»conference topics« · the shore of the arctic seas are generally of low relief ... between tidal...

A » Driving factors and scales of storm surges « Oral 3 Poster 21

B » Contemporary risk and management « Oral 43 Poster 63

C » History and intercultural perception « Oral 69 Poster 83

D » The role of scientifi c information - dealing with uncertainty « Oral 93 Poster 107

E » Institutional and economic dimensions - dealing with uncertainty and disaster reduction « (includes DKKV session) Oral 123 Poster 133

F » Building with nature « (additional session) Oral 143G » Social dimensions of risk and vulnerabililty « (additional session) Oral 147 Poster 157

H » Supporting storm surge applications using satellite data « (additional session) Oral 161 Poster 171

I » The climate of storm surges « (additional session) Oral 179 Poster 187

» The book of abstracts includes oral abstracts as well as poster abstracts «

Abstracts are listed according to i) their session topics (A-I), ii) to their status (oral or poster) and iii) according to their abstract IDs (accending list).

»CONFERENCE TOPICS«

» Driving factorsand scales of storm surges «

»STORM SURGES CONGRESS 2010« 13th - 17th September, 2010

Risk and management of currentand future storm surgesUniversity of Hamburg, Germany

»ORAL«

Storm Surges Congress, Hamburg, Germany13–17 September 2010SSC2010-9© Author(s) 2010

Storm Surge Flooding of Deltas Made Susceptible by Human ActivitiesJ.P.M. Syvitski, G.R. Brakenridge, A.J. Kettner, and I Overeem([email protected]) Community Surface Dynamics Modeling System (CSDMS), Institute of Arctic and AlpineResearch (INSTAAR), University of Colorado, Boulder CO, USA, 80309-0545

Deltas are densely populated, intensively farmed landforms, that are being threatened by marine flooding, not justby rising sea levels, but more so by sediment compaction from water, oil and gas mining, sequestration of sedimentin upstream reservoirs, and from floodplain engineering. Visible and near-infrared Moderate Resolution ImagingSpectroradiometer (MODIS) satellite images in conjunction with the Advanced Microwave Scanning Radiometer(AMSR-E) were used to establish the extent of recent flooding on the deltas, whether the flooding was from coastalstorm surges, and whether the floodwaters carried suspended sediment. Of 33 representative world deltas examined,20 have experienced severe flooding in the last decade from storm surges, temporarily submerging 72,240 km2.Areas vulnerable to flooding may increase by 50% under projected 21st Century eustatic sea level rise, but thisis a conservative estimate given the current trends in the reduction in deltaic sedimentation that would otherwisebuffer deltas. Early indications suggest that the magnitude and frequency of hurricanes and cyclones may increasein coming decades, along with the onset of more intense precipitation events. All trends point to ever-increasingareas of deltas sinking below mean sea level. To keep the ocean off the landscape, coastlines are being strengthenedthrough coastal barriers of untested strength. Human occupation and infrastructure development continues, throughthe development of megacities and their expanding footprint on deltas. River engineering near river mouths canbe counterproductive: confinement of in-channel sediment load favors bed aggradation, even while areas behindprotective levees are sinking. Whereas humans have largely mastered the everyday behavior of lowland rivers, theyappear less able to deal with the fury of storm surges that can temporarily raise sea level by 3 to 10 m. It remainsalarming how often deltas flood, whether from land or from sea, and the trends appear to be worsening.

4


Arctic Ocean storm surges: origin, climatology, impacts, simulations andpredictionsA. Proshutinsky (1), S. Solomon (2), and I. Ashik (3)(1) Woods Hole Oceanographic Institution, Woods Hole, United States ([email protected], 508 289 2796), (2)Geological Survey of Canada, (3) Arctic and Antarctic Research Institute, Russia

Like many other manifestations of climate change, sea level rise is already a problem in the arctic regions. Manyscientists and engineers expect the effects of sea level rise to be profound and costly along the Arctic Ocean coasts.The current rate of sea level rise in the Arctic Ocean estimated based on nine tide gauge stations for 1954-2009 is2.57 ± 0.45 mm yr−1 (after correction for glacio-isostatic adjustment). The 2007 IPCC best-estimate scenario forthe world’s oceans projects a sea-level rise in the range of 18 to 38 cm by 2100, and likely to be substantially greaterthan the increase over the last century. With continuing arctic warming and sea ice declines it is expected that sealevel will rise and storms with storm surges will be stronger and more frequent and coastal communities nowstruggling with erosion will see shoreline retreat accelerate. The shore of the arctic seas are generally of low reliefand the combination of waves and high water levels during late summer and fall storms before the developmentof significant sea-ice cover can be particularly damaging to shorelines. Gravel barrier beaches can be overwashedand eroded while bluffs consisting of unlithified ice-bonded sediment and segregated ice can fail and retreat. Thelow-lying coast and delta plains are subject to extensive inundation causing, in some cases, severe environmentaland economic impacts. The presence of sea ice significantly influences the origin, development and progression ofstorm surges and their effects on environmental and socioeconomic conditions. While coastal erosion from stormsis limited during the winter season because of the protection from land-fast sea ice, storm surges of significancecan occur even under conditions of complete ice cover. Ice roads and other temporary infrastructure utilizingthe ice surface (e.g. hydrocarbon exploration equipment) are subject to damage during the winter surges. Thispresentation will focus on the description of arctic storm surge mechanisms, climatology of storm surge relatedsea levels, statistics of extreme storm surge events and their variability due to climate change. Specific attentionwill be paid to the techniques of arctic storm surge simulation, mechanisms of ice-tide-storm interactions andparameterizations, methods of sea level, ice drift, and water circulation prediction including extreme events.

5


Dynamical Downscaling of Storm Surges in South-China Sea andSingapore StraitP. Kolomiets (1), M. Zheleznyak (2), and P. Tkalich (3)(1) Ukrainian Center of Environmental and Water Projects, Kyiv, Ukraine ([email protected]), (2) Ukrainian Center ofEnvironmental and Water Projects, Kyiv, Ukraine ([email protected]), (3) National University of Singapore, Singapore([email protected])

A projection of sea level extremes is an important part of global climate change studies. In Singapore Strait theextreme sea level anomalies occur during Asian Northeast (NE) and Southwest (SW) monsoons blowing overthe South-China Sea (SCS). The extremes are believed to be associated mainly with storm surges exceedingmagnitude 50 cm relative mean sea level in the Singapore Strait due to the strong persistent monsoon winds offcoast of Vietnam (Tkalich et al., 2009a, 2009b). The paper uses past wind and storm surge data to validate theshallow-water COASTOX-UN model (Kolomiets et al. 2008), and then dynamically downscale future extremestorm surges using projected wind and atmospheric pressure from ECHAM5 global climate model (GCM).Past records from Singapore tide gauges and regional TOPEX/POSEIDON altimetry are analysed together withwind and pressure from NCEP data to identify most prominent storm surge cases. A criterion was found linkingextreme storm surges in Singapore Strait to wind parameters at certain areas of SCS; and then the criterion isused on future wind and atmospheric pressure data to identify probable dates of storm surges. Two 100-years-longwind data sets are used, one is the direct output from ECHAM5 GCM for SRES scenario A2, and another isdynamically downscaled using WRF regional climate model RCM.The wind data are fed into 2D shallow-water unstructured finite-volume model COASTOX-UN covering regionfrom -10◦S to 30◦N and from 90◦E to 130◦E. The mesh comprises of 184,000 elements, having cell sizes rangingfrom 100km at the domain boundaries, to 7km at the shoreline, and with refinement up to 1km near the SingaporeStrait. For past cases the computations showed a good agreement with measurements, with a little or no calibrationrequired. For future wind, projected storm surges show distribution consistent with natural variability rather thancertain trend. For extreme cases the maximum sea surface may reach height up to 1m in Singapore Strait.

References1. Tkalich P., Kolomiets K., Zheleznyak M. (2009a) "Simulation of wind-induced anomalies in the South-ChinaSea" Proc. AOGS 2009, Singapore.2. Tkalich P., Vethamony P., Babu M.T., Pokratath R. (2009b) “Seasonal sea level variability and anomalies in theSingapore Strait” Proc. ICOE 2009, Chennai, INDIA .3. Kolomiets P., Kivva S., Zheleznyak M. “2D unstructured shallow water model”, Proc. Mathematical andImitational System Modeling Conf., IPMMS, Kyiv, Ukraine (2008).

6


Storm surges on the coast of the State of Rio de Janeiro, BrazilC. NevesCOPPE - Ocean Engineering Program, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil([email protected])

Due to the absence of extreme meteorological events in Brazil, not much attention has been called upon stormsurges for the design of harbor structures, beach nourishment or drainage in coastal cities. Yet, discrepanciesbetween tidal predictions and tidal measurements as large as 0.90 m are quite often reported on the Southernportion of Brazilian coast, mostly in last two decades. Such figure is significantly higher than the expected sealevel rise predicted by IPCC for mean global sea level by year 2100. This fact should call the attention of cityplanners and environmental authorities for the urgent need of sea level monitoring along the coast of Brazil, takinginto account the need for anticipating storm surges for operational procedures of various industrial plants, fordesign of coastal structures or simply for urbanization of coastal areas.

Numerical models based on POM have not succeeded in forecasting storm surges along Brazilian conti-nental shelf, which may be attributed to the following reasons: lack of reliable wind data for the forcingmechanism, poor topographic description, overlooking of some physical relevant phenomenon, such as barocliniceffects. For this reason, a series of studies were concentrated along the coast of the State of Rio de Janeiro, inorder to establish correlations between observed storm surges from tidal records, NCEP reanalysis data at selectedgrid points, and meteorological observations at land stations.

The data available consisted of one year of sea level, atmospheric pressure and wind velocity at hourly in-tervals, and NCEP/NCAR data every 6 hours. Three stations were considered, Piraquara, Rio de Janeiro andMacaé. For each location, 4 NCEP grid points were investigated, in order to determine which point showed bettercorrelation with the data from the meteorological stations. At each grid point, wind stresses were computed. AtRio de Janeiro station, tidal records covered a time span of 18 years.

At first, a low pass filter was used in both meteorological and oceanographical data, in order to keep oscil-lations with periods longer than 60 hours. In time domain, the statistic of permanence of sea level was obtained,while in frequency domain, cross correlation studies were carried out between filtered sea level and pressure, andbetween filtered sea level and longshore and normal to shore components of wind stresses. Similar cross-spectralanalysis was carried out with wind components at NCEP grid points and data from meteorological stations.Finally, long term data from Ilha Fiscal tidal station (Rio de Janeiro) was used to obtain the statistics of stormsurges.

Coherent oscillations in meteorological NCEP data and sea level, at well defined periods ranging from 5 to13 days, were found at all stations. Meteorological observations at coastal stations correlated poorly bothwith observed storm surges and with NCEP reanalysis data. The farther away the NCEP data, however, thebetter the spectral correlation with the storm surge, indicating the large scale nature of the phenomenon andsuggesting that future observations of wind should be made offshore. In addition, the study motivates furtherinvestigation on whether the observed surges are a purely local phenomenon or result from long waves propagation.

The paper finally addresses possible engineering responses which should be considered by city planners,mainly at locations in the State of Rio de Janeiro.

7


Developing a storm surge typology for low latitude coral atolls and islandsin the Indian and Pacific OceansR. McLEANRoger McLean, School of Physical, Environmental and Mathematical Sciences, University of New South Wales at theAustralian Defence Force Academy, Canberra ACT 2600 Australia([email protected])

BackgroundIn low latitudes (100N-100S) storms of hurricane/typhoon intensity are not common. And yet, coral rubble ridgesand reef blocks emplaced by storm waves and surges are persistent morphological features on many low-lying atollislands and reefs in equatorial regions. The purpose of this study is to examine this paradox based on extensivefield surveys in Tuvalu central Pacific, and the Maldives central Indian Ocean.

ObjectivesThe general aim is to develop a storm surge typology for coral atolls and reef islands. Specifically the objectives are:(1) to identify the provenance of storm waves and surges in these two atoll states; (2) to determine the magnitudeand frequency of surge events and assess whether previous models used to estimate palaeosurge-event magnitude(based on ridge elevation) is a valid methodology; (3) to investigate the immediate and long-term bio-physicalimpacts on islands, and their role in island destruction and construction; and, (4) to examine how communities andgovernments have adapted to surge events in the historic and recent past and what they are likely to do in the future.

MethodsThe identification and mapping of storm surge deposits using geomorphologic and sedimentologic criteria;morphostratigraphic analysis and radiometric dating of storm deposits to determine palaeosurge-event frequencytogether with analysis of historical and contemporary events to develop a long-term record of surge-events andhuman response in Tuvalu and the Maldives; sequential and on-going monitoring of a massive coral rubblerampart deposited during Hurricane Bebe on Funafuti atoll Tuvalu in October 1972; and, repeated island-beachand reef surveys prior to and following the swell-surge event during May-June 2007 in the Maldives. Both of thelatter serve as modern analogues of past surge events, and of how communities and governments have adapted tosuch events.

ResultsThree surge types impact these and other atoll islands in equatorial regions. These are: (1) infrequent tropicalcyclones generated within low-latitude boundaries; (2) waves and surges that penetrate into the region fromtropical storms and cyclones generated in the adjacent hurricane/typhoon belts to the north and south; and (3)sea-swell and surge set-up that results from extra-tropical and high-latitude low pressure systems, originatingthousands of km from the atoll islands. Process characteristics of these events are described and their physicalsignatures and human impacts assessed in order to develop a storm surge typology for atoll islands.

ConclusionsA storm-surge typology has been developed for low-latitude coral islands and atolls. Three types of storm surgeare described. Each type has a different physical impact, human response and risk profile. Whilst the immediateevent is generally destructive, storm surges have an important role in island formation and development. Low-lyingcoral atolls and reef islands have been identified as extreme examples of the potential negative impacts of climatechange and sea-level rise, with the atoll states of Tuvalu and the Maldives often being cited as the most vulnerable.Despite such concerns the most common catastrophic natural events that have impacted these states in the last 100years or so have been associated with storm surges, and this is also likely to be the case in the future.

8


Storm Surges in the Elbe Estuary - Analysis and Sensitivity StudyE. Rudolph and A. Schulte-RentropBundesanstalt fuer Wasserbau, Dienststelle Hamburg, Hamburg, Germany ([email protected])

Due to changing climatic conditions several parameters influencing water levels during storm surges are expectedto change. In order to find a strategy for adaption to climate change it is important to understand today’s situationand analyse the future situation under the influence of climate change. This concept will be presented by lookinginto storm surges in the Elbe estuary which hosts the waterway from the North Sea to the port of Hamburg. Theseinvestigations are embedded in the scientific joint venture programmes KLIWAS and KLIMZUG Nord.

Using a hydrodynamical numerical model (UNTRIM, V. Casulli and R. A. Walters (2000)) to calculatewater levels and currents in the Elbe and a meteorological model (MKW, H. Schmidt und J. Pätsch (1992))to provide the local wind field over the Elbe allows not only to understand historical storm surges but also toinvestigate probable future storm surge scenarios. Two historical storm surges with different characteristics are thebasis of the sensitivity study:- the storm surge of January 3rd 1976 which caused the highest observed water levels along the Elbe estuary and- the storm surge of January 28th 1994 which had a high fresh water discharge.

The development and intensity of a storm surge in the Elbe estuary is determined by the water level at theboundary to the North Sea, the river runoff into the estuary and the wind field over the estuary. Based on asensitivity study the influence of- sea level changes in the North Sea,- increase / decrease of river runoff into the estuary, and- increase / decrease of the local wind over the estuaryon the highest water level along the Elbe estuary during storm surge are analysed. The parameters mentioned arevaried according to the knowledge about expected changes in a future climate.

The aim of this investigation is to get a better understanding of the probable changes of water levels dur-ing storm surges along the Elbe estuary depending on parameters that might change in a future climate. The resultswill help to identify vulnerabilities of e.g. the shore protection along the Elbe estuary and give us a chance to workon adaption and risk mitigation necessitated under possible climate change.

Casulli, V. and Walters, R. A. (2000). An unstructured, three dimensional model based on the shallow wa-ter equations. International Journal for Numerical Methods in Fluids, 32, S.331 – 348.

Schmidt, H. und Pätsch, J. (1992). Meteorologische Messungen auf Norderney und Modellrechnungen. DieKüste, 54, S. 131 – 142.

9

Storm Surges Congress, Hamburg, Germany13–17 September 2010SSC2010-63-1© Author(s) 2010

Impact of wind gusts on sea surface height in storm surge modelingR. van der Grinten (1,2), H. de Vries (1), and H. de Swart (2)(1) KNMI, Weather/Research, De Bilt, Netherlands ([email protected], +31 30 2210407), (2) Institute for Marine andAtmospheric research, Utrecht University, Netherlands

Storm surges are subject of great interest to low-lying countries adjacent to oceans. In the Netherlands, modelingof storm induced surges has been performed ever since plans for a closure dike on the Zuiderzee in the early1920s. Over time, models have become more advanced and surges are better predicted. The shallow water modelWAQUA/DCSM (Dutch Continental Shelf Model) is used in The Netherlands for operational forecasting of sealevel heights along the Dutch coast. The meteorological input for the model is hourly averaged wind and pressurefrom HiRLAM (High Resolution Limited Area Model).The effect of high-frequent variations in the wind speed is thus far not taken into account explicitly. But asthe surface wind stress that drives the model is determined by the square of the wind speed, such variations docontribute to the total driving forces. A better understanding of the impact of these variations on a storm surge isdesired.To assess the importance of these high-frequent variations, gustiness values, obtained from the ECMWF model,have been compared to the difference between sea level observations and operational model forecasts alongthe Dutch coast. To get a measure for the gustiness averaged over the North Sea, a district method is used thataccounts for a time lag and wind direction for specific parts of the North Sea. An increasing bias between modeland observations is found for increasing averaged gustiness. This is evidence for the impact of wind gusts on thestorm surge height.To understand more about the way that variations influence the sea level height, a more theoretical case has beeninvestigated. Statistical variations have been added to a uniform wind field over the North Sea. These wind fieldshave been used to drive DCSM. The resulting sea levels are compared to those with undisturbed wind. Ensemblesof wind disturbances in speed and direction have been generated with different temporal and spatial correlation.Added variations in wind speed (i.e. gustiness) enhance a surge whereas variations in wind direction reduce asurge. These properties have to be taken into account for storm surge modeling.

10


Extratropical cyclones as a driver of wind wave extremes over the lastcenturyS.K. Gulev, I. Rudeva, and V. GrigorievaIORAS, SAIL, Moscow, Russian Federation ([email protected])

We analyse decadal scale variability of the characteristics of extratropical cyclones in the North Atlantic – Euro-pean sector using results of numerical storm tracking of NCEP-NCAR and ERA-40 reanlayses for the last 50-60years and of 20 century reanalysis for the last 100 years. Cyclone activity is characterized by cyclone numbersand frequencies as well as by numerous characteristics of the cyclone life cycle such as propagation velocities,deepening rates and intensity. For the off-shore European regions we performed a composite analysis to quantifyconditions associated wit the most extreme cyclones. There has been identified growing number of moderately deepand extreme cyclones which affect Northern Europe during the last decades. In the next step we associated leadingmodes of cyclone activity and characteristics of the cyclone life cycle with the parameters of extreme wind wavesin the off-shore European regions. Wind wave characteristics were derived from 100-year archive of visual windwave observations. Extreme wave statistics were quantified through the application of IVD and POT approaches tothe analysis of probability density functions. Joint consideration of cyclone activity and wind wave characteristicsallowed for the identification of wind wave patterns associated with extreme cyclones and for accurate capturingof synoptic patterns leading to extreme storms in the North Atlantic including off-shore European regions. Someinsights on the impact of wind wave storminess on European storm surges are provided.

11


Statistical Downscaling of Storm Surges in Singapore StraitP. TkalichNational University of Singapore, Singapore, Singapore ([email protected])

Projection of storm surges for the 21st century is an essential part of global climate change science. Unlike Re-gional Climate Models (RCMs), coarse computational grids of current Global Climate Models (GCMs) can notresolve sea level extremes around small islands, such as Singapore; therefore, application of various downscal-ing techniques is required. One of the available options (von Storch and Reichardt, 1997; IPCC, 2007) is to usestatistical downscaling method for estimation of future storm surges, where an empirical relationship between his-torical large-scale synoptic conditions (such as wind speed and atmospheric pressure) and observed local sea levelextremes are constructed; and then utilised for assessment of future sea level extremes using synoptic parametersprojected by GCMs. Additionally, downscaling of driving synoptic parameters with RCMs is required in manycases.

Comparison of tidal gauge data in Singapore Strait versus NCEP wind offshore of Vietnam shown that there is asignificant correlation of sea level anomalies (SLAs) in Singapore Strait on wind in the South-China Sea. The dailymean sea level in Singapore Strait may increase gently for a few days up to 50-70 cm due to the remote strongnorth-east (NE) wind. These cases are more frequent during NE monsoons (December-January) characterised by apersistent wind often exceeding 15 m/s (6-hourly). The empirical formulae relating storm surge height to the windspeed offshore of Vietnam is fitted using power function.

Ten GCMs, including BCCR, CCCMA, CNRM, CSIRO, GFDL, IPSL, MIUB, MPI and MRI, have been used tobuild distribution functions of wind speed projections for the specific region of the South-China Sea. The winddistribution functions have been converted into storm surge distribution functions using the developed empiricalrelationship for past storm surges. There is an increase of percentage of 50cm SLA between 1 and 3% fromoverall number of SLA events as well as existence of small but non-zero probability of 1 m storm surge. Since, theprojected storm surges have been driven by GCM-projected wind, and the wind projections do not show a consistentupward trend, the increased number of high storm surge events could be attributed to a natural variability ratherthan to the global climate change.

References

1. IPCC, 2007: Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to theFourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M.Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge UniversityPress, Cambridge, United Kingdom and New York, NY, USA, 996 pp.

2. von Storch, H., and H. Reichardt, (1997) A scenario of storm surge statistics for the German Bight at theexpected time of doubled atmospheric carbon dioxide concentration. J. Clim., 10, 2653–2662.

12


Return periods of extreme sea levels along the Ravenna coastlineM. Masina (1) and P. Ciavola (2)(1) Department of Earth Sciences University of Ferrara, Earth Sciences, Ferrara, Italy ([email protected]), (2)Department of Earth Sciences University of Ferrara, Earth Sciences, Ferrara, Italy ([email protected])

The low-lying coastal areas of the North-western Adriatic are exposed to a high degree of inundation risk byexceptional sea levels due to storm surges. Although the phenomenon of “ extreme high water” (acqua alta) is wellknown and described in Venice, justifying the construction of the MOSE flood gates to protect the lagoon and thecity against flooding, it remains poorly described for other Italian coastal sites.

The aim of the present investigation, conducted within the European Union funded MICORE Project (FP7contract 202798), is the estimation of extreme water levels and their associated return periods along the vulnerablecoastal areas of the Ravenna Province. A better knowledge of the phenomenon can serve as starting point for animproved Civil Protection practice, aiming at developing adequate forecasting and warning systems.

The observations available for Ravenna are measurements at 10-minute interval, recorded by the PortoCorsini tide gauge station, which is part of the National Sea Level Measurement Network. The studied time seriescover a period of ten years, from 2000 to 2009. Following a preliminary quality control of the dataset, a strongcorrelation was found with the hourly sea level data collected by the Centro Previsioni e Segnalazioni Maree of theVenice Municipality for the tide station of Punta della Salute. It can be concluded that there is enough robustnessin the Ravenna dataset to ensure the validity of further probabilistic analyses.

The methodology employed in the analysis was based on the limiting joint Gumbel distribution of the r-largest annual events. The seven largest values of sea level and surge amplitude were extracted for each year fromthe quality checked records of observations. Tidal residuals were computed using the program t_tide to obtainastronomic predictions and subtracting this value from the observed time series. The selection process for an eventconsidered as independent only those separated by at least 78 hours. This is supposed to have eliminated the effectof seiche oscillations that may follow storm surges. Further validation of the representativeness of the datasetcame from its cross-comparison with time series of meteorological parameters and wave measurements collectedat several of the nearest stations along the Adriatic coast.

The analyses identified extreme levels of 0.85 m, 1.05 m, 1.28 m for return periods of 2, 10 and 100years. The equivalent computed residuals were 0.61 m, 0.79 m, 1.02 m. To notice that for a landscape like theRavenna coastal zone, where the + 1 m contour is often located 1 km inland, even an event with a moderatelikelihood can produce disastrous flooding.

Future work will include the identification of representative prototype surge characteristics which will beused to assess the vulnerability of dune ridges, undertaking a detailed sensitivity analysis of the surge parameter.The identified critical thresholds will be included in an operational warning system which is being developed inthe framework of the MICORE Project (www.micore.eu).

13


Global Vulnerability Assessment on Storm Surges due to TropicalCyclones in the 21st CenturyH. Nobuoka (1) and N. Mimura (2)(1) Urban and Civil Engineering, Ibaraki University, Hitachi, Japan([email protected]), (2) Center for WaterEnvironment Studies, Ibaraki University, Hitachi, Japan([email protected])

Coastal zones are one of the severely vulnerable areas in the 21st century, because of increasing impact of sea-level rise according to global warming, and population growth. Though Hoozemans et al.(1993), Robert Nicholset al.(1999), Nobuoka et al.(2007) and their later works carried out the basic assessment of coastal vulnerabilityin the world, the accuracy of impacts and vulnerability assessments are not enough, especially about storm surgesand trend of increasing in the affected population. This paper shows increases in flooded population by stormsurges due to tropical cyclones during the 21st century by use of numerical simulations, with sea-level rises andpopulation growth according to Special Report on Emission Scenarios (SRES).Hindcasts of storm surges on 4544 tropical cyclones in the world were implemented by use of numerical simulationof shallow water theory, of which spatial resolution was 2 arc-minutes. In addition, a method of extreme statisticsanalysis was employed to calculate storm surge return periods in all of the coasts. Global maps of maximum ofthe storm surges deviation and that of 100 years return period due to tropical cyclones were made. For impactassessment, the two types of these storm surges were downscaled in 1 arc-minute. Total impacts in the coasts werethese storm surges with mean high water spring of astronomical tide calculated by ocean tide model (Nao99b) andtwo scenarios of rise in sea level projected by general circulation model of Meteorological Research Institute ofJapan and Tokyo Universities. Comparing between impact levels and altitudes of digital elevation model of author’smodified SRTM30 ver. 2.1 (Shuttle Radar Topography Mission), considering passes from coastal lines, projectedthe flooded areas during the 21st century. Distribution of population for counting the population in the flooded areawas based on GPW ver.3 and downscaled population growth following SRES scenarios in each countries providedby CIESIN of USA.The main results are as follows. As hindcasted storm surges are good agreement with measured/observed data,the global maps of the storm surges presented by this study are useful for global cooperation of prevention ofstorm surges. In the A2 scenario of SRES, which is worst scenarios, the increase in potential flood area andflooded population along the coast in the world by storm surges in 2100 becomes approximately 180,000 km2and 600 million people, respectively. On the other hand, in A1B and B1 scenarios, the increase in the floodedpopulation becomes a peak around the year of 2050, and this number is going to decrease by the year of 2100.The negative population growth of these scenarios from 2050 in Asia contributes to these global results of theaffected population much. Impact of sea-level rise on the relative increase in the flooded population becomes clearafter 2075 in Oceania and South America. These distinctions of projected-results in regions and SRES scenariosconfirm us that we should select primary impact, which is population growth and/or sea-level rise in each region,for ‘wise adaptation in coastal zones’ in the 21st century.

14


Potential impacts of extreme storm surges on a low-lying denselypopulated coastline: The case of Dunkirk area, Northern France.A. MASPATAUD (1,2,3), M-H. RUZ (1,2,3), and S. VANHEE (4)(1) Univ. Lille Nord de France, F-59000 Lille, France ([email protected]), (2) ULCO, LOG, F-62930Wimereux, France, (3) CNRS, UMR8187, R-62930 Wimereux, France, (4) Institution Interdépartementale des Wateringues,62505 Saint-Omer, France

Along the southern coast of the North Sea, a large proportion of the Flemish coastal plain consists of denselypopulated reclaimed land, much of which lying below sea level. This is the case along the northern coast ofFrance, from Dunkirk to the Belgium border, where the shoreline consists of coastal dunes that protect low-lyingreclaimed lands from marine flooding. Coastal dunes have been massively transformed or destructed due tothe urbanisation and to the development of the port of Dunkirk. Along this coastline, the beach is subject to amacro-tidal range that increases from 3.5 m during mean neap tide conditions to 5.4 m during mean spring tides.The coastline is dominantly exposed to offshore to shore parallel winds from a south to southwesterly window.Although this coast is exposed to fetch-limited, relatively low-energy waves, strong onshore winds associated withlow atmospheric pressure frequently result in storm surges (1-1.5 m) responsible for upper beach/dune erosion,and polder drainage problems. Indeed, to prevent the polder from being waterlogged, drainage canals and pumpsdrain excess water from the polder and discharge it directly into the sea at low tide.In November 2007, strong (> 10m/s) direct onshore winds, blowing persistently during 48 hours, combined withan atmospheric pressure decrease of 23 hPa, resulted in an extreme surge with a residual tidal height maxima of2.4 m. This surge, the largest event measured since the 1953 storm surge, fortunately occurred at low tide, duringmoderate tidal range conditions. At high tide, although the surge was less than 1.2 m, it nevertheless induced dunescarping and seafront promenade submersion. During this storm event, the drainage of the polder could not beoperated during 24 hours, because the maximum water level over which gravity drainage can not be performedwas constantly exceeded even at low tide. The peak of the storm surge occurred nearly 5 hours before high tide. Ifthe storm surge took place only a few hours later, the consequences could have been much more severe. Damagescan also occur during moderate surges however, when they are combined with a spring tide, as what happened,when a 1.1 m surge developed in March 2007.In a context of global change and projected sea level rise, such events could induce major damages in the nextfuture, especially if an extreme surge occurs at high tide during a spring tide. Large proportions of the urbanseafront could be submerged. Furthermore, coastal dunes, which act as natural ramparts against marine flooding,could dramatically retreat and the polder could be inundated. Stakeholders will have to face increasing challengesin warning and protecting coastal communities against marine submersion and polder flooding. It is then importantfor the local authorities to take into account the potential impacts of such events, in order to reinforce sea defence,increase pumping station efficiency and plan warning systems, witch is it not the case yet.

15


The interplay between waves and surges in the Dutch Wadden SeaG.Ph. Van Vledder, J, Adema, and O.R. KoopAlkyon Hydraulic Consultancy & Research

The Dutch Wadden Sea is a shallow shelf sea connected to the North Sea by a number of tidal inlets. The mainlandshore consists of dykes protecting the low-lying areas of the northern Netherlands. Tide and extra-tropical stormsresult in hydraulic loads on dykes through severe surge and wave action. Improved understanding of the physi-cal processes responsible for wave, current and water level conditions is essential for the design of such coastaldefenses. This knowledge is required for the 5-year mandatory check to assess the existing level of protection.

Lipari and Adema (2010) address the generating mechanisms of severe surges in the Wadden Sea, whereas thepresent contribution focuses on the wave conditions in the Wadden Sea and on the interplay with surges. Wavesin the Wadden Sea are mainly locally generated and strongly affected by shallow-water and current effects. Inaddition, some waves generated in the North Sea penetrate into the Wadden Sea through the tidal inlets. Becauseof the shallowness, accurate water level predictions are essential for the proper computation of the waves, as wavesare often depth-limited. In turn, wave set-up significantly affects surge levels in the Wadden Sea.

The modeling of waves in the Wadden Sea is rather complicated due to its geometry consisting of tidal channelsand extended tidal flats. Various aspects thereof will be highlighted here.

Firstly, the penetration of North Sea waves into the Wadden Sea is addressed. These waves are strongly affected bythe shallow ebb-tidal deltas where severe wave breaking occurs, causing an abrupt drop in wave height. In addition,refraction effects in the tidal channels turn most of the wave energy to the shoulder’s sides of these channels wherethe waves are dissipated. Only in the wider tidal inlets of the eastern Dutch Wadden Sea a relatively large amountof wave energy is able to penetrate farther up to the dykes. The amount of wave penetration is also affected bycurrent effects, and noticeable differences in wave conditions occur for following and opposing currents.

Secondly, the effect of waves on the surge levels was computed and compared with water-level measurements forthe storm of 9 Nov 2007. The numerical simulations were performed with the Delft3D flow solver of Deltaresin combination with the SWAN model of TU Delft. Various numerical experiments were performed to optimizethe coupling of these models. The results of these two-way coupled simulations showed that wave effects couldcontribute with up to 0.40 m to a peak surge level of about 3 m; but also that inclusion of the wave effect wasrequired to accurately (error less than 0.05 m) predict the water levels, which in turn affect the wave conditions inshallow water.

Thirdly, numerical experiments show that, during severe storms, the wave-current interaction is not limited to thetidal channels, but that it may cover the entire Wadden Sea. Depending on the characteristics of the storm (Lipariand Adema, 2010), this can result in a large-scale flow through the Wadden Sea more or less aligned with thedirection of the wind. Then, the waves encounter a large scale following current yielding significantly lower waveheights and periods.

Finally, the hindcast and numerical experiments show that the largest-wave conditions seldom occur simultane-ously with the water-level peak. Competing mechanisms are identified that lead to depth-limited high wave condi-tions.

The results of this study highlight the importance of including current effects in the derivation of Hydraulic Bound-ary Conditions for the design of the coastal sea defenses. Insights obtained in this study will be explored in furtherstudies.

The research was funded by Rijkswaterstaat – Centre for Water Management – on behalf of the Ministry for

16

Transport, Public Works and Water Management. This funding was realized in the framework of the Dutch re-search program “Strength and Loads of Water Defenses” (SBW). Wind fields were provided by the Dutch RoyalMeteorological Office.

Reference:

Lipari and Adema, 2010: The generation of severe surges in the Dutch Wadden Sea: insights from data, hindcastsand numerical experiments. Abstract submitted to this conference.

17


The Risk and Calculation of Extreme Storm Surges due to ClimateChangeK. Sossidi, G. Gönnert, B. Gerkensmeier, and Th. BußFree and Hanseatic City of Hamburg, Agency of Roads, Bridges, and Waters, Hamburg, Germany([email protected], fax +49 40 427949 164)

The North Sea coast is seriously threatened by storm surges. Climate change and resulting consequences as sealevel rise and possible intensification of storm surges will have serious effects on the safety of people and economicvalues in coastal areas. The uncertainties of climate change necessitate new concepts of coastal protection.

The Project XtremRisK – Extreme Storm Surges at the North Sea Coast and in Estuaries, Risk Calculationand Risk Strategies, funded by the German Federal Government, will assist in facing this challenge. The„Source-Pathway-Receptor“-Concept will be used as a basis for risk analysis and development of new strategies.The project results will be exemplarily applied and further developed at two pilot sites, the island of Sylt and thecity of Hamburg.

The objective of this paper is to determine the relevant extreme events. Within the XtremRisK project,methods will be developed to assess the extreme events under the conditions of today. Starting from there, extremeevents will be assessed using conditions which reflect the climate change scenarios. The presentation will give anoverview about methods and results of the calculated extreme events.

A couple of methods to calculate an extreme storm surge exist, which are often statistical methods. Here, amethod will be used which takes the physics of storm surges into account. With this method, the three componentsof a storm surge, tide, wind surge and external surge, will be analysed and their development over the last 100years will be assessed.

The paper focuses on analysing (i) the highest event of each component and (ii) the interaction betweentide and surge and the interaction between surge and external surge. This detailed analysis is needed because thecomponents do not interact linearly. The non-linear interaction between the components require the considerationof the hydrodynamics. With this knowledge, an extreme event based on the maximum components can becalculated. Investigation on the hydrodynamics and physics of storm surges show that the components have to beadded non-linearly, which leads to a lower water level than in case of linear superposition. With this, a realisticextreme event under present climate conditions can be calculated.

A long consistent time series of water levels is very important for the characterisation of storm surges.Consequently the extreme storm surge for the city of Hamburg will be determined at tidal gauge Cuxhaven and itspropagation upstream to Hamburg subsequently calculated by numerical modeling. The storm surge for Sylt willbe determined at the tidal gauge Hörnum.

Acknowledgement

The project is funded by the German Federal Ministry of Education and Research BMBF (Project No. 03F 0483 C).

18


Human Impacts on Estuarine Storm Surge LevelsH. D. NiemeyerCoastal Research Station Lower Saxony Water Management, Coastal Defence and Nature Conservation AgencyNorderney/East Frisia, Germany ([email protected])

1 Introduction

In the German North Sea estuaries of Ems, Weser and Elbe significant changes of ordinary tidal and storm surgelevels have occurred after the second World War. The public and scientific discussion on the reason(s) for thesedevelopments are still go-ing on. It is often steered by interests of stakeholders, blaming partly coastal protec-tionmeasures and partly deepening of estuarine waterways. Aim of this contribution is to analyze the data in order tosubstantiate the discussion. The means of the analysis are explanatory highlighted in the abstract for one of thethree estuaries.

2 Very High Storm Surge Levels in the Ems-Dollard estuary since 1962

The storm surge levels of February 1962 are effected by a then still open dyke line in the upper part of the estuaryallowing the flooding of large areas. After the closure of that gap in 1967 the storm surges of January 1976, ofJanuary 1994 and January 1995 have the same boundary conditions above MHWL. The only changes in the periodbetween 1976 and 1994 have successive deepenings and streamlining of the estuarine waterway between Pogumand Papenburg. The effect on the peaks of storm surges is evident. E. g. the storm surges of January 1995 andparticularly of January 1994 have relatively high peaks in comparison to that of January 1976 and between Emdenand Leerort also to that of February 1962 though their peaks at Borkum in the mouth of the estuary have beensignificantly higher. These changes in the upper part of the estuary is linked with the waterway deepening andstreamlining there. The peaks of the storm surge of November 2006 are mainly determined by the closure of theEms storm surge barrier: The peak levels downstream of the barrier are higher than those of 1962 and 1976 inrelation to the ones at the estuarine mouth.

3 Change of threshold level in the Weser estuary

The statistical classification of storm surge levels DIN 4049 defines the lowest thresh-old for storm surges by thelowest of the 200 highest set-ups within 20 years. This threshold has climbed in the Weser estuary from the period1968-87 to 1980-99 at the gauge Elsfleth from 112 cm to 120 cm and at the gauge Oslebshausen from 117 cm to132 cm. Major impact is the deepening of the Lower Weser in the 1970s. The erection of storm surge barriers atthe mouth of three tributaries of the Weser is of low importance since their closure is only executed for a muchsmaller number of higher storm surge levels.

19

4 Percentiles of set-up in the Elbe estuary

The 95- and 99,5-percentiles of the set-up storm surges at the tidal gauges of Cux-haven, Stadersand, St. Pauli andOver increase significantly between 1964 and 1975 and again after a slight lowering between 1978 and 1985. Theincrease is much more pronounced at the upstream located gauges than at Cuxhaven at the mouth. Contra-dictorythe lowering of the percentiles does show such significant differences be-tween the data for the gauge Cuxhavenand those ones for the upstream located gauges. This indicates that the reason for these changes are the subsequentdeep-enings to a nautical depth of 12 m between 1964 and 1967 and to 13,5 m between 1974 and 1978.

20

» Driving factorsand scales of storm surges «

»STORM SURGES CONGRESS 2010« 13th - 17th September, 2010

Risk and management of currentand future storm surgesUniversity of Hamburg, Germany

»POSTER«


Relationship between moon phases and ocean surges in the city of Lagoso. ediang (1) and m Makinde Ibraahim (2,3)(1) nigeria meteorological agency, research and trainning, lagos, Nigeria ([email protected]), (2) meteorologydepartment, federal university of technology,akure,ondo Nigeria, (3) Nigerian Maritime Authority

The moon phases we have over the earth do have effects on ocean tides,the rising and falling of ocean waters.Thisfact has been established.Ocean surges is more emphatic than ocean tides in terms of it’s effects on coastal areas-on life and property.As such,this study is to see if there is a relationship between the moon phases and oceansurges. Ocean surges have been a major problem to most coastal countries and often contribute to environmentaldisaster.This is mainly experienced in many of the coastal areas of the world.In this study,some meteorological parameters of the ocean are examined;some of them are velocity of wind movingthe ocean waters,sea surface temperature,percentage of the moon illuminated,moon and sunrise times,transit timesand set times for some specific dates from 1994 to 2007.Statistical analyses of the data gives rise to a clear relationship through graph plotting. The data of moon phases(percentage of moon illumilled) is a daily data, as well as the wind, temperature,etc.This paper conclude with an attempt to investigate the relationship between moon phases (in the period of 1994 to2007) and intensity of ocean surges in the atlantic ocean.

22


Sea level rise in the German Bight as one of the main contributors tochanging storm surge statisticsT. Wahl, T. Frank, and J. JensenUniversity of Siegen, Research Institute for Water and Environment, Siegen, Germany ([email protected], +49 (0)271 740 3462)

Sea Level Rise (SLR) is one of the major consequences we are facing in times of a warming climate. Differentanalyses and recent scientific publications identify mean sea level rise as the main driving factor for increasingstorm surge heights over the last century (e.g. Haigh et al., 2008), resulting in a higher risk of inundation forthe affected coastal areas. Although storm surge heights in the German North Sea area are expected to increaseover the next century due to changes in the wind statistics (Woth et al., 2006), the projected mean sea level risewill remain one of the major contributors. Therefore, regional and global sea level rise are subjects to manyongoing scientific research efforts. Improved estimates of sea level change over the last 165 years in the GermanBight are the outcome of the German research project AMSeL. Different methods are used to generate andanalyse observed mean sea level time series from tide gauges providing high quality data sets. Parametric fittingapproaches as well as non-parametric data adaptive filters, such as Singular System Analysis (SSA) are applied.For padding non-stationary sea level time series, an advanced approach named Monte-Carlo autoregressivepadding (MCAP) has been introduced (Wahl et al., 2010). A ‘virtual station’ is estimated from 15 single stationscovering the entire German North Sea coastline. The estimated linear trend (of relative mean sea level) for theentire period is 1.3 ± 0.15 mm/a, 1.6 ± 0.2 mm/a from 1901 onwards and 3.4 ± 1.4 mm/a for the reducedperiod since 1970 (quoted errors are 2σ-standard errors). Additionally, the reconstructed mean sea level timeseries for the German Bight shows different patterns of sea level change than global reconstructions, which isnot fully understood up to now. Detailed examination of the datum and height changes of respective coastalareas and the gauge sites over time is a crucial part of the overall assessment and a potential source for uncertainties.

In the German Bight, a period of positive acceleration of sea level rise occurred at the end of the 20th cen-tury, followed by a period of negative acceleration. Another still ongoing period of positive acceleration startedaround 1970, leading to high recent and current rates. The latter are in the order of 5 mm/a and slightly higherthan those observed around 1900. However, it will be essential to keep in view the sea level observations withinthe next years, to prove whether a period of negative acceleration will take place again or whether the period,with its start point around 1970 denotes the beginning of an accelerated sea level rise due to anthropogenic impacts.

Furthermore, significant differences exist between the estimated rates of sea level rise for the eastern andthe southern part of the German Bight, indicating stronger land subsidence along the eastern part.

This is a German Coastal Engineering Research Council (KFKI) project, funded by the German FederalMinistry of Education and Research (BMBF) (Project No. 03KIS068).

23


Comparative Study of the Characteristics of Annual Maximum SurgeHeights along the Eastern and Southern Coasts of KoreaS.J. Kwon (1), E. Lee (1), I.J. Moon (2), and S.M. Oh (2)(1) Korea Hydrographic and Oceanographic Administration, Incheon, Korea([email protected] / 82-32-891-3773), (2)Department of Marine meteorology, Cheju National University, Jeju, Korea([email protected] / 82-64-756-3483)

There has been growing interest in how the storm surge height is changing due to typhoons associated with globalwarming. During the past 56 years, the number of typhoons affecting on the Korean Peninsula (KP) was about 3.3per year, and approximately one of them made landfalls over the KP. The intensity of the KP-landfall typhoonsis increasing continuously, which produces more severe damages in Korea. In 2003, the typhoon Maemi (2003)produced 85 deaths and total property damage of about 5 billion US dollars. This study investigates and comparesthe characteristics of the long-term variation of the annual maximum surge heights (AMSH) in the eastern andsouthern coasts of Korea, focusing on the global warming and the resultant change of typhoon and storm surgeintensity. The used data are the sea level records at Sokcho, Mukho, Pohang, Ulsan, Busan, Tongyoung, and Yeosualong the KP coasts over 34 years. In this study, the hourly surge data at 7 tidal stations are constructed throughthe tide filtering and data corrections.Based on the linear regression, it is found that the long-term uptrends in the AMSH of the eastern coast are8.3cm/34yrs at Sokcho, 8.7cm/34yrs at Mukho, 12.1cm/34yrs at Pohang, and 9.9cm/34yrs at Ulsan, while those inthe southern coast are 11.2cm/34yrs at Busan, 33.6cm/31yrs at Tongyoung, and 34.5cm/34yrs at Yeosu, which arerelatively higher than those in the eastern coast. It is also seen that the average and maximum values of AMSH inthe southern coast are obviously higher than those in the eastern coast. The statistical analysis reveals that the 53%of the AMSH in eastern coast and the 68% of the AMSH in southern coast occur during the typhoon period. It isshown that the uptrend in the AMSH seems to be attributed by the increased intensity of typhoon that influencesthe KP with climate change. The continuous efforts monitoring and predicting the extreme surge events in thefuture warm environments will be required to reduce the growing storm surge damage by the intensified typhoonin Korea.

24


Long-term variations of typhoon intensity and storm surge along theKorean coastsS. M. Oh (1), I.-J. Moon (1), S. J. Kwon (2), E.-I. Lee (2), and J. S. Shim (3)(1) Department of Marine Meteorology, Cheju National University, Jeju, Korea, (2) National Oceanographic ResearchInstitute, Incheon, Korea, (3) Korean Ocean Research & Development Institute, Ansan, Korea

Tropical cyclone (TC) best track data from the Joint Typhoon Warning Center (JTWC) and Regional SpecializedMeteorological Center (RSMC) during 1975-2005 have been analyzed to investigate a long-term intensity variationof typhoons which made landfall over the Korean peninsula (KP). It is found that the central pressure of TCs madelandfall over the KP is decreasing during 36 years persistently. The linear regression results reveal that the intensityof the landfall TCs has been dropped by 24 hPa on the average since 1970. Pressure and wind speed data observedat metrological stations show a similar trend that supports the growing intensity of the KP-landfall TCs found inthe best track data

From the analysis of observed sea level data during 45 years, it is also found that the intensified TC in the KPproduces high storm surge along the coasts. The hourly surge data at Busan, which are constructed through the tidefiltering and data corrections, shows a clear uptrend with 15-cm rise averagely during 45 years. It also appears thatthe largest storm surge events are mostly caused by recent record-breaking typhoons. Considering the mean sealevel rise about 2.5 mm per year at Busan due to global warming, the growing typhoon intensity and storm surgealong the Korean coasts will be a great potential danger in this area.

25


A Numerical Study of the Effects of Coastal Geometry on Storm Surges inthe Bohai SeaW.S. Jiang and P. ZhaoPhysical Oceanography Laboratory, Ocean University of China, Qingdao 266100, P. R. China

Strom surges are not only determined by the atmospheric forcing, but also influenced by the geological factorsof the affected areas, such as the coastal geometry and bathymetry. The coastline geometry changes evidently inmany areas of China, especially in the Bohai Sea, where the geometry has changed dramatically both at the tip ofthe Yellow River Delta and in the vicinity of Caofeidian harbor area due to natural and anthropogenic factors. Asthe area of Yellow River Delta grows at a mean net accretion rate of 10.02 km2/a and the planning build-up areaof Caofeidian industrial park will reach 310 km2 by 2030, the coastline of the Bohai Sea has changed dramaticallythese years. The storm surge in the changing Bohai Sea may have new characteristics due to the local changingof the geometry. To find out these new characteristics, the influence of the changes of coastal geometry on stormsurges is primarily investigated by numerical methods. The coastal geometry of the Bohai Sea in 1976, 2010 and2030 are involved as three coastal geometry scenarios in this study. A storm surge occurred in Oct. 2003 in theBohai Sea is modeled based on three coastal geometry scenarios of the Bohai Sea. By analyzing the model resultsof these scenarios, it is found that the southwest part of the Bohai Sea, especially the top of the Bohai Bay and theLaizhou Bay, suffers the most series disaster of storm surge, and the main changes in the distribution of maximumsurge elevation are located in the area south of the Yellow River esturay, where the maximum value of surgeelevation is strengthened. By comparing the risk and vulnerability assessment of storm surge in Tanggu, Huanghuaand Yangjiaogou, which are susceptible to storm surge in the Bohai Sea, the risk of storm surge in these places areall weakened by the development of the coastal geometry. Particularly in Yangjiaogou, though the maximum valuekeep almost the same in the three scenarios, the risk and vulnerability of storm surge is still weakened for the surgeelevation during the process of water level rising is decreased by the evolvement of the Yellow River Delta.

26


Analysis of Anomalous Storm Surge around West Coast of the Sea ofJapanS.Y. Kim (1), Y. Matsumi (2), T. Yasuda (3), and H. Mase (3)(1) Dept of Social Management, Tottori University, 4-101 Koyama-Minami, Tottori 680-8552, [email protected], +81-857-31-5311, (2) Dept of Social Management, Tottori University, 4-101Koyama-Minami, Tottori 680-8552, Japan. , (3) Disaster Prevention Research Inst., Kyoto University, Gokasho, Uji, Kyoto611-0011, Japan.

Introduction

Maximum surges occur after typhoons pass at Sakai Minato in west coasts of the Sea of Japan instead of thetyphoons’ peak at the site. The time lag is more evident from the time series of measured water levels during theTyphoon Songda in 2004 and predicted tides by Japan Meteorological Agency. It denotes that the maximum surgewas delayed 15 hours after the typhoon passed.

Understanding of such anomalous storm surges is poor in the west coast of the Sea of Japan. The anomalous stormsurges are characterized by the fact that typhoon centers are far from the coast. The present study examines themechanisms of anomalous surges from viewpoints of external meteorological forces and induced currents at SakaiMinato during Typhoon Songda (2004).

Numerical analysis of storm surge

A coupling model of surge, wave and tide (called SuWAT), developed by Kim et al. (2008)), was employed inthe present study by using six level grid system (nesting scheme). Two types of external wind and pressure es-timations were used: one is those from an empirical model by Fujii and Mitsuta (referring to FM); another isthose from Weather Research and Forecasting model (WRF) by Skamarock et al. (2008). FM typhoon model givessymmetric wind and pressure fields without considering the terrain effects; while WRF model gives more realis-tic meteorological fields influenced by topography than FM output. WRF was run by a two-way nesting with 12km - 1.3 km grid sizes for six grid regions with the 4D data assimilation method: 1) nudging with the reanalysisFNL data (NCEP/NCAR) in the outer most domain (referring to WRF-NF); and 2) nudging in all nested domains(WRF-NA).

In the storm surge simulations two different drag coefficients were used: one is the wave dependent drag (WDC);another is the empirical drag coefficients (EDC) to estimate the wind stresses.

Results and discussion

The maximum pressure depression of FM model is close to the measurement at Sandozaki. The time series byWRF-NF and WRF-NA outputs agree well to the measurement although their maximum pressure depressionsare a little smaller than the measurement. On the other hand, the wind speed estimated by the FM model areconsiderably larger than the measurement at Mihonoseki, while those predicted by WRF-NF and WRF-NA areclose to the measurement.

The surge heights in the second sea level rise using WRF-NF estimations agree well with the measured sea levelrise. The storm surge heights using WRF-NA estimations are under-predicted for the secondary sea level rise. Thestorm surge heights cannot present the measurement when using estimations by FM model.

From the storm surge simulations with and without Coriolis force, it is found that Coriolis force is an importantfactor for the anomalous storm surge at Sakai Minato as well as the accurate estimations of external forces of windsand pressures.

27

REFERENCES

Fujii, T. and Mitsuta, Y., 1986. Synthesis of a stochastic typhoon model and simulation of typhoon winds. Annualsof Disas. Prev. Res. Inst., Kyoto Univ., 29, B-1: 229-239.

Kim, S.Y., Yasuda, T., Mase, H., 2008. Effects of Tidal Variation on Storm Surges and Waves, Applied OceanResearch, 30, 311-322.

Skamarock, W.C., Klemp, J.B., Dudhia, J., Gill, D.O., Barker, D.M., Duda, M.G., Huang, X.Y., Wang, W., Powers,J.G., 2008. A description of the Advanced Research WRF Version 3, NCAR/TN-475+STR, 113p.

28


Extreme wind waves from the visual VOS observations worldwideV. Grigorieva (1) and S. Gulev (2)(1) P.P. Shirshov Institute of Oceanology RAS, Moscow, Russia, [email protected], (2) P.P. Shirshov Institute of OceanologyRAS, Moscow, Russia, [email protected]

We analyse extreme wind waves using VOS collection of visual wind wave observations assembled in ICAODSarchive and covering the period from 1880 onwards. In order to estimate characteristics of extreme wind waveswe used individual wave reports for the space grid cells from 2º×2º degrees to 10º ×10º degrees. Wind waveheight, swell height and the derived significant wave height were taken at the actual time of sampling. Then forevery calendar season we estimated characteristics of extreme waves using both initial value distribution and peakover threshold methods. In order to apply extreme value statistics to heavily and inhomogeneously undersampledVOS data, we used 6-hourly snapshots of wave characteristics from ERA-40-WAM hindcast covering the periodfrom 1960 to 2002. These model data were subsampled in order to simulate VOS sampling density. Analysis ofprobability distributions shows that VOS-like subsampled wind sea and SWH demonstrate regionally significantlydifferent modal values compared to those reported by WAM. The largest deviations of probability distributionswere found for the wind sea. Comparison of statistical characteristics of the fully sampled WAM data with theVOS-like undersampled subset allowed for estimation of the effect of sampling on probability distributions. Beingarmed with estimates of all types of biases in PDF, we estimated corrected extreme wind wave characteristics.Corrections in poorly sampled regions may be quite high and may even imply the change of sign of linear trendsin extreme waves. Then we estimate long-term tendencies and characteristics of interannual variability in extremewave parameters. During the last 50 years extreme waves in both North Atlantic and North Pacific experienceslight increase, which is, however, superimposed with the pronounced decadal-scale variability.

29


Understanding storm surge risk in Fiji due to climate variability andchangeK L McInnes (1), K Walsh (2), J O’Grady (1), and F Colberg (1)(1) Centre for Australian Weather and Climate Research, CSIRO, Marine and Atmospheric Research, Aspendale, Australia([email protected]), (2) School of Earth Sciences, University of Melbourne

Global warming induced sea level rise and changes in tropical cyclone behaviour have the potential to significantlyalter the risk of extreme sea level and coastal inundation events in the future. However, in the southwestern Pacific,the El Nino Southern Oscillation (ENSO) also causes significant variability in tropical cyclone behaviour. In ElNiño years, cyclone genesis is greater east of the dateline, with the highest density occurring around 10oS, 180oEwhile in La Niña years, the region of greatest formation is near 15oS 170oE, as well as to the south of Fiji. Thereare three preferred tracks of cyclones during the El Niño phase: one passing over the Fiji islands moving southeast,another recurving east of Vanuatu and a third near Samoa, tracking towards the southeast. During La Niña, thetypical track is from a point east of Vanuatu to a location south of Fiji (Chand and Walsh, 2009).

ENSO is also associated with variations in local mean sea level. During El Niño years, sea levels to thewest of about 165◦W and to about 10◦ either side of the equator are lower than average by up to several cmswhile in the eastern equatorial Pacific, sea levels are several cms higher than average (e.g. Church et al, 2006).Confidence in future projections of tropical cyclones and regional sea level therefore will depend to an extent theability of climate models to simulate ENSO behaviour now and into the future.

Tropical cyclone storm surge risk in the Pacific is of interest for impact assessment and adaptation plan-ning despite the high uncertainty surrounding its possible future changes. This study aims to explore howENSO and plausible scenarios of future climate change affect storm surge risk in Fiji using a statistical andhydrodynamical modelling approach similar to that described in McInnes et al (2003). Cyclone-induced stormsurge risk under average climate conditions is firstly quantified by developing a statistical model of cyclones forthe region of interest through an analysis of the historical cyclone records of the region. A simple cyclone vortexmodel is used to simulate the surface wind and pressure field associated with each cyclone and a hydrodynamicmodel is used to simulate the storm surges. With hundreds of plausible cyclones simulated, a map of storm surgehazard can be constructed. The sensitivity of this pattern of surge hazard to different modes of climate variabilityas well as climate change can then be investigated. Results of this study will be presented for Fiji.

Chand, S.S. and K.J.E. Walsh, 2009: Tropical cyclone activity in the Fiji region: spatial patterns and rela-tionship to large-scale circulation. Journal of Climate, 22, 3877–3893.

Church, J.A., White N.J., Hunter, J.R., 2006: Sea-level rise at tropical Pacific and Indian Ocean islands.Glob. Planetary Ch 53: 155–168.

McInnes, K.L., K.J.E. Walsh, G. D. Hubbert and T. Beer, 2003: Impact of Sea-level Rise and Storm Surges on aCoastal Community. Nat. Haz. 30(2) 187-207.

30


Storm Surges in the Canadian Beaufort SeaS.M. Solomon (1), G. Manson (1), D.L. Forbes (1), S.V. Kokelj (2), M. Pisaric (3), T. Lantz (4), E.J. Hart (5), andH. Melling (6)(1) Natural Resources Canada, Dartmouth, Canada ([email protected]), (2) Indian and Northern AffairsCanada, Yellowknife, Northwest Territories, Canada ([email protected], (3) Carleton University, Ottawa, Ontario, Canada([email protected]), (4) University of Victoria, Victoria, British Columbia, Canada ([email protected]), (5) Cochrane,Alberta, Canada ([email protected]), (6) Fisheries and Oceans Canada, Sidney, British Columbia,Canada([email protected])

Northwest winds acting over the extensive shallow shelf of the eastern Beaufort Sea can result in significantstorm surges which impact human activities and natural ecosystems. Open-water season surges contribute to rapiderosion of unlithified, frozen coastal bluffs with 10-15 m of erosion occurring during a single storm. In manylocations surge-elevated waves directly impact coastal bluffs causing deep thermal notching followed by collapseof large blocks directly onto the beach. Where elevations are lower, extensive flooding occurs causing problemsfor both permanent and seasonal (subsistence hunting and fishing) communities. Winter-season surges also occur,despite complete ice cover, causing flooding and breakup of ice roads and potential erosion of ice-constrainedchannel mouths. Ivu, or ice push, associated with winter surges can extend onto land with associated risk to humanlife and infrastructure. Fatalities due to surges are infrequent, but a severe event in 1970 led to two deaths at a campin the outer Mackenzie Delta. The same event also resulted in salt contamination of a community water supply.An unusually severe event in 1999 led to the salinization of 200 km2 of highly productive waterfowl habitat inthe Mackenzie Delta. Occasional occurrences of negative surges (0.4 m below chart datum) due to strong offshorewinds can impact commercial ship traffic transiting the wide, shallow inner shelf.

Storm surge chronologies are challenging to construct due to a lack of observations and the limited timespan covered by the instrumental records. An additional complexity is the variability of Arctic sea-ice cover. Inthe Canadian Beaufort Sea, official coastal meteorological and water level records extend back to the early 1960s,although the records are far from complete with very few reporting stations. We have attempted to supplementthese records using a variety of sources including unofficial records from military sites, commercial ventures(e.g. Hudson Bay Company records), missionary and church records and oral histories based on interviews withaboriginal elders. While these approaches yield interesting observations, they are not continuous and thereforedifficult to use for quantitative analysis. Based on the analysis of the tide gauge record at Tuktoyaktuk (1962-1997),a total water level (relative to chart datum) of 2.36 m represents the 25 year event. The highest recorded waterlevel during that period was in 1993 (2.2 m). Surveys of loglines suggest that the 1970 event reached 2.7 m, butswash driven water levels reach higher than 3.5 m. Relative sea level in the region is rising due to a combinationof subsidence and eustatic sea level rise. Sea ice duration is expected to decrease as the Arctic warms and eustaticsea level rise is predicted to accelerate. Therefore, total water levels associated with storm surges are expected toincrease. The potential trajectory for frequency and duration of storms is not well constrained by climate modelpredictions. Present variability is too large and the record too short to detect trends, but even if storminess remainsunchanged the exposure to damaging surges is likely to increase due to changing sea ice conditions.

31


Xynthia storm impact on french south britanny coast-suscinio bayA. Dubois, M Sedrati, and D Meniermarine geosciences and coastal geomorphology unit, Géoarchitecture EA2219, Université de Bretagne Sud, 56017 cedexVannes, France.

The storm called Xynthia took place first offshore of Portugal and then affected French Atlantic shore during thenight between the 27th and the 28th of February 2010. In the Vendée area, about 170km south, a few seawallsbroke up and massive flood occurred into built areas and farmlands. 53 casualties counted and material damagesevaluated about 1 billon euros cost. This winter storm head towards North-East and crossed France, Belgiumand the Netherlands. Important wind speeds have been recorded on the shore and inland due to the route of thestorm and its low atmospheric pressure (963 hpa, 18 m/s wind speed recorded in Belle-île island, offshore southBrittany). Moreover, this storm was combined with high spring tide (102). Even if the studied area was not locatedon the main route of the storm, it has been under impact of such event.An hydrodynamic and topographic survey was conducted from the 26th of February to the 5th of march onSuscinio beach (south Brittany, western France). Hydrodynamic data were collected with adcp and S4adw currentmeters. Topographic data were conducted with DGPS Trimble GeoXH2008 each daylight tide from the 27th ofFebruary until the 4th of March 2010 where sensors were deployed. Just after the storm event, 9 beach profileswere conducted in order to compare the topographic change with the same last topographic survey conducted onemonth earlier. The studied area is a pocket beach embayed between rocky headlands exposed globally N105. Theonshore of the beach is a steep reflective slope composed by heterogeneous coarse sand. The offshore is a flatdissipative low terrasse composed by fine and cohesive sand. This is a dune system protecting low swamps behindit. The deployment of three currentmeters in three different sites in the embayed beach highlights the behaviour oftidal currents under different windy conditions during high spring tides. The results of this survey during a winterstorm can thus be compared to modal conditions recorded one month earlier. This highlights the role of windforcing a few hours before stormy conditions and the impact on tidal currents. Xynthia storm is marked by a quickchange of wind speed and direction. Sensors recorded a signicative surge about 63cm high above the predicted hightide of Port-Navalo by S.H.O.M. It recorded unexpected moderate Hs and low tidal currents speed. Surge effect hasbeen absorbed because of The north-westward wind direction on the site. The hydrodynamic data recorded duringthe winter storm were less stronger compared to other datasets collected during this survey but the morphology ofthe beach was severely affected by a negative sedimentary balance and over washes deposits as far as the swampsbackward. In the western part, 1,8 m backward movement of the shoreline have been measured. However, themorphological change measured highlighted on this beach thanks to Digital Terrain Model is not homogeneousbut significant. This change is nevertheless representative of the storm impact on beach topography. Because ofsuch results, beach topographic changes seem to be mainly led by the Xynthia storm surge rather than wind forcing.

32


A 3D Model for the Kiel Fjord on the German Baltic Sea CoastN. Jimenez (1) and R. Mayerle (2)(1) FTZ-CORELAB, University of Kiel, Kiel, Germany ([email protected]), (2) FTZ-CORELAB, University ofKiel, Kiel, Germany ([email protected])

In this paper results of the set-up and simulations of the flow and waves in the Kiel Fjord on the German BalticSea are presented. A three-dimensional baroclinic model with 30m grid resolution in the horizontal and 4 and 8layers over the vertical was used. Boundary conditions along the open sea boundaries were obtained from a largescale model covering the entire Baltic Sea. Results of simulations covering moderate conditions and the extremestorm (Daisy) in January 2010 which caused main damage along the western promenade are presented. The abilityof the model in predicting water levels and waves in good agreement with observations resulted quite high. A cleareffect of the storms on the circulation patterns was found. The results obtained showed the capability of the modelin supporting the design of coastal protection measures and for assessing the effects of harbour expansions andthermal plants.

33


Methodology for evaluation of flood risk in Port areas. Application to thePort of Praia da VitóriaP.D. Raposeiro (1), C.J.E.M Fortes (2), M.T. Reis (3), and J.C. Ferreira (4)(1) Universidade Nova de Lisboa, Lisboa, Portugal ([email protected]), (2) Laboratório Nacional de Engenharia Civil,Lisboa, Portugal ([email protected]), (3) Laboratório Nacional de Engenharia Civil, Lisboa, Portugal ([email protected]), (4)Universidade Nova de Lisboa, Lisboa, Portugal ([email protected])

This paper illustrates the methodology developed at the National Civil Engineering Laboratory (LNEC), Portugalwith the partnership of the Sciences and Environmental Engineering Department of the New University of Lisbon(UNL)to evaluate the flood risk areas in the Port of Praia da Vitória (Terceira island, Archipelago of Azores),especially in storm conditions.Such a methodology is based on four key stages: a) division of the study area into sub-areas of similar charac-teristics; b) definition of the qualitative factors associated with the consequences, when pre-defined flood limitsare exceeded; c) determination of the probability of the flooding levels surpassing the pre-defined limits, for eachstudy area; d) assessing the flood risk of the port area.For the determination of the flooding levels the method described in Raposeiro et al. (2009) was used, based on thewave buoy data collected in storm conditions. These data are propagated to the Port of Praia da Vitória by usingwave propagation numerical models, included in the geographical information system (GIS), GUIOMAR, (Neveset al. 2009). Then, the determination of run-up and flooding levels is performed by using empirical formulae forport protection structures. Especial emphasis is placed on the automation of this process in the GUIOMAR systemto make it an efficient tool to produce flood levels and risk maps in the Port of Praia da Vitória.

NEVES, D.R.C.B.; ZÓZIMO, A.C.; PINHEIRO, L.V.; FORTES, C.J.E.M. (2009). GUIOMAR: Geo(graphical)User Interface for cOastal and MARine Modeling. Supported Decision System. Journal of Coastal Research, SI56 (Proc. 10th International Coastal Symposium), pp. 1542-1546.RAPOSEIRO, P.D.; FORTES, C.J.E.M.; REIS, M.T. (2009). Ferramenta de cálculo e análise do espraiamentoem estruturas de enrocamento: caso de estudo Praia de Vale do Lobo. Proc. 3º Encontro Nacional de Riscos,Segurança e Fiabilidade, Lisboa, 3 a 5 de Novembro.

34


ENSURF: Multi-model sea level forecast. Implementation and validationresults for the IBIROOS regionB. Pérez (1), E. Alvarez Fanjul (1), R. Brower (2), J. Beckers (2), D. Paradis (3), C. Balseiro (4), K. Lyons (5),and M. Cure (5)(1) Puertos del Estado, Area de Medio Físico, Madrid, Spain ([email protected]), (2) Deltares, Delft, Netherlands,[email protected], (3) Meteo-France, Toulousse, France ([email protected]), (4) MeteoGalicia, Santiago deCompostela, Spain ([email protected]), (5) Irish Marine Institute, Galway, Ireland ([email protected]

ABSTRACT

ENSURF: Multi-model sea level forecast. Implementation and validation results for the IBIROOS region

B. Pérez1, E. Alvarez1, R. Brower2, J. Beckers2, D. Paradis3, C. Balseiro4, K. Lyons, M. Cure5

1. Area de Medio Físico de Puertos del Estado. Avda. del Partenón, 10, 28042 Madrid, [email protected],[email protected]. Deltares, PO Box 177, 2600 MH Delft, Holanda.3. Meteo-France, Av. Coriolis 42, 31057 Toulouse Cedex, Francia.4. MeteoGalicia, Rúa Roma, nº6 15707 – Santiago de Compostela.5. Irish Marine Institute, Rinville, Oranmore, Co. Galway, Irlanda.

ENSURF is a multi-model application for sea level forecast that makes use of existing storm surge or cir-culation models today operational in Europe, as well as near-real time tide gauge data in the region, with thefollowing main goals:providing an easy access to existing forecasts, as well as to its performance and model validation, by means of anadequate visualization toolgeneration of better forecasts of sea level, including confidence intervals, by means of the Bayesian ModelAverage Techique (BMA)The system was developed and implemented within ECOOP (C.Nº 036355) European Project for the NOOS andthe IBIROOS regions, based on MATROOS visualization tool developed by Deltares. Both systems are todayoperational at Deltares and Puertos del Estado respectively. The Bayesian Modelling Average technique generatesan overall forecast probability density function (PDF) by making a weighted average of the individual forecastsPDF’s; the weights represent the probability that a model will give the correct forecast PDF and are determinedand updated operationally based on the performance of the models during a recent training period. This impliesthe technique needs the availability of sea level data from tide gauges in near-real time. Results of validation ofthe different models and BMA implementation for the main harbours will be presented for the IBIROOS region,where this kind of activity is performed for the first time. The work has proved to be useful to detect problems insome of the circulation models not previously well calibrated with sea level data, to identify the differences onbaroclinic and barotropic models for sea level applications and the general improvement of the BMA forecasts.

REFERENCes- Leamer, E.E., 1978. Specification Searches, Wiley- Alvarez Fanjul, E., Pérez, B., Sánchez-Arévalo, I.R., 2000. Nivmar: A storm surge forecasting system for theSpanish Waters. Scientia Marina, Vol 65, pp. 145-154- Poole, D. and Raftery, A.E. 2000. Inference for Deterministic Simulation Models: The Bayesian MeldingApproach. Journal of the American Statistical Association;ABI/INFORM Global pg. 1244

35

- Raftery, A.E., Gneiting, T., Balabdaoui, F., Polakowsly, M., 2005. Using Bayesian Model Averaging to CalibrateForecast Ensembles. Am. Meteorological Soc. 133, 1155,1174.- Beckers J.V.L., Sprokkereef, E. and Roscoe, K.L. 2008. Use of Bayesian model averaging to determineuncertainties in river discharge and water level forecasts.

36


Intense storms and associated impacts - Storm Xynthia caseM. L. R. Liberato (1,2), J. G. Pinto (3), I. F. Trigo (1,4), and R. M. Trigo (1)(1) CGUL, IDL, University of Lisbon, 1749-016 Lisbon, Portugal ([email protected]), (2) University of Trás-os-Montes e AltoDouro, School of Sciences and Technology, Physics Dept, 5001-801 Vila Real, Portugal, (3) Institute for Geophysics andMeteorology, University of Cologne, Cologne, Germany, (4) Institute of Meteorology 1749-077 Lisbon, Portugal

Intense extratropical cyclones are often associated with extreme weather conditions, in terms of wind and pre-cipitation, being among the most severe natural hazards affecting Europe. In winter 2009 and early spring 2010the Northern part of Iberia and Southern France were hit by destructive windstorms. In this work we assess thesynoptic evolution and the main impacts of storm Xynthia, which caused considerable economical losses on 28thFebruary 2010, mainly in France where 45 people were killed, mostly in the flooding in the Atlantic coast, near LaRochelle. Overall, storm Xynthia was responsible for 59 casualties in Europe - more than storms “Lothar” (1999)and “Kyrill” (2007). The analysis of storm Xynthia is also put into perspective among storminess variability onNorth Atlantic region. The objective of such exercise is to better understand surges driving factors and to assessassociated coastal risk on this region.

37


Modelling of storm surges in the German Baltic SeaG. Bruss, N. Jimenez, H. Eiben, and R. MayerleResearch and Technology Centre - University of Kiel, Germany ([email protected]; [email protected])

This paper summarises the results of numerical model simulations of the storm surges in the German Baltic Seain November 1872, February 2002 and January 2010. The investigations were carried out with simultaneouslycoupled flow and wave models downscaling from the Baltic Sea to high resolving nearshore models of the mostaffected coastal regions. The reconstruction of the hydrodynamic reactions to the three cyclones resulted in goodagreement with observations. A good spatial correlation was found between high energy levels of the computedcoastal impacts and reported damages. The analysis of the temporal and spatial variability of water levels and waveheights during the extreme events provide useful information for the design of coastal protection measures. Both,hazard hot spots common to all of the storms as well as individual patterns were identified along the coast. Resultsof the investigations on the effects of sea level rises are also presented.

38

Storm Surges Congress, Hamburg, Germany13�

»conference topics« · the shore of the arctic seas are generally of low relief ... between tidal...

Documents