Allowances for evolving coastal flood risk under uncertain local sea-level rise

Estimates of future flood hazards made under the assumption of stationary mean sea level are biased low due to sea-level rise (SLR). However, adjustments to flood return levels made assuming fixed increases of sea level are also inadequate when applied to sea level that is rising over time at an uncertain rate. SLR allowances—the height adjustment from historic flood levels that maintain under uncertainty the annual expected probability of flooding—are typically estimated independently of individual decision-makers’ preferences, such as time horizon, risk tolerance, and confidence in SLR projections. We provide a framework of SLR allowances that employs complete probability distributions of local SLR and a range of user-defined flood risk management preferences. Given non-stationary and uncertain sea-level rise, these metrics provide estimates of flood protection heights and offsets for different planning horizons in coastal areas. We illustrate the calculation of various allowance types for a set of long-duration tide gauges along U.S. coastlines.     

Estimating sea-level extremes under conditions of uncertain sea-level rise

Estimation of expected extremes, using combinations of observations and model simulations, is common practice. Many techniques assume that the background statistics are stationary and that the resulting estimates may be used satisfactorily for any time in the future. We are now however in a period of climate change, during which both average values and statistical distributions may change in time. The situation is further complicated by the considerable uncertainty which accompanies the projections of such future change. Any useful technique for the assessment of future risk should combine our knowledge of the present, our best estimate of how the world will change, and the uncertainty in both. A method of combining observations of present sea-level extremes with the (uncertain) projections of sea-level rise during the 21st century is described, using Australian data as an example. The technique makes the assumption that the change of flooding extremes during the 21st century will be dominated by the rise in mean sea level and that the effect of changes in the variability about the mean will be relatively small. The results give engineers, planners and policymakers a way of estimating the probability that a given sea level will be exceeded during any prescribed period during the present century.     

Increasing flood risk and wetland losses due to global sea-level rise: regional and global analyses

To develop improved estimates of (1) flooding due to storm surges, and (2) wetland losses due to accelerated sea-level rise, the work of Hoozemans et al. (1993) is extended to a dynamic analysis. It considers the effects of several simultaneously changing factors, including: (1) global sea-level rise and subsidence; (2) increasing coastal population; and (3) improving standards of flood defence (using GNP/capita as an “ability-to-pay” parameter). The global sea-level rise scenarios are derived from two General Circulation Model (GCM) experiments of the Hadley Centre: (1) the HadCM2 greenhouse gas only ensemble experiment and (2) the more recent HadCM3 greenhouse gas only experiment. In all cases there is a global rise in sea level of about 38 cm from 1990 to the 2080s. No other climate change is considered. Relative to an evolving reference scenario without sea-level rise, this analysis suggests that the number of people flooded by storm surge in a typical year will be more than five times higher due to sea-level rise by the 2080s. Many of these people will experience annual or more frequent flooding, suggesting that the increase in flood frequency will be more than nuisance level and some response (increased protection, migration, etc.) will be required. In absolute terms, the areas most vulnerable to flooding are the southern Mediterranean, Africa, and most particularly, South and South-east Asia where there is a concentration of low-lying populated deltas. However, the Caribbean, the Indian Ocean islands and the Pacific Ocean small islands may experience the largest relative increase in flood risk. By the 2080s, sea-level rise could cause the loss of up to 22% of the world's coastal wetlands. When combined with other losses due to direct human action, up to 70% of the world's coastal wetlands could be lost by the 2080s, although there is considerable uncertainty. Therefore, sea-level rise would reinforce other adverse trends of wetland loss. The largest losses due to sea-level rise will be around the Mediterranean and Baltic and to a lesser extent on the Atlantic coast of Central and North America and the smaller islands of the Caribbean. Collectively, these results show that a relatively small global rise in sea level could have significant adverse impacts if there is no adaptive response. Given the “commitment to sea-level rise” irrespective of any realistic future emissions policy, there is a need to start strategic planning of appropriate responses now. Given that coastal flooding and wetland loss are already important problems, such planning could have immediate benefits.     

Economic impacts of climate change in Europe: sea-level rise

This paper uses two models to examine the direct and indirect costs of sea-level rise for Europe for a range of sea-level rise scenarios for the 2020s and 2080s: (1) the DIVA model to estimate the physical impacts of sea-level rise and the direct economic cost, including adaptation, and (2) the GTAP-EF model to assess the indirect economic implications. Without adaptation, impacts are quite significant with a large land loss and increase in the incidence of coastal flooding. By the end of the century Malta has the largest relative land loss at 12% of its total surface area, followed by Greece at 3.5% land loss. Economic losses are however larger in Poland and Germany (483 and483 and 391 million, respectively). Coastal protection is very effective in reducing these impacts and optimally undertaken leads to protection levels that are higher than 85% in the majority of European states. While the direct economic impact of sea-level rise is always negative, the final impact on countries’ economic performances estimated with the GTAP-EF model may be positive or negative. This is because factor substitution, international trade, and changes in investment patterns interact with possible positive implications. The policy insights are (1) while sea-level rise has negative and huge direct economic effects, overall effects on GDP are quite small (max −0.046% in Poland); (2) the impact of sea-level rise is not confined to the coastal zone and sea-level rise indirectly affects landlocked countries as well (Austria for instance loses −0.003% of its GDP); and (3) adaptation is crucial to keep the negative impacts of sea-level rise at an acceptable level.     

Spatial variations of sea-level rise and impacts: An application of DIVA

Due to complexities of creating sea-level rise scenarios, impacts of climate-induced sea-level rise are often produced from a limited number of models assuming a global uniform rise in sea level. A greater number of models, including those with a pattern reflecting regional variations would help to assure reliability and a range of projections, indicating where models agree and disagree. This paper determines how nine new patterned-scaled sea-level rise scenarios (plus the uniform and patterned ensemble mean rises) influence global and regional coastal impacts (wetland loss, dry land loss due to erosion and the expected number of people flooded per year by extreme sea levels). The DIVA coastal impacts model was used under an A1B scenario, and assumed defences were not upgraded as conditions evolved. For seven out of nine climate models, impacts occurred at a proportional rate to global sea-level rise. For the remaining two models, higher than average rise in sea level was projected in northern latitudes or around populated coasts thus skewing global impact projections compared with the ensemble global mean. Regional variability in impacts were compared using the ensemble mean uniform and patterned scenarios: The largest relative difference in impacts occurred around the Mediterranean coast, and the largest absolute differences around low-lying populated coasts, such as south, south-east and east Asia. Uniform projections of sea-level rise impacts remain a useful method to determine global impacts, but improved regional scale models of sea-level rise, particularly around semi-enclosed seas and densely populated low-lying coasts will provide improved regional impact projections and a characterisation of their uncertainties.     

Exposure of developing countries to sea-level rise and storm surges

An increase in sea surface temperature is strongly evident at all latitudes and in all oceans. The scientific evidence to date suggests that increased sea surface temperature will intensify cyclone activity and heighten storm surges. The paper assesses the exposure of (coastal) developing countries to sea-level rise and the intensification of storm surges. Geographic Information System (GIS) software is used to overlay the best available, spatially-disaggregated global data on critical exposed elements (land, population, GDP, agricultural extent and wetlands) with the inundation zones projected with heightened storm surges and a 1 m sea-level rise. Country-level results indicate a significant increase in exposure of developing countries to these climate-induced changes.     

Toward a physically plausible upper bound of sea-level rise projections

Anthropogenic sea-level rise (SLR) causes considerable risks. Designing a sound SLR risk-management strategy requires careful consideration of decision-relevant uncertainties such as the reasonable upper bound of future SLR. The recent Intergovernmental Panel on Climate Change’s (IPCC) Fourth Assessment reported a likely upper SLR bound in the year 2100 near 0.6 m (meter). More recent studies considering semi-empirical modeling approaches and kinematic constraints on glacial melting suggest a reasonable 2100 SLR upper bound of approximately 2 m. These recent studies have broken important new ground, but they largely neglect uncertainties surrounding thermal expansion (thermosteric SLR) and/or observational constraints on ocean heat uptake. Here we quantify the effects of key parametric uncertainties and observational constraints on thermosteric SLR projections using an Earth system model with a dynamic three-dimensional ocean, which provides a mechanistic representation of deep ocean processes and heat uptake. Considering these effects nearly doubles the contribution of thermosteric SLR compared to previous estimates and increases the reasonable upper bound of 2100 SLR projections by 0.25 m. As an illustrative example of the effect of overconfidence, we show how neglecting thermosteric uncertainty in projections of the SLR upper bound can considerably bias risk analysis and hence the design of adaptation strategies. For conditions close to the Port of Los Angeles, the 0.25 m increase in the reasonable upper bound can result in a flooding-risk increase by roughly three orders of magnitude. Results provide evidence that relatively minor underestimation of the upper bound of projected SLR can lead to major downward biases of future flooding risks.     

Relative sea-level rise and the conterminous United States: Consequences of potential land inundation in terms of population at risk and GDP loss

Global sea-level rise poses a significant threat not only for coastal communities as development continues but also for national economies. This paper presents estimates of how future changes in relative sea-level rise puts coastal populations at risk, as well as affect overall GDP in the conterminous United States. We use four different sea-level rise scenarios for 2010–2100: a low-end scenario (Extended Linear Trend) a second low-end scenario based on a strong mitigative global warming pathway (Global Warming Coupling 2.6), a high-end scenario based on rising radiative forcing (Global Warming Coupling 8.5) and a plausible very high-end scenario, including accelerated ice cap melting (Global Warming Coupling 8.5+). Relative sea-level rise trends for each US state are employed to obtain more reasonable rates for these areas, as long-term rates vary considerably between the US Atlantic, Gulf and Pacific coasts because of the Glacial Isostatic Adjustment, local subsidence and sediment compaction, and other vertical land movement. Using these trends for the four scenarios reveals that the relative sea levels predicted by century's end could range – averaged over all states – from 0.2 to 2.0 m above present levels. The estimates for the amount of land inundated vary from 26,000 to 76,000 km². Upwards of 1.8 to 7.4 million people could be at risk, and GDP could potentially decline by USD 70–289 billion. Unfortunately, there are many uncertainties associated with the impact estimates due to the limitations of the input data, especially the input elevation data. Taking this into account, even the most conservative scenario shows a significant impact for the US, emphasizing the importance of adaptation and mitigation.     

Island abandonment and sea-level rise: An historical analog from the Chesapeake Bay,USA

Small islands are widely agreed to be vulnerable to human-induced sea-level rise during the 21st century and beyond, with forced abandonment of some low-lying oceanic islands being a real possibility. A regional abandonment of islands in the Chesapeake Bay, USA provides an historical analog of such vulnerability as this has been linked to a mid 19th Century acceleration in relative sea-level rise. Using a case study approach for Holland Island, Maryland, this hypothesis was tested using a range of physical and human historical data. While sea-level rise was the underlying driver, this analysis shows that the abandonment was more complex than a direct response to sea-level rise. Between 1850 and 1900, Holland Island was a booming community and population increased from 37 to 253, with immigration causing the majority of the increase. At the same time, the upland area where people made their homes was steadily diminishing, losing about 15 ha or 38% of the total. After 1900, the island experienced a decrease in population to 169 in 1916, with final abandonment in 1918, with the exception of one family who left by 1920. Final abandonment was triggered by this depopulation as the population fell below a level that could support critical community services, and the community lost faith in their future on Holland Island. It is likely that similar social processes determined the abandonment of the other Chesapeake Bay islands. Looking to the future, it shows that many small low-lying islands could be abandoned due to sea-level rise long before they become physically uninhabitable.     

Implications of sea-level rise and extreme events around Europe: a review of coastal energy infrastructure

Sea-level rise and extreme events have the potential to significantly impact coastal energy infrastructure through flooding and erosion. Disruptions to supply, transportation and storage of energy have global ramifications and potential contamination of the natural environment. On a European scale, there is limited information about energy facilities and their strategic plans for adapting to climate change. Using a Geographical Information System this paper assesses coastal energy infrastructure, comprising (1) oil/gas/LNG/tanker terminals and (2) nuclear power stations. It discusses planning and adaptation for sea-level rise and extreme events. Results indicate 158 major oil/gas/LNG/tanker terminals in the European coastal zone, with 40 % located on the North Sea coast. There are 71 operating nuclear reactors on the coast (37 % of the total of European coastal countries), with further locations planned in the Black, Mediterranean and Baltic Seas. The UK has three times more coastal energy facilities than any other country. Many north-west European countries who have a high reliance on coastal energy infrastructure have a high awareness of sea-level rise and plan for future change. With long design lives of energy facilities, anticipating short, medium and long-term environmental and climatic change is crucial in the design, future monitoring and maintenance of facilities. Adaptation of coastal infrastructure is of international importance, so will be an ongoing important issue throughout the 21^st century.     

A simple technique for estimating an allowance for uncertain sea-level rise

Projections of climate change are inherently uncertain, leading to considerable debate over suitable allowances for future changes such as sea-level rise (an ??allowance?? is, in this context, the amount by which something, such as the height of coastal infrastructure, needs to be altered to cope with climate change). Words such as ??plausible?? and ??high-end?? abound, with little objective or statistically valid support. It is firstly shown that, in cases in which extreme events are modified by an uncertain change in the average (e.g. flooding caused by a rise in mean sea level), it is preferable to base future allowances on estimates of the expected frequency of exceedances rather than on the probability of at least one exceedance. A simple method of determining a future sea-level rise allowance is then derived, based on the projected rise in mean sea level and its uncertainty, and on the variability of present tides and storm surges (??storm tides??). The method preserves the expected frequency of flooding events under a given projection of sea-level rise. It is assumed that the statistics of storm tides relative to mean sea level are unchanged. The method is demonstrated using the GESLA (Global Extreme Sea-Level Analysis) data set of roughly hourly sea levels, covering 198 sites over much of the globe. Two possible projections of sea-level rise are assumed for the 21st century: one based on the Third and Fourth Assessment Reports of the Intergovernmental Panel on Climate Change and a larger one based on research since the Fourth Assessment Report.     

Past and future sea-level rise along the coast of North Carolina,USA

We evaluate relative sea level (RSL) trajectories for North Carolina, USA, in the context of tide-gauge measurements and geological sea-level reconstructions spanning the last ~11,000 years. RSL rise was fastest (~7 mm/yr) during the early Holocene and slowed over time with the end of the deglaciation. During the pre-Industrial Common Era (i.e., 0–1800 CE), RSL rise (~0.7 to 1.1 mm/yr) was driven primarily by glacio-isostatic adjustment, though dampened by tectonic uplift along the Cape Fear Arch. Ocean/atmosphere dynamics caused centennial variability of up to ~0.6 mm/yr around the long-term rate. It is extremely likely (probability P=0.95) that 20th century RSL rise at Sand Point, NC, (2.8 ± 0.5 mm/yr) was faster than during any other century in at least 2,900 years. Projections based on a fusion of process models, statistical models, expert elicitation, and expert assessment indicate that RSL at Wilmington, NC, is very likely (P=0.90) to rise by 42–132 cm between 2000 and 2100 under the high-emissions RCP 8.5 pathway. Under all emission pathways, 21st century RSL rise is very likely (P>0.90) to be faster than during the 20th century. Due to RSL rise, under RCP 8.5, the current ‘1-in-100 year’ flood is expected at Wilmington in ~30 of the 50 years between 2050-2100.     

Contribution of glacier melt to sea-level rise since AD 1865: a regionally differentiated calculation

 The contribution of glacier melt, including the Greenland ice-sheet, to sea-level change since AD 1865 is estimated on the basis of modelled sensitivity of glacier mass balance to climate change and historical temperature data. Calculations are done in a regionally differentiated manner to overcome the inhomogeneity of the global distribution of glaciers. A distinction is made between changes in summer temperature and in temperature over the rest of the year. Our best estimate of the ice melt in the period 1865–1990 in terms of sea-level change equivalent is 5.7 cm (2.7 cm for glaciers and 3.0 cm for the Greenland ice-sheet). Additional calculations show that simpler methods, like using annual or even global mean temperature anomaly give estimates that differ by up to 55%. Consequently, a regionally differentiating approach is advised for making projections of glacier melt with GCM output. Received: 6 December 1996/Accepted: 30 May 1997     

The socio-economic impact of sea-level rise on the Netherlands: A study of possible scenarios

This paper describes a simulation study of some of the socio-economic consequences of a rise in sea level on Dutch society. A computer simulation model for the greenhouse problem has been developed, which tries to capture the climate change cause-effect relationship for a combination of greenhouse-gas emissions. The impact of emissions of greenhouse gases on global temperature and sea-level rise can be calculated using the model. Additionally, separate, independent modules have been implemented in order to quantify the socio-economic consequences for the Netherlands. Four consistent sets of scenarios have been developed, based on differences in economic growth, energy use, international environmental measures, etc. On the basis of these scenarios estimates are made of the costs of coastal defence and water management in the Netherlands as a result of adaptation to the impacts of sea-level rise.     

Integrating knowledge to assess coastal vulnerability to sea-level rise: The development of the DIVA tool

This paper describes the development of the DIVA tool, a user-friendly tool for assessing coastal vulnerability from subnational to global levels. The development involved the two major challenges of integrating knowledge in the form of data, scenarios and models from various natural, social and engineering science disciplines and making this integrated knowledge accessible to a broad community of end-users. These challenges were addressed by (i) creating and applying the DIVA method, an iterative, modular method for developing integrating models amongst distributed partners and (ii) making the data, scenarios and integrated model, equipped with a powerful graphical user interface, directly and freely available to end-users.