BACKGROUNDER: SEA-LEVEL RISE AND CLIMATE CHANGE IMPACTS AND ADAPTATION NEEDS ON PRINCE EDWARD ISLAND: STUDY RESULTS

1. Purpose

The goal of the study was to assess the physical and socio-economic impacts of climate change and accelerated rise in sea-level on the coast of Prince Edward Island (P.E.I.), particularly in relation to the following:

• an anticipated increase in the frequency and extent of storm-surge flooding in Charlottetown, and
• an anticipated decrease in sea ice, an increase in wave energy, and a probable increase in rates of shore erosion on the north shore.

Another important objective was to consider feasible and effective adaptation measures that might be adopted in P.E.I. to minimize the impacts of these changes. As this study is one of the first of its kind, it is intended to serve as a template for future studies. Moreover, it was deemed important to assess how well the study met its aims and how future studies could be improved.

The coast of P.E.I. was initially selected for this climate-change study in part because it had been identified as one of the regions most sensitive to sea-level rise in Canada. In Charlottetown Harbour, relative sea level is rising and storm-surge events are increasingly common. For example, it is estimated that unusually high tides accompanied by a storm-surge event on 21 January 2000 caused nearly $1 million worth of damage to P.E.I., a significant underestimate of the true cost.

Also, parts of the north shore of P.E.I. were studied since an energetic wave climate, rising mean sea levels and an increase in severe storm activity in this area could have detrimental effects on shoreline stability and coastal ecosystems, as well as on human activity and well-being.

2.Components of the study

The study, carried out by a team from Natural Resources Canada, Environment Canada, Dalhousie University, the Centre of Geographic Sciences of the Nova Scotia Community College, the City of Charlottetown and other partners, was organized into the following components:

• Digital elevation models (DEMs)were developed using airborne Light Detection and Ranging (LIDAR) surveys. This resulted in high resolution topographic maps of the Charlottetown and Rustico areas for flood impact analysis.
• A climatological analysis of sea level,storm surges, winds, waves and ice cover in the Gulf of St. Lawrence was carried out.
• A meteorological storm surge model was developed.
• Sensitivity studies with the surge model and statistical studies of long-term sea-level rise, astronomical tides and annual maxima, based on observations at Charlottetown, were combined to provide flooding forecasts at three critical levels: 4.23 m above Chart Datum (CD)(the 21 January 2000 storm); 4.70 m CD (a lesser storm surge superimposed on higher sea level); and 4.93 m (the 21 January 2000 storm plus 100 years of predicted relative sea-level rise at Charlottetown). These water levels were then superimposed on the high-resolution topographic map produced in the first step to predict which areas of Charlottetown would be flooded in the three scenarios.
• A socio-economic analysis was carried out to estimate the number and value of properties in Charlottetown at risk from these three flooding scenarios; the effects on the coastal infrastructure; and the effects on health, education and employment. For the north shore, an assessment was carried out on the effects of increased erosion on real property loss for cottage and non-cottage properties, and on non-market values for wetlands, forested land, beaches and dunes.
• A review of adaptation measures demonstrated that proactive retreat or avoidance is feasible and highly cost-effective in many rural areas in P.E.I. but may not be easy to implement where subdivision and dense development have occurred in coastal communities or in urban centres. Criteria for set-back and other adaptive measures were developed and a number of recommendations were made.

3. Summary of findings

Private and public property in both the residential and commercial sectors in Charlottetown are at risk of damage from flooding. With flooding to 4.23 m CD, approximately 460 properties with assessed property values of $172 million either will be flooded or be at risk of flooding. The value of properties at risk for flooding to 4.70 m CD is $190 million, while those at risk for flooding to the 4.93 m CD level are valued at approximately $202 million.

Over 1.1 million tourists visited P.E.I. in 2000 and spent approximately $257 million. Of the total pleasure private motor-vehicle and air visitors to the island in the summer tourism season, 29.9% reported Charlottetown as their "main overnight destination". In Charlottetown, there are approximately 335 municipally designated heritage properties, most of which are in and around the downtown core, and about a dozen federal sites, many of which lie within the probable flood plain outlined by the DEM. A flooding level of 4.23 m CD would render 30 municipally designated heritage properties at risk of flooding (total assessed value of $8.6 million). Flooding levels of 4.70 and 4.93 m CD would render an additional 11 and 19 properties at risk, respectively ($10.5 million and $11.3 million in assessed property values).

The City of Charlottetown has invested millions of dollars in developing and upgrading its stormwater, sewage and waste treatment systems, all of which could be damaged to varying degrees with a rise in sea level.

The total value of the infrastructure at risk in Charlottetown is $46 million. Estimates show that flooding to the level of 4.23 m CD would affect approximately 150,000 m2 of right-of-way (values at $12 million) and sidewalks to the value of $1 million, excluding the value of the land on which they are located.

The replacement cost of the Trigen Energy-from-Waste Facility, which provides heating services in 80 buildings, many of which do not have other sources of heating, is estimated at $25 to $30 million. The Maritime Electric facility, which could be at risk, carries an approximate asset value of $48 million.

Some "community costs and/or damages" due to lost wages and health care costs caused by homeowners taking time from their work for clean-up would occasion a cost directly related to the surge incident, as would businesses being closed for repairs. The employment created in the clean-up efforts to repair, service and rebuild commercial establishments should not be seen as employment revenue that offsets the costs of the storm.

The property of both the Queen Elizabeth Hospital and the Hillsborough Hospital and Special Care Unit would be affected. There does not appear to be any danger of flooding in the main buildings themselves, however, with flooding levels of 4.70 m and 4.93 m, there is some risk to one of the auxiliary buildings.

In some places, rising mean sea levels, less sea ice and higher wave energy along the north shore can be expected to cause more severe damage from erosion and, possibly, rapid coastal change. Photogrammetric analysis dating back to the mid-1930s shows significant change along much of the coast, including severe erosion in places and shoreline recovery or dune growth in others. The longer-term view from marine geological surveys shows that the coast, on average, has been retreating by at least 0.5 m/year for several thousand years. Property is lost, wetlands are encroached upon and migrate inland (but can be permanently lost if migration is limited by infrastructure), and coastal infrastructure and community-related resources are put at risk in a situation of accelerated shoreline erosion. The value of the lands lost due to coastal erosion and the lost value of the services produced by or from it represent a real cost of climate change.

The value of cottage-land lost to erosion between 1935 and 1990 was $816,000 ($15,000/year). Between 1980-81 and 1990, the value was $242,000 ($22,0000/year).

The value of non-cottage land lost to erosion between 1935 and 1990 was $63,400 ($1,100/year) and $10,600 ($1,000/year) from 1980-81 to 1990. This means that approximately one-sixth of the erosion since 1935 occurred in the last decade.

Preliminary estimates of future erosion rates under climate change and relative sea-level rise suggest a potential increase of 1.5 to 2 times the 1935-1990 mean rates of erosion for the study area.

At double the present erosion rate, almost 10 percent of the present area of coastal properties in the study area will be lost within the next 20 years, and almost one-half in the next 100 years.

The study area includes a number of saltwater marshes located further inland. Using a value of $21,206/ha per year, wetlands add a value of $188,448 (in benefits from ecological services) to the assessed value of the land.

Using a value of $2.74/m2 for water filtration, removal of air pollutants and control of runoff, the assessed value of the 18 hectares of forested land in the study area on the north shore can be augmented by $49,800 to account for the inclusion of ecosystem values. The role of forests in erosion control could raise the augmented total value of the land.

The coastal dunes of P.E.I. are among the key natural tourist attractions of the province. The 100 hectares within the study area are at risk of being breached by wave action in severe storms. Apart from its tourism value, the sand dune system on the north shore seems to be the most important land-conservation tool available in nature, and its absence could lead to accelerated rates of erosion in vulnerable areas.

Tide gauge data indicates that the mean sea level at Charlottetown rose 32 cm in this century and 29 cm at Rustico. Part of the long-term sea-level rise (perhaps 20 cm/century) is due to crustal subsidence following post-glacial adjustments to changing ice and water loads. The remaining 12 cm/century is a signal of global and regional sea-level rise. It should be noted that storm surges and ocean waves are also factors at the coastline and are carried to higher levels with rising mean sea level. Even without climate change, the present rate of sea-level rise in P.E.I. will bring challenges to human interests and ecological systems in the coastal zone.

This study has adopted the Intergovernmental Panel on Climate Change (IPCC) central value of about 0.5 m for sea-level rise, combined with a risk-conservative estimate of 0.2 m for crustal subsidence, to obtain a total projection of 0.7 m relative sea-level rise by 2100 in the Charlottetown region. The maximum IPCC projection could raise the mean sea level at Charlottetown to as much as 1.10 m.

Storm surges are the meteorological effects on sea level and can be defined at the coast as the difference between the observed water level and the predicted astronomical tide. Storm surges can be positive or negative, can occur everywhere along our coastlines, anywhere in the tidal cycle, or can last over several tidal cycles. Large positive storm surges at times of high tide lead to coastal inundation.

Storm surges above 60 cm occur about eight times a year in Charlottetown compared, for instance, to twice a year along the Atlantic coast in Halifax. They are mainly associated with winter storms and show great variability over the years. Storm surge events above 120 cm occur, on average, about twice a decade.

The DEM for Charlottetown determined that sea water begins to flood the waterfront at about 3.6 m CD. A storm surge of less than 60 cm combined with the highest predicted tide (2.91 m) cannot reach this level. Of all the storm surge events noted between 1911 and 1998, only six reached this level and had an impact on the Charlottetown waterfront. If sea level rise to the year 2000 is taken into account, then eight of these storms would have caused some flooding of the waterfront if they had occurred in the year 2000. With sea-level rise and global warming, not only will flooding in Charlottetown be higher, but the floods at the lower levels will become much more frequent.

A numerical model to forecast storm surges driven by wind and sea-level atmospheric pressure was developed at Dalhousie University. The model can forecast storm surges to within about 10 cm with a lead time of one day.

The Probable Maximum Storm (PMS) methodology indicates that storm surges in the Gulf of St. Lawrence are highly sensitive to changes in storm trajectory.

The Conditional Probability Method (CPM) was developed to estimate the probability that a specified level will be exceeded at least once by a given date. The CPM suggests that, at the present rate of sea-level rise and with no increase in storminess, the probability that sea level in Charlottetown will exceed 4.22 m CD at least once by 2050 is about 0.8.

High wind speeds are the defining characteristics of storms and contribute to the development of both storm waves and storm surges. The historical wind climatology of P.E.I. and the Magdalen Islands was investigated to identify periods in the wind record during which wind storms were more frequent or intense and to describe past storm events and wind regimes that contributed to storm wave and storm surge events. Knowledge of the historical variability and impacts of storms provides conceptual insight into future frequencies and the severity of coastal flooding, as well as erosion under scenarios of changing climate.

Analysis of storm-surge events indicated a significant correlation between large surges in the southern Gulf of St. Lawrence and strong northeast winds. Other surges are related to winds out of the north and northwest.

Waves are one of the most widely recognized indicators of storm activity and constitute a significant natural hazard for shoreline erosion and infrastructure damage. Coastal erosion, increased sediment mobility and damage to infrastructure can be caused by waves impinging on the shoreline, especially when they are superimposed on higher-than-normal water levels during storm surges.

Wave data was collected off the north shore of P.E.I. and includes one benchmark storm during which waves of up to 14 m were recorded.

Time series of the hindcast data show that waves tend to be largest and most numerous in the fall as waves are fetch-limited by ice in winter and wind storms are uncommon in summer.

The presence of sea ice in the Gulf of St.Lawrence inhibits wave development, thereby reducing winter storm erosion. Waves are expected to increase if sea ice in the Gulf decreases, as predicted in future global change scenarios. There may also be some effect on storm surges.

The latest results of the Canadian Global Circulation Model indicate that the Gulf of St. Lawrence may be free of ice by 2045. The dominant characteristic of the time series is the oscillation between years of maximum ice cover and years in which the total accumulated ice cover is 50 percent less.

In an ice-free Gulf, the length of the wave season would greatly increase, and wave impacts would occur in the winter as well as fall and summer. Wind climatology indicates that this would greatly increase the contribution of wave energy from the northeast. Investigators suggest the possibility of increased winter wind speeds and the largest storm surges occur more frequently in winter. There is potential for increased frequencies of extreme wave events and associated erosion, sediment mobilization and infrastructure damage in the Gulf of St. Lawrence.

To predict areas at risk of coastal storm-surge flooding, it is necessary to have an accurate and high-resolution representation of the topography. LIDAR, an emerging technology, produces a series of point measurements.

Water levels associated with the three flood levels selected for modeling the extent of flooding in Charlottetown at 4.23, 4.70 and 4.93 m CD were projected to the P.E.I. double stereographic map projection and given to the City of Charlottetown for planning purposes.

Estuarine deposits mapped and sampled offshore include several sites, presently at depths of 18 to 25 m, that date back some 6,000 years. As the outer coast at the time may have been between < 1 and 5+ km further seaward, these observations indicate a long-term mean coastal recession rate of at least 50 m/century (0.5 m/year).

Historical map evidence suggests that large-scale shoreline adjustment and landward sand movement occurred in response to large storms prior to 1880 and that extensive washover was maintained on some beaches by storms prior to 1935, after which the dunes began to recover.

Records of erosion and accretion for a 12-km section of the north shore for various periods between 1935 and 1990 show that there is considerable variation in the rates of erosion both in time and space. In some sections, erosion prevailed during the 55-year period, with coastal retreat rates generally ranging from 0.2 to 2.5 m per year.

Our estimates of future coastal retreat range from 0.62 to 0.80 m/year, an increase of about 55 percent over observed rates in the past.

Areas showing significant erosion can be expected to experience long-term future erosion at rates at least equal to those of the past 65 years. Some areas could experience massive breaching and rapid landward migration given a sufficiently energetic storm.

The vulnerability of a community or ecosystem to climate change is a function of its exposure and susceptibility to environmental change, and its inherent or managed adaptive capacity in the face of such change. Adaptations are actions taken in response to a projected or actual change in the climate or other change in the environment. They aim is to maximize positive effects and to minimize adverse impacts, thereby reducing vulnerability. Adaptation may be reactive or proactive, and may occur at any level (local, provincial, national or international) or at a combination of these levels.

There are three broad categories of adaptation to sea-level rise and climate change in the coastal zone: protection, accommodation and retreat (or avoidance):

• Protection is costly. It may have limited long-term effectiveness in exposed locations and may be successful where wave energy is limited. Soft protection options should be considered: their design should take future climate change into account and they must recognize that potential impacts of the protection work on the wider coastal system, including adjacent shore-front properties.
• Accommodation may involve the redesign of structures to minimize impacts; zoning to encourage appropriate land use with low capital investment on vulnerable properties; efforts to increase natural resilience through such measures as coastal dune rehabilitation, dyke opening and wetland renewal; substitution of bridges in place of causeways; or "soft" protection measures such as beach nourishment.
• Retreat (avoidance of risk) represents a form of proactive adaptation to eliminate a direct impact. The simplest form of retreat involves avoidance of vulnerable properties by individual buyers or decisions against building within flood or erosion hazard zones.

Many options have been considered that are not always based on appropriate or accurate data. There may be no one option that is best or effective in isolation. Appropriate adaptation may require a mix of options that will vary with place and time. Local adaptation needs are best solved locally or at least with the participation and buy-in of local stakeholders.

Adaptation strategy needs to be adaptable. Vulnerability, understanding, technology and coastal dynamics may change with time, and adaptation needs should be reassessed regularly.