual damages by region and deCade for the low slr sCenario (billion dollars per year at 2005 prices, no discounting; high-income countries are excluded) World Bank Regions EAP ECA LAC MNA SAS SSA Total Total adaptation costs: Beach nourishment 2010s 0.26 0.06 0.39 0.07 0.11 0.36 1.25 2020s 0.30 0.08 0.48 0.09 0.13 0.43 1.51 2030s 0.35 0.10 0.58 0.10 0.15 0.51 1.79 2040s 0.41 0.13 0.70 0.12 0.17 0.61 2.14 Port upgrades 2010s 0.11 0.03 0.02 0.02 0.01 0.01 0.21 2020s 0.11 0.03 0.02 0.02 0.01 0.01 0.21 2030s 0.11 0.03 0.02 0.02 0.01 0.01 0.21 2040s 0.11 0.03 0.02 0.02 0.01 0.01 0.21 CC MC CC MC CC MC CC MC CC MC CC MC River dikes 2010s 0.03 0.00 0.00 0.00 0.11 0.01 0.00 0.00 0.01 0.00 0.01 0.00 0.17 2020s 0.03 0.00 0.00 0.00 0.11 0.01 0.00 0.00 0.01 0.00 0.01 0.00 0.17 2030s 0.03 0.00 0.00 0.00 0.11 0.02 0.00 0.00 0.01 0.00 0.01 0.00 0.18 2040s 0.03 0.01 0.00 0.00 0.11 0.02 0.00 0.00 0.01 0.00 0.01 0.00 0.19 Sea dikes 2010s 2.79 0.29 0.89 0.09 3.02 0.31 0.34 0.04 0.53 0.05 1.04 0.11 9.50 2020s 2.79 0.57 0.86 0.18 3.04 0.61 0.38 0.07 0.53 0.11 1.05 0.21 10.40 2030s 2.80 0.84 0.85 0.26 3.07 0.92 0.37 0.11 0.53 0.16 1.07 0.32 11.30 2040s 2.80 1.12 0.84 0.35 3.09 1.23 0.37 0.15 0.53 0.21 1.09 0.43 12.21 Total residual damage costs: Land loss, migra- 2010s 0.21 0.01 0.04 0.01 0.03 0.00 0.30 tion, and sea and river flood costs 2020s 0.42 0.02 0.08 0.01 0.05 0.01 0.59 2030s 0.45 0.03 0.09 0.03 0.18 0.01 0.79 2040s 0.49 0.03 0.21 0.03 0.50 0.02 1.28 Note: 2010s=2010­19, 2120s=2020­29, 2030s=2030­39, and 2040s=2040­49; CC=Capital Cost, and mC=maintenance Cost D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 39 table a3.3. inCremental annual Costs of adaptation for Coastal proteCtion and residual damages by region and deCade for the medium slr sCenario billion dollars per year at 2005 prices, no discounting; high-income countries are excluded. World Bank Regions EAP ECA LAC MNA SAS SSA Total Total adaptation costs: Beach nourishment 2010s 0.45 0.13 0.67 0.11 0.19 0.62 2.17 2020s 0.56 0.18 0.87 0.15 0.24 0.79 2.79 2030s 0.72 0.26 1.14 0.18 0.31 1.02 3.63 2040s 0.89 0.34 1.44 0.22 0.38 1.28 4.55 Port upgrades 2010s 0.18 0.06 0.04 0.04 0.02 0.02 0.37 2020s 0.18 0.06 0.04 0.04 0.02 0.02 0.37 2030s 0.18 0.06 0.04 0.04 0.02 0.02 0.37 2040s 0.18 0.06 0.04 0.04 0.02 0.02 0.37 CC MC CC MC CC MC CC MC CC MC CC MC River dikes 2010s 0.07 0.00 0.01 0.00 0.25 0.01 0.00 0.00 0.01 0.00 0.02 0.00 0.37 2020s 0.07 0.01 0.01 0.00 0.25 0.02 0.00 0.00 0.01 0.00 0.02 0.00 0.39 2030s 0.07 0.01 0.01 0.00 0.26 0.04 0.00 0.00 0.01 0.00 0.02 0.00 0.42 2040s 0.07 0.01 0.01 0.00 0.26 0.05 0.00 0.00 0.02 0.00 0.02 0.00 0.44 Sea dikes 2010s 6.39 0.62 2.05 0.20 6.92 0.67 0.78 0.08 1.22 0.12 2.39 0.23 21.67 2020s 6.40 1.26 1.99 0.40 7.06 1.37 0.85 0.16 1.23 0.24 2.43 0.47 23.86 2030s 6.42 1.90 1.96 0.59 7.15 2.09 0.84 0.24 1.23 0.36 2.47 0.72 25.97 2040s 6.43 2.54 1.96 0.79 7.18 2.80 0.84 0.33 1.24 0.49 2.54 0.97 28.11 Total residual damage costs: Land loss, migra- 2010s 0.31 0.01 0.12 0.01 0.04 0.00 0.49 tion, and sea and river flood costs 2020s 0.60 0.03 0.14 0.02 0.10 0.02 0.91 2030s 1.11 0.05 0.23 0.04 0.28 0.03 1.74 2040s 1.34 0.08 0.45 0.07 0.94 0.04 2.92 Note: 2010s=2010­19, 2120s=2020­29, 2030s=2030­39, and 2040s=2040­49; CC=Capital Cost, and mC=maintenance Cost 40 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e table a3.4. inCremental annual Costs of adaptation for Coastal proteCtion and residual damages by region and deCade for the high slr sCenario billion dollars per year at 2005 prices, no discounting; high-income countries are excluded. World Bank Regions EAP ECA LAC MNA SAS SSA Total Total adaptation costs: Beach nourishment 2010s 0.58 0.16 0.84 0.13 0.25 0.78 2.74 2020s 0.77 0.25 1.17 0.19 0.33 1.07 3.78 2030s 1.00 0.36 1.55 0.24 0.43 1.40 4.98 2040s 1.34 0.50 2.09 0.31 0.58 1.87 6.69 Port upgrades 2010s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2020s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2030s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2040s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 CC MC CC MC CC MC CC MC CC MC CC MC River dikes 2010s 0.10 0.00 0.01 0.00 0.37 0.02 0.00 0.00 0.02 0.00 0.02 0.00 0.54 2020s 0.10 0.01 0.01 0.00 0.38 0.04 0.00 0.00 0.02 0.00 0.03 0.00 0.59 2030s 0.11 0.02 0.01 0.00 0.39 0.05 0.00 0.00 0.02 0.00 0.03 0.00 0.63 2040s 0.11 0.02 0.01 0.00 0.40 0.07 0.00 0.00 0.02 0.00 0.03 0.01 0.67 Sea dikes 2010s 9.52 0.86 3.06 0.28 10.4 0.94 1.17 0.11 1.83 0.17 3.58 0.32 32.2 2020s 9.55 1.82 2.96 0.58 10.6 1.99 1.29 0.23 1.84 0.35 3.65 0.69 35.6 2030s 9.58 2.78 2.94 0.87 10.7 3.06 1.25 0.36 1.85 0.53 3.73 1.06 38.7 2040s 9.59 3.73 2.95 1.17 10.7 4.13 1.25 0.48 1.85 0.72 3.83 1.44 41.8 Total residual damage costs: Land loss, migration, 2010s 0.37 0.01 0.14 0.01 0.05 0.01 0.59 and sea and river flood costs 2020s 0.79 0.03 0.20 0.03 0.13 0.03 1.21 2030s 1.30 0.06 0.33 0.06 0.36 0.05 2.16 2040s 1.88 0.10 0.63 0.12 1.48 0.07 4.28 Note: 2010s = 2010­19, 2120s = 2020­29, 2030 s= 2030­39, and 2040s = 2040­49; CC = Capital Cost, and mC=maintenance Cost D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 41 table a3.5. inCremental annual Costs of adaptation for Coastal proteCtion and residual damages by region and deCade for the high slr sCenario with CyClones (billion dollars per year at 2005 prices, no discounting; high-income countries are excluded.) World Bank Regions EAP ECA LAC MNA SAS SSA Total Total adaptation costs: Beach nourish- 2010s 0.58 0.16 0.84 0.13 0.25 0.78 2.74 ment 2020s 0.77 0.25 1.17 0.19 0.33 1.07 3.78 2030s 1.00 0.36 1.55 0.24 0.43 1.40 4.98 2040s 1.34 0.50 2.09 0.31 0.58 1.87 6.69 Port upgrades 2010s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2020s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2030s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2040s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 CC MC CC MC CC MC CC MC CC MC CC MC River dikes 2010s 0.11 0.01 0.01 0.00 0.37 0.02 0.00 0.00 0.02 0.00 0.03 0.00 0.57 2020s 0.12 0.01 0.01 0.00 0.38 0.04 0.00 0.00 0.03 0.00 0.03 0.00 0.62 2030s 0.12 0.02 0.01 0.00 0.39 0.05 0.00 0.00 0.03 0.00 0.03 0.00 0.65 2040s 0.13 0.02 0.01 0.00 0.40 0.07 0.00 0.00 0.03 0.01 0.03 0.01 0.71 Sea dikes 2010s 10.4 1.04 3.09 0.29 10.74 1.01 1.17 0.11 2.01 0.20 3.68 0.34 34.0 2020s 10.6 2.09 3.00 0.59 11.00 2.10 1.29 0.23 2.05 0.40 3.78 0.71 37.8 2030s 10.7 3.16 2.98 0.88 11.09 3.21 1.25 0.36 2.09 0.61 3.88 1.10 41.3 2040s 10.8 4.24 2.99 1.18 11.13 4.32 1.25 0.48 2.12 0.82 3.99 1.49 44.8 Total residual damage costs: Land loss, migra- 2010s 0.37 0.01 0.14 0.01 0.05 0.01 0.59 tion, and sea and river flood costs 2020s 0.79 0.03 0.20 0.03 0.13 0.03 1.21 2030s 1.30 0.06 0.33 0.06 0.36 0.05 2.16 2040s 1.88 0.10 0.63 0.12 1.48 0.07 4.28 42 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e table a3.6. inCremental annual Costs of adaptation for Coastal proteCtion and residual damages by region and deCade for the high slr sCenario with no population growth billion dollars per year at 2005 prices, no discounting; high-income countries are excluded. World Bank Regions EAP ECA LAC MNA SAS SSA Total Total adaptation costs: Beach nourishment 2010s 0.58 0.16 0.84 0.13 0.24 0.78 2.73 2020s 0.77 0.25 1.16 0.18 0.33 1.07 3.76 2030s 0.99 0.36 1.55 0.23 0.42 1.40 4.95 2040s 1.33 0.50 2.08 0.30 0.57 1.86 6.64 Port upgrades 2010s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2020s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2030s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 2040s 0.24 0.08 0.05 0.05 0.03 0.03 0.49 CC MC CC MC CC MC CC MC CC MC CC MC River dikes 2010s 0.09 0.00 0.02 0.00 0.40 0.01 0.00 0.00 0.02 0.00 0.02 0.00 0.56 2020s 0.10 0.01 0.03 0.00 0.38 0.04 0.00 0.00 0.02 0.00 0.03 0.00 0.61 2030s 0.11 0.01 0.02 0.00 0.39 0.06 0.00 0.00 0.02 0.00 0.03 0.00 0.64 2040s 0.13 0.02 0.03 0.00 0.40 0.08 0.00 0.00 0.02 0.00 0.03 0.01 0.72 Sea dikes 2010s 9.45 0.86 3.25 0.29 10.14 0.91 1.14 0.10 1.83 0.17 3.52 0.32 31.98 2020s 9.50 1.81 3.33 0.62 10.49 1.96 1.16 0.22 1.84 0.35 3.67 0.68 35.63 2030s 9.55 2.76 3.39 0.96 10.54 3.01 1.18 0.34 1.85 0.53 3.72 1.05 38.88 2040s 9.75 3.73 3.41 1.30 10.59 4.07 1.18 0.45 1.85 0.72 3.72 1.43 42.20 Total residual damage costs: Land loss, migra- 2010s 0.26 0.01 0.10 0.01 0.05 0.00 0.43 tion, and sea and river flood costs 2020s 0.62 0.02 0.14 0.02 0.09 0.01 0.90 2030s 1.28 0.05 0.22 0.03 0.20 0.03 1.81 2040s 1.81 0.11 0.39 0.06 0.61 0.04 3.02 note: 2010s = 2010­19, 2120s = 2020­29, 2030s = 2030­39, and 2040s = 2040­49; CC = Capital Cost, and mC = maintenance Cost D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 43 referenCes Bosello, F., Kuik, O., Tol, R., and Watkiss, P. 2007. Costs of Adaptation to Climate Change: A review of assessment studies with a focus on methodologies used. Adger, W.N., Agrawala, S., Mirza, M.M.Q., Conde, C., Final Report to the European Agency. 2 February. O'Brien, K., Pulhin, J., Pulwarty, R., Smit, B., and Takahashi, K. 2007. Assessment of adaptation practices, Brown, K., Tompkins, E.L., and Adger, N. 2002. options, constraints and capacity. In: M.L. Parry, O.F. Making Waves: Integrating Coastal Conservation and Canziani, J.P. Palutikof, P.J. van der Linden, and C.E. Development. Earthscan Publications Limited, London. Hanson, eds., Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Burton, I. 2004. Climate change and the adaptation Fourth Assessment Report of the Intergovernmental Panel deficit. In: A. Fenech, D. Maclver, H. Auld, R. Bing on Climate Change. Cambridge University Press, Rong, and Yongyuan Yin, eds., Building the Adaptive Cambridge, UK, pp. 717­43. Capacity. Meteorological Service of Canada, Toronto, pp. 25­33. Anthoff, D., Nicholls, R.J., Tol, R.J.S., and Vafeidis, A.T. 2006. Global and regional exposure to large rises in Cicin-Sain, B., and Knecht, R. 1998. Integrated Coastal sea-level: a sensitivity analysis. Tyndall Working Paper 96. and Ocean Management: Concepts and Practices. Island Tyndall Centre for Climate Change Research. Press, Washington, DC. Anthoff, D., Nicholls, R.J., and Tol, R.S.J. 2010. The CIESIN and CIAT. 2004. Gridded Population of the economic impact of substantial sea-level rise. Mitigation World (GPW), Version 3. Center for International and Adaptation Strategies for Global Change 15(4), 321­35. Earth Science Information Network, Columbia University and Centro Internacional de Agricultura Barth, M.C., and Titus, J.G., eds. 1986. Greenhouse Tropical, Palisades. Columbia University, New York. Effect and Sea Level Rise: A Challenge for This Generation. Van Nostrand Reinhold, New York. Clark, M. 1998. Flood insurance as a management strategy for UK coastal resilience. The Geographical Bijlsma, L., Ehler, C.N., Klein, R.J.T., Kulshrestha, Journal 164(3), 333­43. S.M., McLean, R.F., Mimura, N., Nicholls, R.J., Nurse, L.A., Pérez Nieto, H., Stakhiv, E.Z., Turner, R.K., and Dasgupta, S., Laplante, B., Murry, S., and Wheeler, D. Warrick, R.A. 1996. Coastal zones and small islands. In: 2009a. Sea-Level Rise and Storm Surges: A Comparative R.T. Watson, M.C. Zinyowera, and R.H. Moss, eds., Analysis of Impacts in Developing Countries. Policy Climate Change 1995: Impacts, Adaptations and Research Working Paper 4901. World Bank, Mitigation of Climate Change: Scientific-Technical Washington, DC. Analyses, Contribution of Working Group II to the Second Assessment Report of the Intergovernmental Panel on Dasgupta, S., Laplante, B., Meisner, C., Wheeler, D., Climate Change. Cambridge University Press, and Yan, J. 2009b. The impact of sea level rise on devel- Cambridge, UK, pp. 289­324. oping countries: a comparative analysis. Climatic Change 93(3), 379­88. Bindoff, N., Willebrand, J., Artale, V., Cazenave, A., Gregory, J., Gulev, S., Hanawa, K., Quéré, C.L., Levitus, Deaton, A., and Heston, A. 2009. Understanding PPPs S., Nojiri, Y., Shum, C.K., Talley, L., and Unnikrishnan, and PPP-Based National Accounts. Princeton University A. 2007. Observations: oceanic climate change and sea and University of Pennsylvania, November. At http:// level. In: S. Solomon, D. Qin, M. Manning, Z. Chen, M. www.princeton.edu/~deaton/downloads/deaton_heston_ Marquis, K.B. Averyt, M. Tignor, and H.L. Miller, eds., complete_nov10.pdf. Climate Change 2007: The Physical Science Basis. Working Group I Contribution to the Intergovernmental Panel on DEFRA. 2006. Flood and Coastal Defence Appraisal Climate Change Fourth Assessment Report. Cambridge Guidance. FCDPAG3 Economic Appraisal. University Press, Cambridge, UK, pp. 385­432. Supplementary Note to Operating Authorities ­ 44 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e Climate Change Impacts. October. Department of Hamilton, J.M., Maddison, D.J., and Tol, R.S.J. 2005b. Environment, Food and Rural Affairs (DEFRA), The effects of climate change on international tourism. London. Climate Research 29, 255­68. Deltacommissie. 2008. Working Together with Water. A Hanson, S., Nicholls, R.J., Hallegatte, S., and Corfee- Living Land Builds for Its Future. Findings of the Morlot, J. 2009. The effects of climate mitigation on the Deltacommissie. Delta Commissie, The Netherlands. exposure of the worlds large port cities to extreme coastal water levels. AVOID Programme DECC/ Dike and Channel Maintenance and Habitat DEFRA, Project Report, London. Subcommittee. 2001. Comprehensive Management for Flood Protection Works, Fraser Basin Council, British Herberger, H., Cooley, H., Herrera, P., Gleick, P.H., and Columbia. Moore, E. 2009. The impact of sea-level rise on the Californian coast. Climate Change Centre, University Environment Agency. 2009. Managing Flood Risk through of California, Berkeley. London and the Thames Estuary. Thames Estuary 2100, Strategic Environmental Assessment, Environmental Hoozemans, F.M.J., Marchand, M., and Pennekamp, Report Summary. Environment Agency, Bristol. H.A. 1993. A Global Vulnerability Analysis: Vulnerability Assessment for Population, Coastal Wetlands and Rise Ericson, J.P., Vörösmarty, C.J., Dingman, S.L., Ward, Production on a Global Scale, 2nd ed., Delft Hydraulics, L.G., and Meybeck, M. 2006. Effective sea-level rise Netherlands. and deltas: Causes of change and human dimension implications. Global and Planetary Change 50, 63­82. Hope, C. 2006. The marginal impact of CO2 from PAGE2002: an integrated assessment model incorporat- European Union. 2010. Coastal Zone Policy, Integrated ing the IPCC's five reasons for concern. The Integrated Coastal Zone Management (ICZM). Assessment Journal, Bridge and Science 6(1), 19­56. Fankhauser, S. 1995. Valuing Climate Change ­ The IPCC. 1990. Sea-level Rise: A World-wide Cost Estimate Economics of the Greenhouse, 1st ed. Earthscan, London. of Basic Coastal Defence Measures. Report prepared for Intergovernmental Panel for Climate Change and Fossey, J. ed. 2009. Containerisation International Rijkswaterstaat. Delft Hydraulics, Delft, Netherlands. Yearbook. Informa, London. IPCC. 2007. Climate Change 2007: Synthesis Report. Grinsted, A., Moore, J.C., and Jevrejeva, A. 2009. Intergovernmental Panel on Climate Change, Geneva. Reconstructing sea level from palaeo and projected temperatures 200 to 2100 AD. Climate Dynamics 34(4), Jevrejeva, S., Moore, J.C., Grinsted, A., and Woodworth, 461­72. P.L. 2008. Recent global sea level acceleration started over 200 years ago? Geophysical Research Letters 35, Grossi, P., and Muir-Wood, R. 2006. Flood Risk in New L08715. Orleans: Implication for Future Management and Insurability. Risk Management Solutions, London. Kabat, P., Fresco, L.O., Stive, M.J.F., Veerman, C.P., van Alphen, J.S.L.J., Parmet, B.W.A.H., Hazeleger, W., and Haigh, I.D., Nicholls, R.J., and Wells, N.C. In press. Katsman, C.A. 2009. Dutch coasts in transition. Nature Assessing changes in extreme sea levels: Application to Geoscience 2(7), 450­52. the English Channel, 1900­2006. Continental Shelf Research. Kay, R., and Alder, J. 2005. Coastal Planning and Management, 2nd ed. E & FN Spon, London. Hamilton, J.M., Maddison, D.J., and Tol, R.S.J. 2005a. Climate change and international tourism: a simulation Klein, R.J.T., and Nicholls, R.J. 1999. Assessment of study. Global Environmental Change 15(3), 253­66. coastal vulnerability to sea-level rise. Ambio 28, 183­87. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 45 Klein, R.J.T., Buckley, E.N., Aston, J., Nicholls, R.J., Meehl, G.A., Stocker, T.F., Collins, W.D., Ragoonaden, S., Capobianco, M., Mizutani, N., and Friedlingstein, P., Gaye, A.T., Gregory, J.M., Kitoh, A., Nunn, P.D. 2000. Coastal Adaptation. In: B. Metz, O.R. Knutti, R., Murphy, J.M., Noda, A., Raper, S.C.B., Davidson, J.W. Martens, S.N.M. Van Rooijen, and L.L. Watterson, I.G., Weaver, A.J., and Zhao, Z.-C. 2007. Van Wie Mcgrory, eds., Methodological and Technological Global climate projections. In: S. Solomon, D. Qin, M. Issues in Technology Transfer. Special Report of the Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Intergovernmental Panel on Climate Change, Cambridge Tignor, and H.L. Miller, eds., Climate Change 2007: The University Press, Cambridge, UK, pp. 349­72. Physical Science Basis. Contribution of Working Group 1 to the Fourth Assessment Report of the Intergovernmental Klein, R.J.T., Nicholls, R.J., Ragoonaden, S., Panel on Climate Change. Cambridge University Press, Capobianco, M., Aston, J., and Buckley, E.N. 2001. Cambridge, UK, pp 433­97. Technological options for adaptation to climate change in coastal zone. Journal of Coastal Research 17, 531­43. Milliman, J.D., Broadus, J.M., and Gable, F. 1989. Environmental and economic implications of rising sea Kok, M., Jonkman, B., Kanning, W., Rijcken, T., and level and subsiding deltas: the Nile and Bengal exam- Stijnen, J. 2008. Deltacommissie: Toekomst voor het ples. Ambio 18, 340­45. Nederlandse polderconcept. Technische en financiele houdbaarheid. Mimura, N., Nurse, L., McLean, R.F., Agard, J., Briguglio, L., Lefale, P., Payet, R., and Sem, G. 2007. Lloyd's List. 2009. Ports of the World 2009: Volume 1: Small islands. In: M.L. Parry, O.F. Canziani, J.P. Albania ­ Lithuania, and Volume 2: Madagascar - Yemen. Palutikof, P.J. van der Linden, and C.E. Hanson, eds., Informa Maritime and Professional, London. Climate Change 2007: Impacts, Adaptation and Vulnerability. Contribution of Working Group II to the Lowe, J.A., Howard, T.P., Pardaens, A., Tinker, J., Holt, Fourth Assessment Report of the Intergovernmental Panel J., Wakelin, S., Milne, G., Leake, J., Wolf, J., Horsburgh, on Climate Change. Cambridge University Press, K., Reeder, T., Jenkins, G., Ridley, J., Dye, S., and Cambridge, UK, pp. 687­716. Bradley, S. 2009. UK Climate Projections Science Report: Marine and Coastal Projections. Met Office Hadley Moser, S.C., and Tribbia, J. 2008. Vulnerability to inun- Centre, Exeter, UK. dation and climate change impacts in California: coastal managers' attitudes and perceptions. Marine Society McFadden, L., Nicholls, R.J., Vafeidis, A.T., and Tol, Technology Journal 40, 35­44. R.S.J. 2007. A methodology for modeling coastal space for global assessment. Journal of Coastal Research 23(4), Myers, N. 2001. Environmental refugees: our latest 911­20. understanding. Philosophical Transactions of the Royal Society B 356, 16.1­16.5. McGranahan, G., BAlk, D., and Anderson, B. 2007. The rising tide: assessing the risks of climate change Nakicenovic, N., and Swart, R., eds. 2000. Emissions and human settlements in low elevation coastal zones. Scenarios. Special Report of the Intergovernmental Panel on Environment and Urbanisation 19(1), 17­37. Climate Change. Cambridge University Press, Cambridge, UK. McLean, R., Tsyban, A., Burkett, V., Codignotto, J.O., Forbes, D.L., Mimura, N., Beamish, R.J., and Ittekkot, Narita, D., Tol, R.S.J., and Anthoff, D. 2009. Economic V. 2001. Coastal zone and marine ecosystems. In: J.J. costs of extratropical storms under climate change: an McCarthy, O.F. Canziani, N.A., Leary, D.J., Dokken, application of FUND. Working Papers, WP274, and K.S. White, eds., Climate Change 2001: Impacts, Economic and Social Research Institute, Dublin. Adaptation and Vulnerability. Contribution of Working Group II to the Fourth Assessment Report of the Neumann, J. 2009. Unpublished memo on sea-level rise Intergovernmental Panel on Climate Change. Cambridge scenarios for the World Bank Economics of Adaptation University Press, Cambridge, UK, pp. 343­79. to Climate Change (EACC) study, 9 April. 46 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e Nicholls, R.J. 2002. Analysis of global impacts of sea- Nicholls, R.J., Klein, R.J.T., and Tol, R.S.J. 2007b. level rise: a case study of flooding. Physics and Chemistry Managing coastal vulnerability and climate change: a of the Earth 27, 1455­66. national to global perspective. In: L. McFadden, R.J. Nicholls, and E.C. Penning-Rowsell, eds., Managing Nicholls, R.J. 2004. Coastal flooding and wetland loss Coastal Vulnerability. Elsevier, Oxford, pp. 223­41. in the 21st Century: changes under the SRES climate and socio-economic scenarios. Global Environmental Nicholls, R.J., Hanson, S., Herweijer, C., Patmore, N., Change 14(1), 69­86. Hallegatte, S., Corfee-Morlot, J., Chateau, J., and Muir- Wood, R. 2008a. Ranking Port Cities with High Exposure Nicholls, R.J. 2007. Adaptation Options for Coastal Areas and Vulnerability to Climate Extremes: Exposure and Infrastructure: An Analysis for 2030. Report to the Estimates. OECD Environment Working Papers No. 1. UNFCCC, Bonn. OECD Publishing, Paris. Nicholls, R.J. In press. Impacts of and responses to sea- Nicholls, R.J., Wong, P.P., Burkett, V.R., Woodroffe, level rise. In: J.A. Church, P.L. Woodworth, T. Aarup, C.D., and Hay, J.E. 2008b. Climate change and coastal and S. Wilson, eds., Understanding Sea-Level Rise and vulnerability assessment: scenarios for integrated assess- Variability. Wiley-Blackwell, London. ment. Sustainability Science 3(1), 89­102. Nicholls, R.J., and Hoozemans, F.M.J. 2005. Global Nicholls, R.J., Woodroffe, C.D., and Burkett, V.R. 2009. vulnerability analysis. In: M. Schwartz, ed., Encyclopedia Coastal degradation as an indicator of global change. In: of Coastal Science. Kluwer Academic Publishers, T.M. Letcher, ed., Climate Change: Observed Impacts on Dordrecht, Netherlands, pp. 486­91. Planet Earth. Elsevier, Oxford, pp. 409­24. Nicholls, R.J., and Klein, R.J.T. 2005. Climate change Nordhaus, W.D., 2001. Documentation of the RICE-99 and coastal management on Europe's coast. In: J. Excel spreadsheet. Available at http://www.econ.yale. Vermaat, L. Bouwer, K. Turner, and W. Salomons, eds., edu/~nordhaus/homepage/RICEModelDiscussionas Managing European Coasts. Past, Present and Future. ofSeptember 30.htm. Springer, Berlin, pp. 199­226. Parry, M.L., Arnell, N.W., Berry, P.M., Dodman, D., Nicholls, R.J., and Lowe, J.A. 2004. Benefits of mitiga- Fankhauser, S., Hope, C., Kovats, S., Nicholls, R.J., tion of climate change for coastal areas. Global Satterthwaite, D., Tiffin, R., and Wheeler, T. 2009. Environmental Change 14, 229­44. Assessing the Costs of Adaptation to Climate Change: A Review of the UNFCCC and Other Recent Estimates. Nicholls, R.J., and Tol, R.S.J. 2006. Impacts and International Institute for Environment and responses to sea-level rise: a global analysis of the Development and Grantham Institute for Climate SRES scenarios over the twenty-first century, Change, London. 111 pp. Philosophical Transactions of the Royal Society A 364, 1073­95. Patt, A.G., Tadross, M., Nussbaumer, P., Asante, K., Metzger, M., Rafael, J., Goujon, A., and Brundrit, G. Nicholls, R.J., Wong, P.P., Burkett, V.R., Codignotto, 2010. Estimating least-developed countries' vulnerability J.O., Hay, J.E., McLean, R.F., Ragoonaden, S., and to climate-related extreme events over the next 50 years. Woodroffe, C.D. 2007a. Coastal systems and low-lying Proceedings of the National Academy of Sciences 107(4), areas. In: M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J. 1333­37. van der Linden, and C.E. Hanson, eds., Climate Change 2007: Impacts, Adaptation and Vulnerability. Peltier, W.R. 2000a. Global glacial isostatic adjustment Contribution of Working Group II to the Fourth and modern instrumental records of relative sea level Assessment Report of the Intergovernmental Panel on history. In: B.C. Douglas, M.S. Kearny, and S.P. Climate Change. Cambridge University Press, Leatherman, eds., Sea Level Rise: History and Cambridge, UK, pp. 315­57. Consequences. Academic Press, San Diego, pp. 65­95. D E V E L O P M E N T A N D C L I M AT E C H A N G E D I S C u S S I O N PA P E R S 47 Peltier, W.R. 2000b. ICE4G (VM2) Glacial isostatic Sugiyama, M., Nicholls, R. J., and Vafeidis, A.T. 2008. adjustment corrections, In: B.C. Douglas, M.S. Kearny, Estimating the Economic Cost of Sea-Level Rise. MIT and S.P. Leatherman, eds., Sea Level Rise: History and Global Change Joint Program Publications, Report 156. Consequences. Academic Press, San Diego. Syvitski, J.P.M., Kettner, A.J., Overeem, I., Hutton, Pielke, Jr., R.A., Gratz, J., Landsea, C.W., Collins, D., E.W.H., Hannon, M.T., Brakenridge, G.R., Day, J., Saunders, M., and Musulin, R. 2008. Normalized hurri- Vörösmarty, C., Saito, Y., Giosan, L., and Nicholls, R.J. cane damages in the United States: 1900­2005. Natural 2009. Sinking deltas due to human activities. Nature Hazards Review 9, 29­42. Geoscience, Vol. 2. Rahmstorf, S. 2007. A semi-empirical approach to Tol, R.S.J. 2006. The DIVA model: socio-economic projecting future sea-level rise. Science 315, 368­70. scenarios, impacts and adaptation and world heritage. In: DINAS-COAST Consortium 2006. DIVA 1.5.5, RSBC (Revised Status of British Columbia). 1996. Potsdam Institute for Climate Impact Research, Drainage, Ditch and Dike Act. Available at http://www. Potsdam, Germany, CD-ROM. bclaws.ca/Recon/document/freeside/--%20d%20--/ drainage%20%20ditch%20and%20dike%20act%20%20 Tol, R.J.S. 2007. The double trade-off between adapta- rsbc%201996%20%20c.%20102/00_96102_01. tion and mitigation for sea-level rise: an application of xml#section102. Accessed June 2009. FUND. Mitigation and Adaptation Strategies for Global Change 12(5), 741­53. Sachs, J.D., Mellinger, A.D., and Gallup, J.L. 2001. The geography of poverty and wealth. Scientific American Tol, R.S.J. 2008. Climate, development, and malaria: an 284(3), 70­75. application of FUND. Climatic Change 88(1), 21­34. Small, C., and Nicholls, R.J. 2003. A global analysis of Tol, R.S.J., and Yohe, G.W. 2007. The weakest link human settlement in coastal zones. Journal of Coastal hypothesis for adaptive capacity: an empirical test. Research 19(3), 584­99. Global Environmental Change 17, 218­27. Smedema, L.K., Vlotman, W., and Rycroft, D.W. 2004. Tol, R.S.J., Klein, R.J.T., and Nicholls, R.J. 2008. Modern Land Drainage. Planning, Design and Towards successful adaptation to sea-level rise along Management of Agriculture Drainage Systems. Leiden, Europe's coasts. Journal of Coastal Research 24 (2), Netherlands: Taylor and Francis. 432­50. Smit, B., Pilifosova, O., Burton, I., Challenger, B., UNCTAD. 1985. Port Development. A Handbook for Huq, S., Klein, R., and Yohe, G. 2001. Adaptation to Planners in Developing Countries. 2nd ed. United climate change in the context of sustainable develop- Nations Conference on Trade and Development, New ment and equity. In: J.J. McCarthy, O. Canziana, N.A. York. Leary, D.J. Dokken, and K.S. White, eds., Climate Change 2001: Impacts, Adaptation and Vulnerability. UNCTAD. 2005. Review of Maritime Transport. United Contribution of Working Group II to the Fourth Nations Conference on Trade and Development, New Assessment Report of the Intergovernmental Panel on York and Geneva. Climate Change. Cambridge University Press, Cambridge, UK, pp. 877­912. UNCTAD. 2008. Review of Maritime Transport. United Nations Conference on Trade and Development, New Stive, M.J., Capobianco, M., Wang, Z., Ruol, P., and York and Geneva. Buijsman, M. 1998. Morphodynamics of a tidal lagoon and the adjacent coast. In: J. Dronkers and M. Scheffers, UNDP. 2004. Reducing Disaster Risk ­ A Challenge for eds., Physics of Estuaries and Coastal Seas. Balkema, Development. United Nations Development Programme, Rotterdam, pp. 397­407. New York. 48 e C o n o m i C s o f C o a s ta l Z o n e a d a p tat i o n to C l i m at e C h a n g e UNFCCC. 1999. Coastal Adaptation Technologies, Warrick, R.A., Barrow, E.M., and Wigley, T.M.L., eds. Technical Paper FCCC/TP/1999/1, United Nations 1993. Climate and Sea Level Change: Observations, Framework Convention on Climate Change Secretariat, Projections, Implications. Cambridge University Press, Bonn. Cambridge, UK. UNPD. 2006. World Population Prospects. Department of Ward, P.J., Strzepek, K.M., Pauw, P., Brander, L.M., and Economic and Social Affairs, United Nations Aerts, J. 2010. Costs of Adaptation Related to Industrial Population Division. At http://www.un.org/esa/popula- and Municipal Water Supply and Riverine Flood tion/publications/wpp2006/English.pdf. Protection. Economics of Adaptation to Climate Change Aggregate Track, World Bank, Washington, DC. May. Vafeidis, A.T., Nicholls, R.J., McFadden, L., Tol, R.S.J., Hinkel, J., Spencer, T., Grashoff, P.S., Boot, G., and Williams, A., and Micallef, A. 2009. Beach Management: Klein, R.J.T. 2008. A new global coastal database for Principles and Practice. Earthscan Publications Limited, impact and vulnerability analysis to sea-level rise. London. Journal of Coastal Research 24, 917­24. Woodroffe, C.D., Kennedy, D.M., Brooke, B.P., and van Goor, M., Zitman, T., Wang, Z., and Stive, M. Dickson, M.E. 2006. "Geomorphological 2003. Impact of sea-level rise on the morphological equilibrium state of tidal inlets. Marine Geology 202, Evolution of Lord Howe Island and Carbonate 211­27. Production at the Latitudinal Limit to Reef van Koningsveld, M., Mulder, J.P.M., Stive, M.J.F., van Growth." Journal of Coastal Research. der Valk, L., and van der Weck, A.W. 2008. Living with sea-level rise and climate change: a case study of the Woodworth, P.L., and Blackman, D.L. 2004. Evidence Netherlands. Journal of Coastal Research 24(2), 367­79. for systematic changes in extreme high waters since the mid-1970s. Journal of Climate 17, 1190­97. Verhagen, H.J. 1998. Hydraulic boundary conditions. In: K.W. Pilarczyk, ed., Dikes and Revetments: Design, Zhang, K., Douglas, B.C., and Leatherman, S.P. 2000. Maintenance and Safety Assessment. Balkema, Rotterdam. Twentieth-century storm activity along the US east coast. Journal of Climate 13, 1748­61. Vermeer, M., and Rahmstorf, S. 2009. Global sea level linked to global temperature. Proceedings of the National Zhang, K., Douglas, B.C., and Leatherman, S.P. 2004. Academy of Sciences 106(51), 21527­32. Global warming and coastal erosion. Climatic Change 64, 41­58. von Storch, H., and Woth, K. 2008. Storm surges: perspectives and options. Sustainability Science 3(1), 33­43. The World Bank Group 1818 H Street, NW Washington, D.C. 20433 USA Tel: 202-473-1000 Fax: 202-477-6391 Internet: www.worldbank.org/climatechange