The University of the West Indies Organization of American States

PROFESSIONAL DEVELOPMENT PROGRAMME: COASTAL INFRASTRUCTURE DESIGN, CONSTRUCTION AND MAINTENANCE

A COURSE IN COASTAL DEFENSE SYSTEMS II

CHAPTER 4

DESIGN OF COASTAL DEFENSE WORKS

DAVID SMITH, PhD Smith Warner International Limited, Consulting Engineers, Kingston Jamaica.

Organized by Department of Civil Engineering, The University of the West Indies, in conjunction with Old Dominion University, Norfolk, VA, USA and Coastal Engineering Research Centre, US Army, Corps of Engineers, Vicksburg, MS, USA. Dominica, West Indies, July 30-August3, 2001

Design of Coastal Defense Works: Marine and Coastal Processes David A.Y. Smith, Ph.D., P.Eng.1

Part I Overview of the Processes The islands of the Eastern Caribbean stretch from the Virgin Islands in the north, to Trinidad in the south. Geologically, these islands differ, however the majority have volcanic origins. Exceptions to this majority include Barbados and Antigua, which have large coral caps. These islands all have dual weather exposure, with their eastern shorelines exposed to the Atlantic and their western shorelines open to the Caribbean . From an overview perspective, these islands are exposed to the following forces and elements: • The Trade ; • Waves which are generated by: the Trade Winds; by passing hurricanes; and by North Atlantic storms; • Oceanic and tidally driven currents; and • change. These four parameters are the primary driving forces that contribute to ongoing marine and coastal processes in the islands of the Eastern Caribbean. They therefore need to be understood and/or quantified in order to properly design coastal defense works.

1.1 The Trade Winds The Trade Winds blow with great constancy primarily from the north-east to the south-east. Some seasonal changes occur within this pattern as a result of the relative position of the sun and the ’s surface. On March 21st, the sun is overhead at the equator. It moves overhead the Tropic of Cancer (22 ½oN) on June 21st, and returns overhead the equator again on September 21st. Between September 21st and March 21st the sun is overhead south of the equator. These celestial movements result in a natural division of the annual climate into four seasons: a. December to February: Winds are primarily from the NE to ENE. b. March to May: Winds are mainly from the East. c. June to August: Winds are primarily from the E to ESE. d. September to November: Winds are mainly from the E to SE. Wind speeds are also influenced by the location of the Inter-tropical , or ITC. The ITC is formed as a result of the convergence of north-east and south-east Trade Winds in a belt around the equator. This belt migrates north or south of the equator along with the sun’s motion. Since the ITC is characterized by wind uplift (as a result of convergence), surface wind speeds tend to be low in the vicinity of this feature. The ITC is closest to the Eastern Caribbean Islands

1 Smith Warner International Ltd. Unit 2, Seymour Park, 2 Seymour Avenue Kingston 10, Jamaica Coastal Defense Works – Caribbean Marine and Coastal Processes 2 Coastal Defense Systems II CDCM Professional Development Programme, 2001 between June and November. These months, therefore, have the lowest average wind speeds as compared with the rest of the year. These seasonal variations in wind directions result in a corresponding variation in wave directions. Disturbances to this normal circulation occur throughout the year as a result of the passage of easterly waves, hurricanes, tropical storms, and localized meteorological phenomena such as thunderstorms.

1.2 Wave Climate of the Eastern Caribbean The wave climate of the Eastern Caribbean Islands has three primary components: • Day-to-day (or operational) waves; • waves; and • Hurricanes.

1.2.1 Operational and Swell Waves The day-to-day wave climate occurs as a result of the action of the Trade Winds on the waters of the , and is observed throughout the year, primarily from directions NE, through E, to SE. Because of the constancy of the winds, the windward shores of these islands are exposed to high-energy wave conditions on a near-constant basis. Interestingly, recent work on available wave energy has shown that the frequency of occurrence of a given has increased over the past three decades. By contrast, and again as a result of the directional characteristics of the Trade Winds, the west coasts of these islands are relatively sheltered (compared to their east coasts), and the day-to-day wave climate along such coastlines are largely as a result of diffracted waves traveling around their north and south tips. Because of the predominance of the north-easterly component of the Trades, the south-going diffracted wave climate typically prevails, although there are times of the year when the predominant diffracted wave direction is to the north. Between November and March, the islands, and in particular their west coasts, are subjected to swell waves which are generated by extra-tropical storms occurring in the North Atlantic. Specifically, during these winter months, a large number of originate over the Gulf of Mexico and track in a north-easterly direction along the east coast of the USA, as far north as Newfoundland. These extra-tropical depressions are slow-moving and so the winds under their influence have ample time to generate an active . These generated waves move southwards, away from the cyclones, and travel over 1,000 km to Eastern Caribbean shorelines. During this passage, the waves become regular in shape (i.e. sinusoidal) and have long wave lengths (i.e. long period waves). Because of their direction of travel (i.e. from the north to NNE), they have the most impact on leeward shorelines, which are otherwise sheltered, and can cause a great degree of damage to these shorelines. These swell events last on average between 1 to 3 days, and there are usually 5 to 9 of them in any one swell season. Because of the long wave period characteristics of these waves, they experience a great degree of shoaling and refraction in the nearshore waters of these leeward coasts, and can contribute substantially to the movement of sand in the , in a southerly direction.

Coastal Defense Works – Caribbean Marine and Coastal Processes 3 Coastal Defense Systems II CDCM Professional Development Programme, 2001

Recently measured wave data collected on the lee side of the Barbadian coast has shown that during the summer months, there is some evidence of swell from the south-west. These events appear to be caused by tropical waves and storms passing through the southern Caribbean. The water waves generated by these tropical waves travel in a north-easterly direction and impact on the leeward shores of these islands, moving sand in the surf zone in a northerly direction. They are similar in wave height characteristics to the north swell, with the exception that their wave periods are somewhat shorter. This is to be expected, since they travel over a smaller area of water than the swell originating in the North Atlantic. These events are more rare than northern swell and occur, on average, once every two to three years or so. These occurrences are also known to create severe erosion on the lee side beaches

1.2.2 Hurricane Waves The third component of the wave climate that affects Eastern Caribbean shorelines is due to the passage of tropical storms. These meteorological features traverse the Caribbean between June and November (the hurricane season). They have an organized circulation structure and are characterized by winds rotating around a central core, or “eye”. In the , the winds rotate in an anti-clockwise direction, whereas in the they rotate in a clockwise direction and are called typhoons. The majority of cyclones that affect the Caribbean have their genesis on the African continent ( region) and travel across the Atlantic. Usually, as they make this trans-Atlantic crossing, they gain energy from the waters over which they travel, and develop a more organized structure. For these storms, their first landfall are the islands of the Eastern Caribbean. Less frequently, tropical cyclones originate in the south-west of the . These usually affect the north-western Caribbean but, as in the case of Hurricane Lenny, can travel eastward across the Caribbean sea. These storms can also be quite damaging. The term “tropical ” refers to any non-frontal, low pressure, large-scale weather system that develops over tropical or sub-tropical waters, and possesses a definite organized circulation. They have historically been classified according to their maximum sustained wind speeds. Cyclones with wind speeds below 34 knots (63 km/hr) are known as Tropical Depressions. Those with wind speeds between 34-64 knots (63-118 km/hr) are termed Tropical Storms, while the term Hurricane is used for tropical cyclones with sustained wind speeds over 64 knots (118 km/hr). In the USA, a further classification of hurricanes is in common usage. This classification system, known as the Saffir/Simpson Hurricane Scale, describes five scales of hurricane strengths according to wind speed. These range from Category No. 1, starting at 74 mph (119 km/hr) up to a Category No. 5, with speeds in excess of 155 mph (250 km/hr). Historically, however, Category No. 5 hurricanes have not been observed in the Eastern Caribbean. It is interesting to note that an earlier categorization was developed for the Eastern Caribbean (Depradine and Rudder, 1973), and is given in the following Table 1.1.

Coastal Defense Works – Caribbean Marine and Coastal Processes 4 Coastal Defense Systems II CDCM Professional Development Programme, 2001

Table 1.1 Regional Caribbean Classification of Tropical Cyclones Storm Category Wind Speed (knots) Central Pressure (mb) I 20-44 1006+ II 45-70 1000-1006 III 65-109 970-1000 IV 110+ <970

In addition, this classification was aided by a division of categories III and IV into three latitude groups. Group A encompassed a range of from 10-15oN, Group B from 15-20oN and Group C was north of 20oN. Observation of historical hurricane data revealed that for Category I and II storms, the influence of latitude on the parameters given above was minimal. For Category III and IV storms, however, there were notable changes in the storm parameters with latitude. Generally, central pressure and forward speed of these storms decreased with increasing latitude (i.e. the cyclones became more intense with increasing latitude), whereas radius to maximum winds increased with increasing latitude (potentially larger storms).

1.3 Currents and

1.3.1 Oceanic Currents The primary ocean currents that influence the islands of the Eastern Caribbean are the North Equatorial and Currents. The former current crosses the Atlantic, running in an east to west direction just north of the equator, at a mean speed of 0.5 knot. The Guyana Current runs parallel to the coast of , generally flowing from the south-east to the north-west at speeds of up to 1 knot. The Guyana Current appears to dominate flow around the southern Caribbean between January and April, when it brings low salinity waters north from the Amazon and Orinoco Rivers. This influx of fresh water has been found have a profound effect on nearshore water levels.

1.3.2 Tidal Action In general, tides are caused by the gravitational effects of the sun and moon on the of the world. The periodicity of the rise and fall of the , known as the flood and ebb, is determined by the periodicity of these gravitational effects. Additionally, the height of the tide is a function of the sum of water displacements produced by these gravitational forces. Closer in to shore, the effect of the tides becomes noticeable, as their amplitudes and velocities (but not frequencies) are modified by coastal . Typically, tidal fluctuations occur once daily (diurnal tide) or twice daily (semidiurnal tide), with unequal amplitudes. Observations of tides over a lunar month reveal periodic variations in the (difference between successive high and low tides). From these records, times of maximum and minimum tidal range are observed. The maxima are called spring tides and the minima, neap tides. These occur at approximately two-week intervals.

Coastal Defense Works – Caribbean Marine and Coastal Processes 5 Coastal Defense Systems II CDCM Professional Development Programme, 2001

1.4 Sea Level Change Two primary forms of quantitative information are available for the assessment of sea-level change in the Caribbean. These are changes deduced from the sedimentary record, and from in situ measurements. Sea-Level Movements Deduced From the Sedimentary Record of Coastal and Shelf Areas: Event timing has been generated through radiometric dating of sedimentary samples that were deposited at, or near to, sea level. The nature of these records is such that resolution greater than 100 years is difficult to achieve, and inherent measurement errors are typically on a scale of 100 years. Thus, while these records are useful as indicators of longer-term sea-surface changes for pre-historical periods, they cannot be used accurately for detection of sea-surface changes on a decadal, annual or intra-annual frequency. Records from Tide Gauges: Within the wider Caribbean region there are sixty-four tide-gauge stations with data of sufficiently high quality to have their data placed in the revised local reference file at the Permanent Service for Mean Sea Level (Hanson and Maul, 1993). There is a marked inconsistency in both the direction and rate of sea-level change from these recent historical records. This is typical for the Caribbean in general and is caused by local variation in factors that include the rate of tectonic displacement (land movement) at each location. For example, within the wider Caribbean, Port-au-Prince records one of the highest historical rates of sea-level rise. This appears to be due primarily to locally rapid subsidence of the crustal block on which Port-au-Prince is located. The global warming factor also needs to be taken into account. At present, while there are indications of atmospheric warming, there is as yet no definitive signal in the tide gauge data for the Caribbean to indicate accelerations in the rate of sea-level change. For the purpose of estimates, however, it is considered advisable to err on the side of caution when predicting future change, by taking into account some element of potential from this phenomenon. Assumptions for future global sea-level rise under a "greenhouse" scenario vary substantially between 0.3 cm/yr. to 1.0 cm/yr. The UNEP/IOC Task Team adopted a figure of about 0.5 cm/yr. for modeling purposes. The actual rate of change will obviously vary enormously depending on the local nature of tectonic movement, subsidence, fluid withdrawal and other contributors to relative sea level rise, and needs to be considered on an island-by-island basis.

1.5 Coastal Process/Shoreline Interactions in the Eastern Caribbean The elements and driving forces described in the previous sections have interacted with Eastern Caribbean shorelines in a number of different ways, and have influenced, in a profound manner, their present-day morphology. First, the sand dunes found on the windward shores of these islands form as a direct result of aeolian transport of sand. Essentially, sand that is carried to the back of the beach, by wave action, is dried and transported further inland by the prevailing wind. Any trees or shrubbery in the back beach area will trap this sand, resulting in the start of a sand dune. Once the dune forms, it will continue to grow (under the action of wind), and will only be eroded either by sand mining or during periods of high water levels and wave action.

Coastal Defense Works – Caribbean Marine and Coastal Processes 6 Coastal Defense Systems II CDCM Professional Development Programme, 2001

The waves that are generated in deep water undergo a number of changes as they approach these shorelines. These include: • Shoaling and refraction, which occur when the waves “feel” the ; • “Whitecapping” which takes place when the waves become over-steepened; • Interaction with, and reflection from, reefs where these occur adjacent to a shoreline; • Diffraction around headlands and obstructions (in fact, at a larger scale, this occurs around the tips of the islands); and • Run-up on shorelines. These changes affect, in a major way, the amount of wave energy remaining at the shoreline that is able to influence the shoreline morphology and to inflict damage to coastal infrastructure. For example, the emergence of headlands and bays comes about as an interaction between the shorelines’ geological formation and nearshore wave climate, and typically occurs over geologic time scales. These interactions further result in shoreline erosion and accretion and therefore are an integral part of the development of a sediment budget, with impacts on the medium and long-term equilibrium of the shoreline. Additionally, it is important to be able to quantify the nearshore and extreme wave climate, in order to facilitate the design of any coastal/marine works. The extreme wave climate, which is used in the design of any coastal structure, is usually derived from an analysis of hurricane parameters. Finally, an understanding of water levels is important. These can be divided into two main categories, day-to-day and event related. Water levels change on a daily basis as a result of tidal action. Typically, two high tides and two lows occur on a daily basis in the Eastern Caribbean. This daily movement of the mean sea level affects the shaping of shorelines, and the extent of the normal beach processes. During extreme events, such as a hurricane, water levels can increase as a result of storm surge. The components of storm surge are: • Wind-induced surge; • Inverse barometric pressure rise; and • Wave set-up. Storm surge can be devastating for a shoreline because it causes: • More wave energy to be transmitted over bank reefs into the nearshore zone, and, • Inundation of shoreline areas that would not normally be flooded. This increases the potential for backshore erosion. Another type of extreme event is , a phenomenon of long period ocean waves. In deep water, these waves may be hundreds of kilometres in length and only a metre or more in height. As they enter shoaling coastal waters, their wave lengths diminish and wave heights increase. This phenomenon can be devastating for low-lying coastal areas. may be generated by submarine earthquakes, volcanic eruptions, landslides, slumps and explosions. In the eastern Caribbean, the most likely source of tsunamis is the nearly continuous belt of shallow-focus seismicity, which can be traced from through the Greater and Lesser Antilles, to north-east Venezuela.

Coastal Defense Works – Caribbean Marine and Coastal Processes 7 Coastal Defense Systems II CDCM Professional Development Programme, 2001

In addition, a third category of water level change must be considered in the context of long term development planning. This is global sea level rise, which is a long-term phenomenon. As outlined above, these phenomena all combine to interact with, and shape, the Eastern Caribbean shorelines. These shorelines, however, differ, and may consist of a number of different types: sand beaches, rocky cliffs or mangrove marshes. The beaches of the Eastern Caribbean are typically formed from one, or (a combination), of three sources: • Volcanic sand which produces black sand beaches; • Riverine (i.e. terrestrial), sand which forms brown sand beaches, and • Coral fragments, which form white sand beaches. In many locations, sand from these sources mix at the shoreline. The ongoing transport of sand in an alongshore direction (as a result of wave action) contributes to the sediment budget, whereby sand is brought both into, and out of, a section of shoreline. Depending on the relative differences between these two rates, the shoreline may either erode or accrete. In the development of a Coastal Zone Management Plan, or in the preparation of coastal infrastructure designs, it is essential to have some knowledge of the sediment budget and the impact(s) of any proposed structures on it. As is to be expected, rocky shorelines are much more resistant to erosion than their sandy counterparts. These rock cliffs may be coral, or may be composed of a harder substrate such as granite. Coral cliffs have been documented to erode at the rate of approximately 5 mm/year, however, harder substrate types are much more resistant to erosion. Mangrove/marsh type shorelines typically occur in low wave energy environments. Because of the complex root structure of these plants, they tend to anchor the shoreline substrate. They also facilitate the settlement of silt and mud, which may be washed out of backshore areas during times of heavy rainfall.

Coastal Defense Works – Caribbean Marine and Coastal Processes 8 Coastal Defense Systems II CDCM Professional Development Programme, 2001

Part 2 Quantification of Coastal Process Elements

2.1 Reasons for Quantification In Part I, a description was given of the various coastal process components that contribute to the shaping of Eastern Caribbean shorelines. In order to be proactive in the evaluation of development proposals that are coastal or marine related, or to produce designs of coastal/marine projects which work in harmony with their environment, it is essential that a proper understanding of these processes be gained. This requires both an appreciation for the way in which these processes work and for their relative magnitude. In the Caribbean region, there is generally a lack of sound data on which to base a design project, to properly evaluate such a project, or to develop a coastal zone management plan. This means that short-term measured data must often be combined with empirical or regional observations to complete the background picture. It is important, therefore, to be aware of what data does exist, how this may be accessed and where the data gaps in the database lie. Further, it is important to be aware of the techniques of field measurement, what should be measured, where the limitations lie and what can reasonably be expected from such measurement programs. Because there have been few comprehensive data collection programs mounted in the Eastern Caribbean, additional techniques are often required to interpret measurements collected at spot locations, within the context of a spatial basis. Very often, numerical modeling (computer) techniques are used to fill in the spatial “picture” of coastal processes. It is important, however, to realize that these models should never be used in a “black box” manner. In other words, they should either be calibrated or their limitations should be fully understood. In the following sections, the available database is explored and methods are given for the quantification of the various parameters.

2.2 Wind Data At a regional level, wind data archives are held by the Caribbean Meteorological Institute (CMI), although not necessarily in a form which is directly suitable for computer analysis. On an island- by-island basis, wind data may be obtained from local airports. These data are typically in the form of wind speed and direction observations that have been made at specific intervals which range from once per hour, per day or per month. If a good digital database is to be collected and analysed, then it is often necessary to convert the airport wind data from an analog to a digital form. This information may be used to obtain an overview of the prevailing wind direction(s), and consequently of the impact of wind on coastal features or facilities. Further, it may be used in a wave hindcasting procedure to evaluate the characteristics of locally generated waves. In using this collected wind data, it is important to properly interpret the data in the context of the location of data collection vis-à-vis the site under consideration.

2.3 Wave Data The wave climate at a given location is typically a collection of wave statistics, which represents the long-term average frequency of occurrence of conditions at that location. The

Coastal Defense Works – Caribbean Marine and Coastal Processes 9 Coastal Defense Systems II CDCM Professional Development Programme, 2001 wave climate usually includes ordinary wave conditions, and/or special events such as hurricanes. The data within the wave climate are usually defined in terms of: • The event or season; • The wave direction; • A measurement of wave height; and • A measurement of wave period. Ideally, the wave climate at a location is best obtained through a long-term period of measurement, spanning years. For the evaluation of hurricane waves, it is customary to review records of greater than 100 years. At this point it is instructive to present some definitions of and statistics for, water waves. Waves generated in the oceans are known as gravity, or wind-generated, waves, with periods ranging from 1 to 30 seconds typically. These waves form when winds blow over a flat area of sea, and change from ripples to full waves, through an extremely complex development process. Other factors that play a part in this process are depth of water, storm duration and distance over which the wind influences the waves (known as the “fetch”). For conditions where water depth, wind duration and fetch are unlimited, a Fully Arisen Sea develops. Waves that are still within the area of generation, are termed sea, whereas swell wave are those which have traveled some distance away from the area of generation into the observer area. During this period of travel, the smaller, short-period waves are “eaten up” by the longer period components of the sea. In addition, the energy of the individual waves is dissipated to some extent. The net result is that when the swell waves are observed after traveling some distance away from their generation area, they have longer wave periods and smaller, more regular wave heights. Because a sea state is comprised of waves with differing heights and periods, a tremendous amount of research has gone into the statistical and/or probabilistic definition of these waves. One of the most common, and more important, definitions of wave height is the significant wave height, Hs. This definition seems to some closest to the visual estimate that would be made by an experienced observer.

Hs - significant wave height, or average height of the highest 1/3 of the waves In the design of a coastal or marine facility, there is interest not only in what takes place during a particular storm, but also over the expected life of the structure. Research has shown that a binomial distribution will give the probability that an event will not occur during the life of a structure. The return period for this is calculated as: 1 T = 1 11−−()R N Where T = design return period; R = permissible risk of failure; and N = expected project life. For example, if a structure with an expected 50-year life was designed for a 20-year return period storm, there is a 33% chance this event will occur during the life of the structure.

Coastal Defense Works – Caribbean Marine and Coastal Processes 10 Coastal Defense Systems II CDCM Professional Development Programme, 2001

Measured wave data may be subjected to a wave spectrum analysis (done by computer analysis only). In this method, the wave record is digitized and a wave spectrum computed. This is a measure of the distribution of wave energy with frequency. Integration of the wave spectrum gives another estimate of the variance of the water surface from the mean water level (σ), which can be used to estimate characteristic wave heights (note that Hs = 4.0σ). The wave spectrum can also be used to give an energy density spectrum, which gives the distribution of energy with frequency. This allows the study of resonant systems. It is normal to characterize the spectrum by its peak frequency, fp. From this, the peak period is given as,

Tp = 1/fp.

2.3.1 Deep Water Waves In the absence of detailed wave measurements at a particular location, it is possible to rely on long term observations of sea state conditions that have been collected by volunteer observer ships (VOS). Essentially, this data has been derived from a quality enhancing analysis of a massive number of visual observations of both waves and winds, which are reported from ships in normal service all over the world, using a computer program called NMIMET. The statistics are presented for waves only, but the wind data has been used to improve the reliability of wave statistics. The parameters measured correlate to what are known as “significant wave heights” and “zero- crossing” wave periods (described in the previous section). The areas of wave parameter recordings that are applicable to the Eastern Caribbean are Areas 47 and 48. For the islands of the Eastern Caribbean, this data source is considered to be quite reliable for the exposed, or windward, shorelines. For the sheltered shorelines, however, it is not so reliable as information on secondary wave trains is typically masked in the visual observations. In reality, these secondary wave trains combine with diffracted waves, around the tips of these islands, to produce a leeward coast wave climate. These secondary wave trains tend to be of the same order of magnitude as the diffracted waves, on these leeward shorelines. A NOAA wave buoy moored in the south-central Caribbean Sea provides one potential data source for the lee coast wave climate. This buoy has been measuring wave heights and periods, and wind directions, since 1994. This data is available on the internet, from the NOAA web site. The one drawback to this data is that it contains no directional wave data. The assumption most therefore be made that the wave and wind directions are coincident. At a few locations within the Caribbean, such as Barbados, extensive wave measurements have been made on both leeward and windward shoreline. These data have shed quite some insight into the characteristics of swell that is experienced on leeward coastlines. For locations where no measured data exists, a hindcast model data may be obtained. Two such avenues are: • It is possible to specially order VOS wave statistics for small areas adjacent to any of the Eastern Caribbean islands, from British Maritime Technology (BMT); • It is also possible to order wave data generated from a Global Spectral Ocean Wave Model (GSOWM). These data are available in 12-hour intervals since June 1986, and consist of a time series of directional frequency spectra with a directional resolution of 15o and a frequency range giving wave periods ranging from 3 to 25 seconds.

Coastal Defense Works – Caribbean Marine and Coastal Processes 11 Coastal Defense Systems II CDCM Professional Development Programme, 2001

In order to assess a hurricane wave climate, the recommended procedure would be to search the database of storms held at the National Climatic Data Centre (in cooperation with the National Hurricane Centre). These records date back to 1871 and include nearly 1000 storms. The records of these storms are updated on an annual basis, and include data on position, wind speed and central pressure. A regional analysis of hurricanes was done as part of a Feasibility Study on Coastal Conservation for the Government of Barbados. This work gave the following values of regional hurricane wave parameters: Table 2.2 Extremal Analysis of Maximum Regional Hurricane Wave Heights

Return Period Wave Height, Hs (m) 95% Confidence Peak Period, Tp (sec) (Years) Interval (m) 5 9.5 8.2 - 10.9 12.0 10 11.8 10.3 - 13.7 14.0 25 14.8 12.6 - 17.0 15.3 50 16.9 14.5 - 19.4 16.0 100 19.0 16.2 - 21.8 16.6

2.3.2 Nearshore Wave Climate In order to develop a wave climate that is representative of nearshore conditions in the vicinity of a project or beach, it is necessary to undertake a wave transformation procedure on the deep water waves. Waves begin to “feel” the seabed when they travel into water depths characterized by:

Lo Lo 5 < d < 2 As these waves propagate in to shore, they are subject to: shoaling; refraction; diffraction around seabed and shoreline features; and energy losses due to bed friction and “whitecapping”. These changes primarily affect the wave height, and to a lesser extent the wave period. There are a number of computer programs that are currently in use to carry out this type of transformation procedure. These typically require bathymetric input as well as deep-water wave characteristics, and fall generally into one of two categories. For the first, the tracks of wave rays are computed in reverse, starting at the inshore location and ending in deep water. A large number of wave rays may be subjected to this procedure, to include all possible wave directions. In addition, a representative range of wave spectrum peak periods is typically used. For this procedure, it is possible to determine the range of offshore directions that are able to arrive at a site in question. Further, it is possible to compute the inshore wave spectrum corresponding to a particular offshore spectrum. For a particular offshore wave condition (wave height, period and direction), this procedure can be used to give the corresponding nearshore conditions of wave height and direction. The second method (known as the forward tracking procedure) is more traditional, and traces the travel of wave rays (which are perpendicular to the wave crests) from deep water in to shore. For

Coastal Defense Works – Caribbean Marine and Coastal Processes 12 Coastal Defense Systems II CDCM Professional Development Programme, 2001 this method, a different wave direction and period are used for each simulation, with simple monochromatic waves as input. The strong point of this method is that it gives a good graphical representation of the patterns of wave propagation over the nearshore shallow water regions. This is useful in identifying areas along the shore where wave energy concentrates or disperses.

2.4 Water Levels The development of coastal zone management plans, emergency response plans and the design of coastal structures rely heavily on an assessment of extreme water levels. For design purposes, the 50 or 100 year return period events are usually considered. The design water level is made up of four major components: • Tidal fluctuations; • Inverse barometric effect; • Wind surge; and • . For long-term planning scenarios, global sea level rise should also be included. These components are discussed in detail following.

2.4.1 Tidal Fluctuations As previously described, tides in the Eastern Caribbean are primarily semi-diurnal in nature (i.e. two highs and two lows per day). The parameters that are usually used to characterize the tide are: • Mean high high water; • Mean low high water; • Mean sea level; • Mean high low water; and • Mean low low water. In addition, the tidal range is usually of importance, for both spring and neap conditions. In the Eastern Caribbean, the tidal range is typically less than 1.0 metre.

2.4.2 Inverse Barometric and Wind Surge Effects As a zone of low pressure (from a or hurricane) moves over a body of water, the reduction in atmospheric pressure (below ambient levels) results in the raising of the mean water level. The empirical formula used to predict this effect is given by:

R S = ()−−()r h 01.Pn Peo 1

Where: Sh = The increase in water level (metres)

Po = The central pressure (kPa)

Pn = The ambient pressure (kPa) R = Radius to maximum winds (km)

Coastal Defense Works – Caribbean Marine and Coastal Processes 13 Coastal Defense Systems II CDCM Professional Development Programme, 2001

r = Radius to a particular location (km) The wind set-up component can be evaluated using a one-dimensional surge model, given by:

2 F Sw = KW d

Where, Sw = The wind surge W = The wind speed F = The fetch length d = The water depth K = A calibration coefficient. These expressions are given in the Shore Protection Manual (1977). The water level increase resulting from low pressure is a maximum at the centre, or “eye”, of the cyclone. By contrast, the increase resulting from wind set-up is a maximum at the radius of maximum winds. The two effects are, therefore, not directly additive. Investigation into the relative additive strengths of these two components has shown that the inverse barometric pressure effect predominates when the two effects are combined.

2.4.3 Wave Set-up Wave set-up is defined as the increase in average water level due to waves breaking in the surf zone. This phenomenon results from the conversion of dissipated kinetic wave energy into potential energy, which takes the form of an increase in the average water level. The calculation of wave set-up is therefore triggered at the point at which wave breaking starts. In reality, the larger waves break further offshore and the smaller waves travel closer in to shore. All breaking waves, however, contribute to the wave set-up.

2.4.4 Tsunami Three submarine volcanoes have been reported as being active in recent times in the Eastern Caribbean. These are: • Kick-em-Jenny, north of Grenada • Holder’s Volcano, west of St. Lucia and • An un-named volcano north of Marie Galante. A historical summary of tsunamis affecting the Eastern Caribbean has been prepared (Deane et al, 1973). These date back to 1530. In recent years, tide gauge anomalies have been compared with data on earthquakes prepared by the US Department of Commerce and the Seismic Research Unit. These have resulted in an updating of the historical records. An evaluation of tsunami return periods was developed by Deane et al (1973). These were translated to design water levels (from tsunami) based on the assumption that tsunami will undergo little or no shoaling within the Eastern Caribbean islands. The tsunami wave can then be treated as a solitary wave with most of the height being above the mean sea level (Delcan, 1994). This approach resulted in the following design values. Table 2.3 Design Tsunami Water Levels

Coastal Defense Works – Caribbean Marine and Coastal Processes 14 Coastal Defense Systems II CDCM Professional Development Programme, 2001

Return Period (years) Water Levels Relative to LLW (m) 4 1.3 20 1.7 75 2.5 200 4.1 - 5.7

2.5 Currents There is no long-term database of current data for the Eastern Caribbean that can be accessed in the same way that is available for wave data. Reliance must, therefore, be placed on the collection of measured data. A number of different options exist for the measurement of currents in the field. These are: • Using drogues • Deploying moored current meters. Drogue tracking techniques give an estimate of the spatial picture of nearshore current characteristics. The drogues are deployed with sails set at varying depths, and are then tracked to record speeds and directions. These measurements, therefore, give rise to an estimate of water particle movements in a spatial sense. Currents may also be measured using moored current meters. These may be moored close to the seabed, or within the . These meters measure currents at the location of measurement. Where flow is 2-dimensional, a single (or double) meter deployment will give a good representation of the current characteristics within the water column. An acoustic doppler current meter may also be used to measure currents throughout the water column. These are usually mounted on the seabed and scan upwards to the surface. This type of instrument gives a good indication of any 3-dimensional characteristics within the flow.

2.6 Sediment Transport Characteristics An evaluation of the sediment transport regime is essential in developing an understanding of the sediment budget along a shoreline, or in estimating the impact of proposed coastal/marine developments on adjacent shorelines. Given that there are very few locations within the Caribbean where sediment transport rates have been actively measured, it is necessary, in most cases, to estimate this parameter through empirical methods or computer modeling. Sediment transport rates may be measured when a , for example, is constructed in the surf zone and the resulting sand fillet surveyed over known periods of time. In the computation of transport rates, and also to aid the understanding of beach erosion and/or accretion trends, the following input data is required: • Beach profile data; • Sediment size characteristics; and • Nearshore wave data or deep-water data plus offshore bathymetry. Since 1988, a series of quarterly beach profile measurements has been carried out at 21 locations throughout the Eastern Caribbean. This has been done as part of a long-term regional beach

Coastal Defense Works – Caribbean Marine and Coastal Processes 15 Coastal Defense Systems II CDCM Professional Development Programme, 2001 monitoring program (COLSAC). Where no such data is available at a project location, it is necessary to carry out a beach profile survey. Such a survey should extend from the back of beach area (landward of expected wave uprush) out to approximately a 1 metre water depth. Sediment samples should be collected from the active beach face and from the nearshore area in approximately 0 to 1.0 metre water depth. These samples should then be air dried (if sand) and subjected to a standard sieve analysis. Where there is an appreciable quantity of silt in the beach sample, it may be necessary to do a hydrometer analysis. The computation of potential sediment transport may be carried out using either a “bulk” predictor or a detailed predictor model. The bulk model uses one characteristic wave height, the deep-water significant wave height, to compute wave energy and subsequently potential alongshore sediment transport. As may be expected, this method typically overestimates the actual rate of sediment transport, and works best in situations where there is a virtually unlimited supply of sand. The detailed predictors, by contrast, provide for the division of the cross-shore profile into segments. The incoming waves are tracked through each segment and the amount of wave breaking and remaining wave energy accounted for. This approach, therefore, provides a profile of alongshore sediment transport throughout the surf zone. It is possible, with these predictions, to include the effects of tidal or oceanic currents on sand transport. It is also possible to model, in a spatial sense, the response of a shoreline or beach to the addition of a groyne, breakwater, or dredge programme. These morphological models provide good insight into the changes that can take place to a shoreline in those development scenarios. They are, therefore, useful planning tools. Often, in the development of a coastal project, both types of sediment transport models will be used. The 1-line model, to get a sense of shoreline adjustment potential, in a cross-shore sense, and the planform model, to estimate the impacts on adjacent properties.

2.7 Computer Modeling in the Marine Zone A number of computer modeling techniques have been described in the foregoing sections. These have dealt primarily with wave and sediment transport processes. It is also possible to model, by computer, current hydrodynamics. Essentially, 2- or 3-dimensional hydrodynamic models may be used to model current phenomena in a spatial manner. These models use either finite element or finite difference techniques. The finite element models have become more popular in recent years for the following reasons: • It is possible to provide better representation of the model land boundaries; • it is possible to obtain better detail of specific areas within the model grid, then with finite difference models which have uniform grid spacing. As may also be expected, the 2-dimensional models are much more efficient to run than the 3- dimensional models. These models, as with all computer models, should be calibrated. The driving parameters are usually tide and/or wind action, although river inflows may also be simulated. Results obtained give current patterns over large-scale areas that should include points of measurement.

Coastal Defense Works – Caribbean Marine and Coastal Processes 16 Coastal Defense Systems II CDCM Professional Development Programme, 2001 References Battjes, J.A. and Stive, M.J.F., (1985). Calibration and Verification of a Dissipation Model for Random Breaking Waves. J. Geophys. Res. Vol. 90, No. C5, pp. 569-588. Deane, C.A.W., (1969). Hindcast Wave Statistics for Atlantic Coasts of . Journal assocn. Of Prof. Engrs. (Trinidad and Tobago) Vol. 8, No. 2, pp. 27-50. Deane, C.A.W., Thom, M., Edmunds, H., (1973). Eastern Caribbean Coastal Investigations (1970- 73). Volume II - Natural Forces. Regional beach Erosion Control Programme, Faculty of Engineering, U.W.I. Delcan (1994). Water Levels for Barbados. Feasibility Studies on Coastal Conservation. Government of Barbados/In ter-American Development Bank. Depradine, C.A., Rudder, G.M., Lamming, S.D., (1973). Some Characteristics of Hurricanes in the Eastern Caribbean. Caribbean Meteorological Institute. Harrison, K., and Maul, G.A., (1993). Analysis of temperature, and sea-level variability with concentration on , , for evidence of trace-gas-induced climate change. Pp. 193-211 in Maul, G.A., editor, Climatic Change in the Intra-America’s Sea, Edward Arnold, London, p. 389. U.S. Army Coastal Engineering Research Centre (CERC). 1977. Shore Protection Manual.