Storm Surge

A storm surge is an abnormal rise in coastal water level generated by a storm, measured above the predicted astronomical tide. It is produced chiefly by wind pushing water toward the shore, with reduced atmospheric pressure also contributing. Storm surge and astronomical tide together constitute the storm tide; waves can raise the water’s reach still further. Both tropical cyclones and extratropical storms can generate damaging surges. (nhc.noaa.gov)

Storm surge is a major coastal hazard because elevated water can inundate low-lying land, penetrate inland through waterways, and allow waves to attack normally sheltered ground and structures. Its severity depends on the storm’s characteristics, coastal configuration, and timing relative to the tide. Understanding surge involves coastal physics, observation, numerical forecasting, and flood-risk assessment; protection combines warning systems, coastal engineering, and land-use measures. Sea-level rise increases the baseline from which storm-driven flooding occurs, linking this episodic hazard to long-term coastal change. (nhc.noaa.gov)

Definitions and Water-Level Terminology

Storm surge is the abnormal rise in water level generated by a storm above the predicted astronomical tide. It is a difference between water levels, not an elevation above a fixed reference surface. A surge of 2 metres means that the water stands 2 metres above the level expected from the astronomical tide at that place and time; it does not mean that the water surface is 2 metres above mean sea level or that nearby land is under 2 metres of water. The National Hurricane Center’s 2013 terminology statement explicitly distinguishes this tidal departure from elevations referenced to a datum. Wind and atmospheric pressure generate the storm-related departure, whereas astronomical tide predictions describe the expected tidal fluctuations. (nhc.noaa.gov)

Storm tide is the water level resulting from the combination of astronomical tide and storm surge. Unlike surge, its elevation requires a stated vertical datum: a reference surface from which heights or depths are measured. A tidal datum is a reference elevation defined using tidal observations, such as mean higher high water. With all elevations expressed relative to the same datum, the relationship is $s(t)=\eta_{\mathrm{ST}}(t)-\eta_{\mathrm{T}}(t)$, where $s$ is surge, $\eta_{\mathrm{ST}}$ is storm-tide elevation, and $\eta_{\mathrm{T}}$ is predicted astronomical-tide elevation. Changing the common reference datum changes the two elevations but not their difference. (nhc.noaa.gov)

Inundation depth measures water above normally dry local ground, rather than above the predicted tide or a datum. Where storm-tide water covers the ground, its depth is the storm-tide elevation minus the ground elevation, with both referenced consistently. In an illustrative calculation, a predicted tide of 0.8 metres and a surge of 2.0 metres produce a storm-tide elevation of 2.8 metres. At a flooded location whose ground elevation is 1.3 metres in the same datum, the depth is 1.5 metres. These are three different quantities—2.0 metres of surge, 2.8 metres of storm-tide elevation, and 1.5 metres of inundation—not interchangeable descriptions of the flood. (nhc.noaa.gov)

Surge and storm tide are also distinct from individual wind waves. Waves produce oscillating crests and troughs on the elevated water surface; their height is not the height of the surge. At the shore, breaking waves additionally produce wave setup, a rise in the time-averaged water level, and wave runup, the fluctuating reach of water up the beach. Runup incorporates the mean setup and the time-varying uprush and backwash known as swash. Its magnitude depends on wave height, wave period, and beach slope. A shoreline water-level estimate must therefore specify whether wave effects are included. (nhc.noaa.gov)

These distinctions explain why surge height alone cannot describe shoreline flooding. The US Geological Survey’s storm-impact framework compares water levels incorporating tide, surge, and wave runup with local dune elevations. Waves may briefly overtop a dune even when the underlying storm tide remains below its crest; sustained submergence is a different condition. Likewise, identical storm-tide elevations can correspond to different inundation depths where ground elevations differ. (usgs.gov)

Physical Generation

Storm surge is generated chiefly by the transfer of momentum from storm winds to the sea through surface stress. Onshore winds drive water toward the coast; where the shoreline and shallow seabed restrict its redistribution, water accumulates and the sea surface develops a landward-rising slope. The resulting pressure gradient acts against further accumulation. Wind stress increases approximately with the square of wind speed when the air–sea drag coefficient is held constant, so stronger winds can greatly increase the forcing. That coefficient is not constant across all conditions, however: observations show that its increase levels off at high wind speeds. Wind speed therefore does not translate into surge height through a simple universal square-law relationship. (co-ops.nos.noaa.gov)

The direct effect of atmospheric pressure is described by the inverse-barometer effect. In an ideal equilibrium, a pressure decrease allows the water surface to rise according to

where $\Delta \eta$ is the water-level change, $\Delta p$ the atmospheric-pressure change, $\rho_w$ seawater density, and $g$ gravitational acceleration. Over the open ocean, a decrease of 1 hectopascal—equivalent to 1 millibar—corresponds approximately to a 1-centimetre rise. A 50-hectopascal pressure deficit would thus imply roughly 0.5 metres of equilibrium elevation, not several metres. This is an equilibrium approximation rather than a complete coastal prediction: storm-driven currents and changing winds also determine the water level. NOAA’s analysis of Hurricane Floyd in 1999 estimated maximum pressure contributions of about 10 centimetres at Virginia Key, Florida, and 50 centimetres at Duck, North Carolina, while identifying wind stress as often the substantially larger influence during storms. (co-ops.nos.noaa.gov)

Storm intensity and wind-field size control different aspects of that forcing. Stronger winds generally exert greater stress, whereas a larger wind field applies stress over a greater ocean area and can maintain strong winds at a location for longer. Consequently, two storms with the same maximum sustained wind can move different volumes of water toward land. The distribution of winds around the centre matters alongside the radius of maximum winds; the maximum speed alone does not describe the full area supplying shoreward transport. Coastal depth and geometry further modify the response to this forcing. (nhc.noaa.gov)

Forward speed, track, and approach direction determine where the winds act and how long they persist. A track shift can relocate the strongest onshore winds, while a more nearly perpendicular coastal approach can concentrate shoreward forcing differently from a coast-parallel passage. Forward speed has no universally applicable effect: National Hurricane Center training examples show faster storms producing higher immediate open-coast surges, but slower storms sustaining forcing long enough to produce greater rises in enclosed waters or carry flooding farther inland. These controls interact rather than operating as independent multipliers. (nhc.noaa.gov)

The Saffir–Simpson Hurricane Wind Scale therefore cannot specify surge height, because it classifies hurricanes only by maximum sustained wind. Hurricane Ike reached Texas in 2008 as a Category 2 hurricane but produced a peak surge of approximately 20 feet (6.1 metres); its hurricane-force winds extended as far as 125 miles (201 kilometres) from the centre. Hurricane Charley struck Florida in 2004 as a Category 4 hurricane, yet its much smaller wind field—hurricane-force winds extended at most 25 miles (40 kilometres)—produced a peak surge of about 7 feet (2.1 metres). These contrasting events illustrate why surge estimates require the storm’s spatial wind structure and motion as well as its intensity. (nhc.noaa.gov)

Coastal Controls and Compound Flooding

A broad, shallow continental shelf generally supports greater wind-driven water accumulation than a narrow, steep shelf under comparable storm forcing. Shallow water requires a larger water-surface slope to balance the applied wind stress, allowing substantial coastal setup across a wide shelf. The National Hurricane Center identifies the gently sloping shelves off Louisiana and Mississippi as particularly favorable for large surges. Coastline shape further modifies this response: an inward-curving coast, such as Florida’s Apalachee Bay, can concentrate storm-driven water, whereas an outward-projecting coast allows different patterns of redistribution. These controls mean that similar offshore storms need not produce similar coastal water levels. (nhc.noaa.gov)

Bays and estuaries can carry elevated water far inland, but narrowing does not invariably amplify a surge. Channel convergence, depth, bottom friction, and the duration of elevated coastal water jointly determine whether the disturbance grows or is damped as it travels upstream. In a 2022 idealized modeling study, Ramin Familkhalili, Stefan Talke, and David Jay found that deeper estuarine channels generally reduced surge damping, while greater river discharge increased it. Channel deepening also moved upstream the transition between reaches dominated by coastal water levels and those dominated by river flow. Consequently, altered channel geometry can redistribute flooding along an estuary rather than change water levels uniformly throughout it. (os.copernicus.org)

The timing of a surge relative to the astronomical tide is especially important around the United Kingdom, where tidal ranges are large. The maximum surge—the storm-generated departure from the predicted tide—need not occur at the maximum total water level. A large departure near low tide can produce a lower flood level than a smaller departure near high tide; therefore, the separate maxima of tide and surge cannot simply be added unless they coincide. The Met Office notes that a timing difference of only a couple of hours can determine whether coastal flooding occurs. On 5 December 2013, high tidal conditions combined with a large surge to produce widespread flooding along the UK east coast. (weather.metoffice.gov.uk)

Compound flooding occurs when coastal water levels, heavy rainfall, and river flows combine, simultaneously or in succession. Elevated sea levels can impede the drainage of surface runoff and river water, creating backwater effects even where seawater does not directly inundate the land. Rainfall and discharge also have different travel times through a drainage basin, so their peaks may arrive after the surge peak. These are interacting hydraulic processes, not merely independently estimated flood maxima added together: water depth, friction, storage, and flow gradients change during the event. A 2022 study by Nahal Maymandi, Michelle Hummel, and Yu Zhang modeled coastal, riverine, and rainfall flooding in Texas’s Sabine–Neches Estuary. Among the hurricanes examined, extended rainfall associated with Harvey produced the greatest extent and longest duration of compound flooding, with evolving patterns and multiple flood peaks. (doi.org)

Floodwater can also cross a barrier island from its landward side and drain toward the ocean. During Hurricane Dorian in 2019, inundation of North Core Banks—a 36-kilometre-long island in North Carolina’s Outer Banks—was driven primarily from Pamlico Sound rather than the ocean shore. Christopher Sherwood and colleagues reported in 2023 that sound-side flooding and seaward drainage formed more than 80 washout channels, extending across the island through gaps in the foredunes. The island lost approximately 18% of its above-water sediment volume, with sediment exported toward offshore bars. This case demonstrates why ocean-facing shoreline conditions alone cannot establish the direction or severity of barrier-island flooding. (usgs.gov)

Storm Types and Geographic Distribution

Both tropical cyclones and extratropical cyclones generate extreme coastal water levels through strong surface winds and low atmospheric pressure. Their meteorological structures differ: tropical cyclones are warm-core systems whose energy supply depends on the warm ocean, whereas extratropical cyclones develop along temperature contrasts and are associated with fronts. Tropical cyclones dominate major surge hazards in several low-latitude regions, including the Bay of Bengal and the southern United States; extratropical storms are important surge generators at higher latitudes, notably around the North Sea. These geographical patterns overlap rather than forming an absolute boundary between two kinds of coastal hazard. (archive.ipcc.ch)

The Bay of Bengal is a major tropical-cyclone surge setting, exposing the low-lying coast of Bangladesh and the eastern coast of India. Its extensive shallow northern coastal waters favour large water-level responses to storm winds, while islands, tidal flats, and coastal plains provide pathways for widespread inundation. Cyclone Bhola struck what was then East Pakistan on 12–13 November 1970, overwhelming low-lying islands and tidal flats; the disaster illustrates the northern bay’s exceptional exposure without making its consequences representative of every cyclone. The hazard also crosses national borders within individual events. During Cyclone Amphan in May 2020, the India Meteorological Department forecast surge heights above astronomical tide of 4–5 metres for parts of West Bengal and 2.5–3.5 metres for Bangladesh; these were forecast values, not a uniform observed water level around the bay. (mausamaudit.imd.gov.in)

Along the United States Atlantic and Gulf coasts, hurricane-surge exposure extends well beyond the immediate landfall point. The National Hurricane Center identifies all locations along these coasts as vulnerable, including inland areas reached through rivers and canals. Louisiana and Mississippi are particularly susceptible because the seabed deepens gradually offshore, allowing substantial wind-driven water accumulation over a broad shallow shelf. Exposure is therefore not restricted to the places experiencing the strongest winds: storm size, track, approach angle, and local coastal geometry all influence where the largest surge develops. (nhc.noaa.gov)

In the western North Pacific, typhoons produce the corresponding hazard in the Philippines. Low-lying coastal communities are exposed where cyclone winds drive water ashore, with local inundation depending on storm strength, surge height, and ground elevation. Typhoon Haiyan, locally named Yolanda, reached the Philippines on 8 November 2013 and generated destructive coastal flooding around Tacloban. Its peak surge there occurred during relatively low tide; NOAA’s account estimated that coincidence with high tide would have raised the water level by another 0.5 metres. The event thus illustrates both the severity of Philippine typhoon surges and the importance of distinguishing the meteorological surge from the resulting total water level. (pagasa.dost.gov.ph)

The North Sea is a contrasting extratropical-surge setting in which one storm can affect several national coastlines sequentially. On 31 January 1953, persistent northerly storm winds drove water southward; the surge intensified along eastern Scotland and England before travelling northeastward along the Netherlands coast, reaching IJmuiden around 04:00 on 1 February. This was a basin-wide event rather than separate local storms. Another major event on 5–6 December 2013 combined northerly gales with spring tides, producing a surge generally around 2 metres above astronomical tide along the British North Sea coast. (weather.metoffice.gov.uk)

Observation and Event Reconstruction

Permanent coastal water-level stations provide continuous records against which storm flooding can be reconstructed, while temporary instruments fill gaps along the expected storm track. The U.S. Geological Survey (USGS) supplements fixed gauges with rapidly deployed radar gauges that transmit observations and submerged pressure sensors that generally store records for retrieval after the storm. Temporary sensors can be distributed across barrier islands, wetlands, coastal waterways, and low-lying inland areas to resolve how flooding spreads and recedes. In the monitoring procedures described by USGS in 2018, most temporary sites recorded water-level and atmospheric pressure at 30-second intervals; selected beach sensors sampled at two-second intervals to capture wave fluctuations. These time series reveal arrival times, rates of rise, peak levels, and flood duration rather than only a maximum elevation. (usgs.gov)

A non-vented submerged sensor records absolute pressure, which includes both atmospheric pressure and the pressure exerted by the overlying water. Under the hydrostatic approximation, water depth above the sensor is calculated as $h=(p_{\mathrm{sensor}}-p_{\mathrm{air}})/(\rho g)$, where $\rho$ is water density and $g$ is gravitational acceleration. Separate barometers, installed above anticipated flood levels, provide the atmospheric correction; salinity-dependent density corrections also affect the conversion from pressure to depth. The sensor’s surveyed elevation is then added to obtain water-surface elevation. During the USGS investigation of Hurricane Wilma’s October 2005 flooding in southwestern Florida, pressure records were corrected for atmospheric pressure and salinity, and optical-level surveys referenced both instrument records and high-water marks to the North American Vertical Datum of 1988. (pubs.usgs.gov)

Post-storm surveys extend the spatial coverage through high-water marks: sediment stains, fine seed lines, and debris left on buildings, vegetation, or other surfaces. Surveyors document the evidence and assign measurement uncertainty rather than treating every mark as an equally precise flood peak. Todd A. Koenig and colleagues’ 2016 USGS manual formalized procedures for identifying, preserving, and assessing such evidence. Porous surfaces can wick water above the actual waterline, while debris may sag or be displaced; sheltered, corroborating marks therefore carry particular value. A maximum mark generally supplies neither the time nor duration of flooding, and cannot by itself establish the sequence in which water arrived through different pathways. Wave action can further complicate interpretation of exposed marks. (pubs.usgs.gov)

Combining observations across a storm-affected coastline requires consistent vertical referencing. Global Positioning System surveys connect field sites to geodetic control, while leveling transfers elevations between benchmarks, local reference marks, and sensor mounting points. Water depth above one instrument cannot be compared directly with depth above another instrument mounted at a different elevation. USGS surveying guidance therefore emphasizes a common datum, consistent control coordinates, and checks on benchmark stability: storm damage, ground movement, or displaced monuments can invalidate an apparently established reference. Local reference marks also preserve the connection needed to re-survey a sensor site after its instrument has been removed. (water.usgs.gov)

Aerial photography and laser-altimetry surveys provide complementary remote-sensing evidence of the landscape response. Photographs document features such as overwash deposits and breaches, while pre- and post-storm elevation measurements quantify the magnitude and spatial variability of beach and dune change. Following Hurricane Irene’s North Carolina landfall on 27 August 2011, USGS collected aerial photographs and laser-altimetry observations to document coastal changes. Comparisons with pre-storm surveys identified erosion, sediment deposition, and island breaching, and supported refinement of coastal-impact predictions that compared modeled storm water levels and wave runup with measured dune elevations. These surveys document morphological outcomes rather than a continuous history of water levels. (usgs.gov)

Numerical Modeling and Forecast Products

The Sea, Lake, and Overland Surges from Hurricanes (SLOSH) model, used by the United States National Hurricane Center (NHC), calculates storm-driven coastal water levels with a two-dimensional hydrodynamic simulation. It solves simplified equations of momentum using the finite-difference method. Its applications differ principally in the meteorological information supplied: historical reconstructions use information about past hurricanes; hypothetical simulations explore prescribed storms for planning; and event forecasts use predictions for an approaching cyclone. These are distinct uses of the same modeling framework, rather than interchangeable statements about future flooding. (slosh.nws.noaa.gov)

Storm inputs include central atmospheric pressure, track, forward speed, and size, commonly characterized by the radius of maximum winds. The coastal domain supplies underwater depths and land elevations. SLOSH uses curvilinear grids with cells ranging from tens or hundreds of metres to a kilometre or more; channels, levees, roads, and other features smaller than individual cells can be represented through parameterizations. Consequently, both the storm specification and the representation of pathways and obstructions affect modeled inundation. Astronomical tides can be represented by a specified initial level or by spatially and temporally varying tidal calculations. A simulation initialized at high tide therefore represents a different water-level assumption from one initialized at mean tide. (github.com)

Ensemble forecasting addresses the sensitivity of surge to uncertain storm evolution instead of treating one forecast track as certain. NOAA’s Probabilistic Storm Surge system, P-Surge, constructs roughly 500–1,000 representative wind-and-pressure inputs using the official NHC forecast and historical forecast errors in track, size, and intensity. The resulting simulations support location-specific probability distributions of water levels. Two complementary outputs answer different questions: an exceedance-probability product gives the chance of exceeding a specified height, whereas an exceedance-height product gives the height associated with a specified probability. These probabilities describe uncertainty for the particular forecast event, not the annual frequency of coastal flooding. (vlab.noaa.gov)

The NHC’s Potential Storm Surge Flooding Map, introduced experimentally in 2014 and made operational in 2016, uses the 10-percent-exceedance height: at each location, the guidance assigns approximately a one-in-ten chance of water exceeding the mapped level. This is a location-specific, reasonable upper-bound scenario, not the most likely outcome or an absolute maximum. Processing in a geographic information system combines the water-level guidance with high-resolution land elevations to depict potential water depth above normally dry ground. The product incorporates astronomical tides and forecast uncertainty; it is therefore not simply a map of surge height above the predicted tide. (nhc.noaa.gov)

Composite scenario maps have a different construction. A Maximum Envelope of Water (MEOW) retains the greatest water level in each grid cell across hypothetical storms sharing a category, forward speed, approach direction, and initial tide level, but following different parallel tracks. In NHC’s illustrative example, five simulations produce cell values of 4.1, 7.1, 5.3, 6.3, and 3.8 feet; the MEOW retains 7.1 feet. A Maximum of MEOWs (MOM) takes a further cell-by-cell maximum across MEOWs for a hurricane category. Adjacent maxima may originate from different storms and need not occur simultaneously: neither composite represents the regional flooding from one hurricane. These planning envelopes also do not, by themselves, assign annual exceedance probabilities. (nhc.noaa.gov)

Coastal Erosion and Landscape Change

Elevated coastal water levels shift wave attack upward and landward across the beach profile. Waves that would otherwise break or dissipate below a dune can reach its base, remove sand, and steepen the seaward face. Continued undercutting causes sections of the dune to collapse and the dune edge to retreat. Buildings and infrastructure on or near dunes can consequently lose supporting sediment; erosion also exposes previously sheltered inland structures to waves and flooding. Thus, damage involves both water entering developed areas and physical removal of the ground beneath them. These mechanisms underpin the 2013 USGS Mid-Atlantic coastal-erosion assessment by Kara Doran and colleagues. (pubs.usgs.gov)

The USGS storm-impact scale, proposed by coastal scientist Asbury H. Sallenger Jr. in 2000, classifies these changes by comparing storm water levels with local beach and dune elevations, rather than by storm intensity alone. Let $D_{\mathrm{low}}$ denote the dune-toe elevation, $D_{\mathrm{high}}$ the dune crest, and $R_{\mathrm{high}}$ the upper reach of wave runup, including the underlying elevated water level. In the swash regime, $R_{\mathrm{high}}<D_{\mathrm{low}}$: wave action remains on the beach, redistributing sand without directly attacking the dune. In the collision regime, $D_{\mathrm{low}}<R_{\mathrm{high}}<D_{\mathrm{high}}$: waves strike the dune face and cause net dune erosion. The thresholds therefore couple hydrodynamic forcing to coastal geometry: the same water level can produce different impacts on adjacent stretches of coast. (usgs.gov)

In the overwash regime, $R_{\mathrm{high}}>D_{\mathrm{high}}$, allowing individual wave uprushes to cross the dune or beach berm and carry sediment inland. In the inundation regime, the sustained water level associated with surge and tide itself exceeds the crest, $R_{\mathrm{low}}>D_{\mathrm{high}}$, permitting continuous submergence rather than intermittent overtopping. Overwash leaves sheets or fan-shaped deposits behind the beach; inundation can transport sediment farther across a barrier island. These processes redistribute sand rather than simply remove it from the coastal system. Repeated landward sediment transfer contributes to barrier migration, even while the ocean-facing shoreline retreats. (usgs.gov)

Storm-driven landscape change can also involve breaches through the island, connecting the ocean with a back-barrier bay or estuary. During Hurricane Sandy in 2012, two breaches formed on Island Beach near Mantoloking, New Jersey, accompanied by sediment deposition in the estuary. Such openings alter barrier–estuary connectivity and sediment exchange, while storm currents reshape both the island and adjacent seabed. At Fire Island, New York, USGS measurements found that Sandy removed or redistributed 54.4 percent of the pre-storm beach-and-dune volume, with overwash affecting 46.6 percent of the island’s length. These measurements describe changes in sediment storage and distribution, not equivalent permanent losses of sand to deep water. (usgs.gov)

Recovery proceeds on different timescales for different parts of the profile. Sand displaced offshore during swash-dominated erosion can return to the beach over weeks to months; USGS observations after Hurricane Dennis in 1999 documented this recovery while the dune remained untouched. Dunes rebuild more slowly, gaining elevation and volume through deposits of wind-blown sand over years. A recovering shoreline therefore does not necessarily indicate a restored dune barrier. Following Sandy, Fire Island experienced further erosion during the 2012–13 winter storm season; by April 2013, beach volumes at 90 percent of surveyed profiles were similar to those measured immediately after Sandy—not to pre-storm conditions. (usgs.gov)

Historical Disasters and Institutional Responses

The North Sea flood of 31 January–1 February 1953 exposed failures in both coastal defences and warning transmission. Along eastern England, dunes were breached, embankments overtopped, and concrete seawalls collapsed; approximately 160,000 hectares were flooded and 307 people died. Around 1,800 people drowned in the Netherlands. The Met Office and Dutch Surge Warning Service had forecast dangerously high water several hours beforehand, but damaged telephone lines prevented effective transmission of warnings farther south along the English coast. The subsequent British inquiry recommended a dedicated flood-warning organization, leading to the Storm Tide Warning Service. Coastal defences were progressively strengthened, while protection of London became a major engineering priority. The Thames Barrier became operational in 1982 and was officially opened on 8 May 1984. Its development followed investigations prompted by the disaster, including Hermann Bondi’s assessment of barrier protection near Woolwich. (weather.metoffice.gov.uk)

The Bhola cyclone struck on 12–13 November 1970 in what was then East Pakistan, now Bangladesh. Its surge overwhelmed low-lying islands and tidal flats bordering the Bay of Bengal. The World Meteorological Organization (WMO) gives an estimated 300,000–500,000 fatalities for the cyclone disaster, mostly resulting from the surge; this range is not a separately established count of surge deaths. The catastrophe stimulated United Nations resolutions calling for coordinated measures against tropical-cyclone losses. WMO’s Sixth Congress initiated the Tropical Cyclone Project in 1971, and the WMO/ESCAP Panel on Tropical Cyclones was established in 1972. Following a decision in 1979, the project became the broader Tropical Cyclone Programme in 1980, institutionalizing international cooperation in monitoring, forecasting, warning, research, and training. (wmo.int)

Hurricane Sandy in October 2012 illustrated a different institutional response: coordinated documentation of coastal flooding across a large region. The United States Geological Survey (USGS) deployed a temporary network of water-level and barometric-pressure sensors at 224 Atlantic-coast locations from Virginia to Maine. Continuous records documented the timing and magnitude of storm tide, rather than only the maximum elevations left by flooding. Post-flood high-water-mark surveys supplemented the instruments, and USGS distributed monitoring information through its map-based Flood Event Viewer. The combination provided a regional observational record linking local flood measurements with the storm’s progression, supporting assessment of affected areas and subsequent scientific reconstruction. (usgs.gov)

Typhoon Haiyan, which crossed the Philippines on 8 November 2013, demonstrated the difficulty of translating meteorological warnings into locally understood inundation hazards. A WMO–UN-ESCAP–Typhoon Committee expert mission in April 2014 found that forecasts of the storm’s movement and intensity had been provided days in advance, yet unfamiliarity with the term “storm surge” contributed to underestimation of the danger. The mission identified shortcomings in explaining flood extent, water height, and urgency, and recommended relating forecast heights to recognizable buildings or landmarks and showing potentially affected areas graphically. These recommendations concerned communication of physical consequences, not merely replacement of one technical term with another. (old.wmo.int)

The post-Haiyan mission also recommended training forecasters to write in plain language, strengthening coordination among meteorological services, local authorities, and broadcasters, and improving warning dissemination at neighbourhood level. Subsequent Philippine practice incorporated probabilistic surge forecasts paired with impact matrices. WMO reported in June 2026 that these products help local governments interpret which areas may flood and when, directly addressing the gap between storm prediction and community-specific warnings exposed by Haiyan. (old.wmo.int)

Hazard Assessment, Exposure, and Loss

Flood hazard, exposure, and vulnerability describe different components of storm-surge risk. Hazard concerns the physical flooding; exposure identifies people, buildings, infrastructure, and other assets within the affected area; vulnerability describes their susceptibility to harm and capacity to cope. Consequently, similar coastal water levels need not produce similar losses: an unoccupied shore and a densely developed waterfront have different exposure, while buildings with different floor elevations and construction characteristics have different vulnerability. These distinctions follow the risk framework used by the Intergovernmental Panel on Climate Change (IPCC). (ipcc.ch) Building-level loss models make this distinction operational through depth–damage functions, which relate water depth relative to a building’s floor to the expected fraction of damage. FEMA’s Flood Assessment Structure Tool selects these functions using occupancy, first-floor elevation, foundation type, and number of storeys rather than water-surface elevation alone. (github.com)

A return level connects a specified water level to its annual exceedance probability. Under a stationary model, if $M$ is the annual maximum water level, the $T$-year return level $z_T$ satisfies $P(M>z_T)=1/T$. Thus, a 100-year level has a 1% annual exceedance probability, not a schedule of occurrence once per century. Assuming independent years and an unchanged probability, the chance of at least one exceedance over 30 years is $1-(1-0.01)^{30}$, approximately 26%. (usgs.gov) This interpretation depends on the flood-generating probability distribution remaining unchanged. In nonstationary extreme-value estimation, distribution parameters can vary with time or environmental covariates. Eleanor D’Arcy and Jonathan A. Tawn’s 2022 study illustrated this approach using four United Kingdom tide gauges: a nonstationary generalized Pareto model represented extreme skew surges, incorporating climate-related covariates, seasonality, and surge–tide interaction. Such methods distinguish changes in extreme-event behaviour from changes attributable solely to mean sea level. (arxiv.org)

Scenario maps answer a different question: which land could be inundated under specified storm conditions? The United States National Hurricane Center’s national maps use SLOSH simulations covering hurricane Categories 1–5, with variations in track, forward speed, storm size, and other inputs; the published mapping uses a high-tide initial water level. Maximum-envelope products combine local maxima from many hypothetical storms, so the mapped maximum at one location need not occur during the same storm as the maximum elsewhere. These composite category maps therefore describe potential inundation, not the footprint of a single event, and do not themselves assign annual probabilities. A Category 3 inundation area cannot be interpreted as a “100-year floodplain” without a separate analysis of storm occurrence and water-level probabilities. (nhc.noaa.gov)

Direct damage includes losses to inundated structures and their contents, but structural consequences depend on how flooding acts on the building. FEMA’s post–Hurricane Ian guidance distinguishes relatively slow inundation without waves from flooding involving stronger dynamic forces: wet contents do not necessarily imply extensive structural failure, whereas unequal water pressures can damage enclosing walls. Damage may increase abruptly once water exceeds the lowest floor. Loss assessments therefore require building characteristics and flood conditions, not simply a count of properties inside an inundation boundary. (fema.gov)

Wider economic consequences arise when damaged facilities interrupt services and supply networks beyond the flooded area. Dependencies among the electrical grid, transport, communications, and water systems create pathways for cascading disruption; firms outside the inundation zone may lose access to essential services, inputs, or distribution routes. The IPCC’s 2022 assessment concluded with high confidence that infrastructure connectivity can propagate severe-storm impacts beyond the hazard footprint, producing widespread regional disruption. Physical repair costs and destroyed inventories are therefore distinct from business interruption and losses transmitted through supply chains; the latter require analysis of network dependencies and service restoration rather than inundation depth alone. (ipcc.ch)

Warning Systems and Coastal Protection

Operational forecasts estimate the timing and magnitude of coastal water levels; inundation maps translate these estimates into potentially flooded land and water depth above ground. Watches and warnings communicate the threat, while emergency-management messages connect it to local evacuation and response arrangements. In the United States, the National Hurricane Center defines a storm-surge watch as possible life-threatening inundation, generally within 48 hours, and a warning as danger of such inundation, generally within 36 hours. These are jurisdiction-specific definitions, not universal thresholds. Its Potential Storm Surge Flooding Map incorporates forecast uncertainty and depicts depths with a 10% chance of being exceeded at each location—a reasonable worst-case depiction rather than a single deterministic prediction. The warning graphic identifies threatened areas but does not replace instructions from local emergency-management authorities. (nhc.noaa.gov)

Engineered protection interrupts different flood pathways. Seawalls, levees, and floodwalls form fixed barriers along shorelines or around protected land and assets; their design depends on the total water levels they must withstand, not surge height alone. Movable surge barriers, by contrast, close waterways through which elevated coastal water could enter an estuary, river, or inlet. They therefore exploit narrow openings rather than attempting to wall off an entire open coastline. Both approaches have finite design limits, and coastal protection planning also considers effects on adjacent properties, beaches, and natural resources. NOAA distinguishes these barriers from building-level flood-proofing and devices that prevent tidal backflow through drainage pipes. (coast.noaa.gov)

The Thames Barrier illustrates waterway closure combined with forecasting. Its ten steel gates span 520 metres across the River Thames near Woolwich. Normally, the gates permit river traffic and tidal exchange; during a closure, they form a barrier against incoming tidal water. The Environment Agency bases closure decisions on a combination of predicted tides, surge, and river flow, rather than on a single offshore surge threshold. The structure consequently functions as an operated component of a wider flood-defence system, not as a permanently closed dam. (gov.uk)

Natural and nature-based protection modifies waves, currents, and shoreline stability without necessarily excluding high water. Dunes provide elevated sandy barriers, while wetlands act as permeable buffers that slow waves and surge through friction. Vegetated shorelines also stabilize sediment: roots hold soil, and stems interrupt water movement. A living shoreline may combine plantings with low stone sills or other structural elements. These functions must be distinguished from complete inundation prevention: NOAA’s shoreline-stabilization guidance explicitly identifies some vegetated and low-sill treatments as providing no high-water protection, despite reducing wave energy and slowing inland water transfer. Suitability depends on local wave exposure, tidal range, shoreline slope, and available space; natural and structural measures can also be combined. (coast.noaa.gov)

The North Sea floods of 1953 and 2013 demonstrate how warning and protection can change outcomes during severe coastal events. On 31 January–1 February 1953, flooding killed 307 people in eastern England and inundated approximately 24,000 properties. The 5–6 December 2013 event produced the highest coastal water levels since 1953, with records exceeded at some locations, yet around 1,400 properties flooded on England’s east coast and no flood-related deaths were reported. The Met Office’s 2015 review records that coupled weather and surge models identified the 2013 risk about seven days ahead; increasing confidence prompted alerts to government and emergency responders and a red flood-risk assessment. The surge persisted over three high tides. The review attributes the reduced consequences to the combination of advance warning, emergency preparedness, and improved sea defences, rather than forecasting alone. (weather.metoffice.gov.uk)

Sea-Level Rise and Future Research

Sea-level rise raises the baseline from which storm-driven flooding develops; it does not necessarily increase the meteorological surge itself. If tides and surge characteristics remain unchanged, a higher mean sea level shifts the distribution of coastal water levels upward, making a previously rare elevation easier to exceed. In its 2021 assessment, the Intergovernmental Panel on Climate Change (IPCC) projected that, under unchanged tides, surges, and wave climate, extreme sea levels formerly reached once per century would occur annually or more frequently at approximately 19–31% of tide gauges by 2050. By 2100, the corresponding proportions were about 60% under the low-emissions scenario SSP1-2.6 and 82% under the very-high-emissions scenario SSP5-8.5, with medium confidence. These are changes in exceedance frequency, not equivalent increases in surge amplitude. (ipcc.ch)

Changes in storms can modify this baseline effect. The IPCC assessed that average peak tropical-cyclone wind speeds and the proportion of Category 4–5 cyclones will very likely increase globally with warming, whereas the global frequency of cyclone formation is projected to decrease or remain unchanged, with medium confidence. Tropical-cyclone rainfall rates are projected to increase globally with high confidence. Regional qualifications remain important: the projected poleward migration of tropical-cyclone activity in the western North Pacific has medium confidence, while extratropical-cyclone wind changes can have opposite signs in different regions as storm tracks shift. These findings do not imply a uniform increase in surge along every coastline: regional storm behaviour must be considered alongside the rising water-level baseline. (ipcc.ch)

Future coastal flooding also depends on interacting marine and terrestrial extremes. The IPCC’s 2021 assessment assigned high confidence to increasingly probable flooding in coastal cities from the combination of more frequent extreme sea levels and extreme rainfall or riverflow events. Exposure is not fixed: urban growth can place additional people and infrastructure in affected areas, while adaptation and protection standards alter resulting risk. The IPCC’s 2022 assessment estimated that, holding population and protection heights constant, the population at risk from a 100-year coastal flood increases by approximately 20% with 0.15 m of global mean sea-level rise, doubles with 0.75 m, and triples with 1.4 m, with high confidence. These conditional estimates separate the sea-level contribution from future demographic and protection changes. (ipcc.ch)

A statistical research priority is to distinguish mean-level trends from changes in the extremes themselves. In a 2022 study, Eleanor D’Arcy and Jonathan A. Tawn developed a nonstationary generalized Pareto model for extreme skew surge—the difference between observed and predicted tidal-cycle maxima—using records from four United Kingdom tide gauges. Their method incorporates climate-change covariates, seasonality, and dependence between skew surge and peak tide, then combines the components through a joint-probability framework. Such approaches allow extreme-water-level estimates to change with environmental conditions rather than assuming that removing a mean sea-level trend leaves an otherwise unchanging distribution. (arxiv.org)

Probabilistic forecasting is already operational, whereas faster emulators of coupled coastal processes remain an active research field. The United States National Hurricane Center’s P-Surge system uses forecast-error distributions to generate ensembles accounting for uncertainty in cyclone landfall location, forward speed, intensity, and size. Research seeks to make larger, more locally detailed ensembles computationally affordable. Saeed Saviz Naeini and Reda Snaiki’s 2024 study, for example, investigated a hybrid machine-learning model for rapid prediction of wave and storm-surge responses over an extended coastal region. These surrogate approaches emulate expensive numerical simulations; their reported efficiency and accuracy establish research potential, not operational status or demonstrated reliability for every coastline and future climate. (nhc.noaa.gov)

References

1. Defining Storm Surge, Storm Tide and Inundationnhc.noaa.gov
2. Storm Surge Overviewnhc.noaa.gov
3. Hurricane Readiness for Coastal Communities: Student Manual, April 2025nhc.noaa.gov
4. Drag coefficients derived from ocean current and temperature profiles at high wind speedsrepository.library.noaa.gov
5. L0311: Hurricane Readinessnhc.noaa.gov
6. The Saffir-Simpson Hurricane Wind Scalenhc.noaa.gov
7. Compound flooding in convergent estuaries: insights from an analytical modelos.copernicus.org
8. Storm surge - Met Officeweather.metoffice.gov.uk
9. Sound-side inundation and seaward erosion of a barrier island during hurricane landfallusgs.gov
10. TCFAQ A7) What is an extra-tropical cyclone?aoml.noaa.gov
11. Extra-Tropical Transitionaoml.noaa.gov
12. IPCC Third Assessment Report: Sea-level extremesarchive.ipcc.ch
13. World’s deadliest tropical cyclone was 50 years agowmo.int
14. Cyclone Amphan highlights the value of multi-hazard early warningspublic.wmo.int
15. PAGASA: Storm Surgepagasa.dost.gov.ph
16. 2013 State of the Climate: Record-breaking Super Typhoon Haiyanclimate.gov
17. Floods and floodingweather.metoffice.gov.uk
18. Data Series 220pubs.usgs.gov
19. USGS Data Series 294 — Monitoring the Storm Tide of Hurricane Wilma in Southwestern Florida, October 2005pubs.usgs.gov
20. Identifying and preserving high-water mark datapubs.usgs.gov
21. Hurricane Storm-Surge GNSS Surveying Campaignswater.usgs.gov
22. Operational Storm Surge Forecasting at the National Hurricane Center: The Case for Probabilistic Guidance and the Evaluation of Improved Storm Size Forecasts Used to Define the Wind Forcingslosh.nws.noaa.gov
23. UNIT 3: Forecast Uncertaintynhc.noaa.gov
24. Storm Surge – Basicsgithub.com
25. National Storm Surge Risk Maps - Version 4nhc.noaa.gov
26. P-Surge - MDL - Virtual Labvlab.noaa.gov
27. Potential Storm Surge Flooding Mapnhc.noaa.gov
28. Update on National Hurricane Center Products and Services for 2016nhc.noaa.gov
29. Storm Surge Maximum Envelope of Water (MEOW)nhc.noaa.gov
30. Storm Surge Maximum of the Maximum (MOM)nhc.noaa.gov
31. National assessment of hurricane-induced coastal erosion hazards: Mid-Atlantic Coastpubs.usgs.gov
32. Coastal change from Hurricane Sandy and the 2012-13 winter storm season: Fire Island, New Yorkpubs.usgs.gov
33. The Thames Barrier - GOV.UKgov.uk
34. The Thames Barrier – protecting London and the Thames Estuary for 40 yearsgov.uk
35. 40th Anniversary of the Tropical Cyclone Programmewmo.int
36. Hurricane Sandy: Tracking Water Levelsusgs.gov
37. Monitoring Storm Tide and Flooding from Hurricane Sandy along the Atlantic Coast of the United States, October 2012pubs.usgs.gov
38. Report of the WMO/UN-ESCAP/Typhoon Committee Expert Mission to the Philippines, 7–12 April 2014old.wmo.int
39. Using Probabilities to Improve Tropical Cyclone Preparednesspublic.wmo.int
40. Annex II: Glossary | Climate Change 2022: Impacts, Adaptation and Vulnerabilityipcc.ch
41. GitHub - nhrap-hazus/FAST: Hazus - Flood Assessment Structure Toolgithub.com
42. 100-year flood–it's all about chanceusgs.gov
43. The 100-Year Floodpubs.usgs.gov
44. Accounting for Climate Change in Extreme Sea Level Estimationarxiv.org
45. National Storm Surge Risk Maps - Version 4nhc.noaa.gov
46. Hurricane FAQ - NOAA/AOMLaoml.noaa.gov
47. Designing for Flood Levels Above the Minimum Required Elevation After Hurricane Ianfema.gov
48. Technical Summary | Climate Change 2022: Impacts, Adaptation and Vulnerabilityipcc.ch
49. Storm Surge Productsnhc.noaa.gov
50. Take Actioncoast.noaa.gov
51. Thames Barrier closed for 200th timegov.uk
52. Wetland Benefitscoast.noaa.gov
53. Natural and Structural Measures for Shoreline Stabilization: Living Shorelinescoast.noaa.gov
54. Public Weather Service Value for Money Review — March 2015metoffice.gov.uk
55. Chapter 9: Ocean, Cryosphere and Sea Level Changeipcc.ch
56. Chapter 11: Weather and Climate Extreme Events in a Changing Climateipcc.ch
57. Figure 11.19ipcc.ch
58. IPCC AR6 Working Group 1: Summary for Policymakersipcc.ch
59. A Novel Hybrid Machine Learning Model for Rapid Assessment of Wave and Storm Surge Responses Over an Extended Coastal Regionarxiv.org