6

Coastal Processes

David M. Bush, Duke University, Durham, North Carolina

INTRODUCTION

Field observations for this report were made while the author was a member of the Hurricane Hugo postdisaster study team sponsored by the National Research Council's Committee on Natural Disasters. Field work in Puerto Rico was done from September 21 to September 24, 1989, with the team; and from September 25 to September 30, 1989, with both the United States Geological Surveys (USGS) Marine Geology Project Office in San Juan and the Puerto Rico Department of Natural Resources (PRDNR).

Hurricane Hugo left a pronounced mark on the shoreline of Puerto Rico and the Virgin Islands. The impact of the storm was intensified in places by extremely dense shorefront development and shoreline structures (i.e., seawalls). Effects of the storm were mitigated by low storm-surge height, the rocky shoreline, and the steep coastal grade in some areas. The following sections detail the geologic processes active during hurricanes and the impact of Hugo on the Puerto Rican and Virgin Islands shorelines.

STORM SURGE AND SHORE PROCESSES

During storms such as Hurricane Hugo coastal-zone physical processes such as storm surge, wind, waves, overwash, and storm-surge ebb are intensified. Storm surge is the difference between fairweather and storm water levels, due largely to the combined effects of wind pushing water onshore—literally piling the water up against the land—and extremely low atmospheric pressures, which actually bulges the ocean upward. These forces combine to force the ocean landward, bringing slightly deeper water to places that have normally shallow water, or bringing water into places that are normally above sea level. As water depths increase, maximum potential wave heights also increase. Thus, there is a direct relationship between bringing deeper water farther inland by a positive storm surge and allowing larger waves to travel farther inland. In other words, not only are increased winds and wave energy



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA 6 Coastal Processes David M. Bush, Duke University, Durham, North Carolina INTRODUCTION Field observations for this report were made while the author was a member of the Hurricane Hugo postdisaster study team sponsored by the National Research Council's Committee on Natural Disasters. Field work in Puerto Rico was done from September 21 to September 24, 1989, with the team; and from September 25 to September 30, 1989, with both the United States Geological Surveys (USGS) Marine Geology Project Office in San Juan and the Puerto Rico Department of Natural Resources (PRDNR). Hurricane Hugo left a pronounced mark on the shoreline of Puerto Rico and the Virgin Islands. The impact of the storm was intensified in places by extremely dense shorefront development and shoreline structures (i.e., seawalls). Effects of the storm were mitigated by low storm-surge height, the rocky shoreline, and the steep coastal grade in some areas. The following sections detail the geologic processes active during hurricanes and the impact of Hugo on the Puerto Rican and Virgin Islands shorelines. STORM SURGE AND SHORE PROCESSES During storms such as Hurricane Hugo coastal-zone physical processes such as storm surge, wind, waves, overwash, and storm-surge ebb are intensified. Storm surge is the difference between fairweather and storm water levels, due largely to the combined effects of wind pushing water onshore—literally piling the water up against the land—and extremely low atmospheric pressures, which actually bulges the ocean upward. These forces combine to force the ocean landward, bringing slightly deeper water to places that have normally shallow water, or bringing water into places that are normally above sea level. As water depths increase, maximum potential wave heights also increase. Thus, there is a direct relationship between bringing deeper water farther inland by a positive storm surge and allowing larger waves to travel farther inland. In other words, not only are increased winds and wave energy

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA associated these storms but the landward incursion of storm water means the impact of those forces are felt farther inland. There is also a direct positive correlation between wind velocity and wave height. Therefore hurricanes, with characteristically high wind velocities, drive larger waves, and a positive storm surge allows these waves to penetrate farther inland. The wind and wind waves can also force additional water shoreward, increasing flooding. When the storm passes and the driving forces of wind and air pressure are relaxed, the water that has been piled up on the land surface begins to flow back toward the sea, essentially flowing downhill. This phenomenon is called storm-surge ebb (or return) flow. This flow is commonly funneled through constrictions that typically are narrow seaward, such as tidal channels and rivers or buildings and homes near the shoreline. Water velocity increases simply by moving a given volume of water through a decreasing cross-sectional area. This flow is either entirely gravity driven or is increased by reversed wind flow associated with the passing of the hurricane. Winds that first blow onshore, working to pile-up water against the land, turn offshore, accelerating the storm-surge return flow seaward. The geologic processes driven by the aforementioned physical forces are wave scour and turbulence, wave attack on structures and the shoreline, wave reflection, storm water (and sand) overwash, and dune and beach retreat. Eroded beach and dune sand can be moved offshore into deep water and out of the beach system by storm-surge ebb or by offshore-moving pressure-gradient currents. Sand is thereby transported into water so deep that normal, “fairweather” waves that typically act to move sand onshore during summer months cannot do so. SHORELINE DESCRIPTIONS The Puerto Rican shoreline is very rocky in places, especially along the northern coast where several rows of Pleistocene sand dunes have been cemented to form a moderately resistant rock called eolianite. Eolianite is only moderately resistant to erosion by the sea, but offers much more protection for the coast than unconsolidated beach sand. In addition to offering physical protection from waves, the rocky shoreline affords some elevation upon which structures are located above some storm surges and, thus, less likely to be affected by direct storm-wave attack. Probably the best action that can be taken to reduce the risk from storms is to elevate structures and people at risk—get them up out of the storm waters. Perhaps the best example of this is the city of Old San Juan, built at high elevation near the entrance to San Juan Harbor. The city has resisted uncounted attacks (from both nature and warring nations), and undoubtedly its elevation has helped on both counts. The coastal lowlands of Puerto Rico are narrow, and the island is very steep, meaning that the storm-surge effects are not felt very far inland. Unfortunately, the narrow lowland area has more or less constrained much of the development to this hazardous zone.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA The shoreline setting of the Virgin Islands is very similar to that of Puerto Rico. There are stretches of sandy beaches, but there is also a significant amount of rocky shoreline. Similar to Puerto Rico, coastal areas are relatively steep. Both the rocky shoreline and the steepness of the coastal areas limit the destructive effects of hurricanes, particularly by reducing the potential for inland penetration of storm waters. COMPARISON WITH THE SOUTH CAROLINA SHORELINE Hurricane Hugo presented the unique opportunity to study the impact of the same storm on two vastly different geologic shoreline types. The storm was essentially the same intensity when it passed over the Caribbean and when it made landfall in South Carolina. The differing geologic settings, however, produced drastically different shoreline responses to the hurricane-associated physical forces. The steep, narrow nature of the Puerto Rico and Virgin Island insular shelves is in stark contrast to the broad, gentle nature of the South Carolina continental shelf. In Puerto Rico, there is so little shelf water to be piled up into a storm surge that surges are inherently lower. The wide South Carolina continental shelf is essentially a broad, shallow “pan” of water that can be forced landward by storms, leading to inherently higher storm surges than in Puerto Rico from the same forcing storm. The Puerto Rico and Virgin Island shorelines were described briefly in the previous section. The South Carolina shoreline is entirely unconsolidated sand, and the slope of the coastal plain and shelf is very gentle. This means that the same amount of vertical rise in water level (the storm surge) leads to a much greater horizontal transport of the storm water overland. That is, the potential for flooding a much greater distance inland is inherent in the South Carolina setting compared with the Puerto Rico setting. Another consequence of inherently higher potential storm surge in South Carolina (compared with Puerto Rico) is that the ebb flow of subsiding flood water is greater. Because of this, dozens of houses were undermined in South Carolina as storm-surge return waters eroded their foundations. New inlets were cut, severing ties with roads and utilities. By contrast, in Puerto Rico the impact of storm waters, waves, and overwash was almost nil just a few blocks inland, except in the very flattest nearshore swampy areas. GENERAL OVERVIEW OF HUGO'S IMPACT Response of the shoreline to severe storms is a function of two things: 1) the type, intensity, and direction of the storm and 2) shoreline characteristics. These characteristics include geologic features (rocky or sandy shoreline, elevation, dune heights, etc.), the state of development (type and density of structures), and the type

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA and amount of shoreline engineering (seawalls, jetties, etc.). Development in the coastal zone puts property in danger, and the type of development and shoreline engineering can increase or decrease the amount of damage, depending on all the variables mentioned above. The northeastern coast of Puerto Rico was the most severely damaged coastal sector of the island. Observations of the coastal impact of Hugo were made from Dorado, west of San Juan, around to Cabo Mala Pascua, west of the southeastern corner of the island (Figure 6-1). The metropolitan San Juan area and outlying coastal sectors were observed (see Figure 6-2). From Luquillo to the west, the hurricane-force winds were blowing directly onshore and caused considerable damage from direct-wave attack as well as from surge and sand overwash. Fajardo, on the east coast, is near where Hugo made landfall and was one of the hardest hit areas. To the south, on the east coast and the eastern portions of the south coast, winds were largely offshore throughout the storm. While there was some damage from locally generated waves, storm surge was less, and waves were essentially being blown offshore, helping to lessen the damage. Erosion of the coast seemed localized to the sandy beach areas, though not all beaches eroded the same, and at least one (Taino Park in the Condado area of San Juan) actually accreted. Damage from direct-wave attack was usually limited to structures built on rocky headlands (for example, Punta las Marias and Punta El Medio in San Juan). Rocky points do not erode nearly as much as sandy beaches, and their elevation does afford structures some protection from minor storms. However, large storms like Hugo with giant waves and storm-wave swash, do impact structures on the lower-elevation portions of the headlands. FIGURE 6-1 Study area map showing eastern portion of Puerto Rico and highlighting locations visited during field work. Numbers in squares are estimates of surge heights during Hugo as described in text.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-2 Detailed map of San Juan metropolitan area. The area from Old San Juan to the east was the hardest-hit area observed. Almost all major developed portions of the shoreline were visited by the team. The following section details observations made from field visits. On the outlying major islands of Puerto Rico (Culebra and Vieques) and the U.S. Virgin Islands observed by the field study team, damage was very similar to those observed on the main island; that is, largely wind damage with water damage (wave and storm surge) restricted to very low penetration distances. STORM SURGE, PREDICTED AND OBSERVED Computer models developed by NOAA, called SLOSH models, are used to predict storm-surge levels and determine what parts of the coastline will be flooded, given the characteristics of the pending hurricane. Storm surges have been predicted for Puerto Rico using SLOSH models by Aurelio Mercado of the University of Puerto Rico at Mayaguez (Aurelio Mercado, UPR-Mayaguez, personal communication, 1989). Mercado also simulated surge after Hugo using the official storm track, a 1.1-ft high tide, and an assumed constant radius of maximum winds of 19 mi. The model suggests that a maximum surge of about 5 to 6 ft should have occurred in the Vieques Passage between the main island of Puerto Rico and the island of Vieques. The model also predicted that maximum surge in San Juan Harbor should have been about 1.5 ft, when in actuality it was measured at a tidal gauge to have been 3.5 ft. The predicted surge heights for the near shore were very close to those observed in places, but far from accurate in other places. The irregular

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA shoreline and irregular nearshore sea-bottom topography of Puerto Rico probably causes the SLOSH model to break down somewhat in this zone. Field estimates of storm-surge heights were made from observations of debris or rack lines, sand overwash, and other obvious signs that water had impacted the area. Figure 6-1 also shows some estimated surge heights. It is difficult to predict surge heights precisely, because surges are technically mean-water levels, and waves on top of surge increase the penetration of high water to even higher elevations. In addition, the SLOSH model does not account for waves. Luquillo was the only place other than the San Juan tide gage station where a reliable storm surge was measured. A videotape taken during Hugo by Jim Leonard recorded water levels in front of a condominium right on the shoreline (Leonard, 1989). In front of the condo is a distinctive rock formation (Figure 6-3). By comparing the normal water level on the rock formation with that during the storm, a reliable storm surge of about 8 ft can be measured. Storm waves on top of the surge were estimated to be at least 10 ft. The observed surge at Luquillo was about twice that predicted by the SLOSH model. In Fajardo, the surge was high enough to ground one of the large ferries that serve Vieques and Culebra (Figure 6-4). Many boats were destroyed in a supposedly “hurricane-proof” harbor of Ensenada Honda on Culebra. Figure 6-5 shows some of the boats that were floated inland and grounded by the storm surge. It is also evident in the photo that storm-surge waters did not penetrate very far inland in this steep setting. Figure 6-6 shows a similar example at the south shore of St. Croix. Figure 6-7 illustrates the situation on the northern shore of St. Croix, west of Christiansted. FIGURE 6-3 Luquillo shorefront showing seawall that had reportedly fallen during Hugo. Storm surge and wave heights near the time of passage of the eye were determined by observing water levels on the distinctive rock formations.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-4 Grounded ferry in Fajardo. FIGURE 6-5 Ensenada Honda, the “hurricane proof” harbor of Culebra, showing ships grounded by storm surge, notice that ships did not float very far inland—testimony to the small penetration of storm surge owing to the steep terrain. Courtesy: Richard Marshall, 1989.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-6 Boats washed ashore by storm on the southern shore of St. Croix. As in Figure 6-5, storm surge did not penetrate very far inland. Courtesy: Richard Marshall, 1989. FIGURE 6-7 St. Croix, northern shore, west of Christiansted. Reduced penetration of storm surge owing to steep terrain (see also Figure 6-5 and Figure 6-6). Courtesy: Richard Marshall, 1989.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA SAND OVERWASH The section of the coast from Boca de Cangrejos, near the Luis Munoz Marin International Airport, east to Punta Vacia Talega, in the area of Pinones, suffered extensive coastal erosion. Overwash covered Route 187 at the shorefront—the only road in this area—with up to 1 m of sand in places but did not penetrate very far inland, barely over the road in many places. The extent of shoreline affected by overwash was a stretch approximately 10 km long. Figure 6-8 shows sand overwash penetration and thickness for the Pinones area. Sand that was overwashed during the course of Hugo has been permanently removed from the beach system. In an undeveloped setting, this sand would be available to be blown into dunes and to add elevation to the coastal lowlands. On developed shorelines, however, the overwash sand is often regarded as “in the way” and quickly cleared off roads, used for construction, or dumped in empty lots. It is important, especially on shorelines prone to erosion, that overwash sand be returned to the beach system as a means of beach reconstitution. To the credit of the PRDNR, heavy equipment was in the Pinones area moving some of the overwashed sand back to the beach within 1 week of the storm. This should continue throughout the affected areas of Puerto Rico. Figure 6-9 shows one of many piles of overwash sand in the Pinones area. FIGURE 6-8 Overwash sand penetration and relative thickness for Pinones area.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-9 Overwash sand piled up along Route 187 in Pinones. Escambrón recreation area had extensive overwash, though for only a few hundred meters of shoreline. Major overwash occurred at Barbosa Park, just west of Punta las Marias, as well (see Figure 6-2). Other areas did not experience significant sand overwash. Some was noted in the Cabezas de San Juan area at the northeastern corner of the island. Moderate amounts of overwash were observed in Farjardo and along the east coast. Figure 6-10 shows modest amounts of sand overwash on the east side of St. Croix. This is similar to that encountered on the other islands studied. FIGURE 6-10 Sand overwash on the eastern shore of St. Croix. Courtesy: Richard Marshall, 1989.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA COASTAL FLOODING Coastal flooding was not a major problem during Hugo. Coastal flooding must be distinguished from river flooding. The former is simply the result of rising storm-surge waters and storm-wave swash inundating lowland areas. Fields and Jordan (1972) made a series of maps showing areas of major flooding and the effect of storm-wave swash on the north coast of Puerto Rico. The surge associated with Hugo was sufficiently low to limit coastal flooding in Puerto Rico. WAVE ATTACK Buildings were damaged by direct-wave attack in many places around the island, but the damage seemed to be localized. The metropolitan San Juan area was exceedingly hard hit. The major rocky headlands of Punta Escambrón, Punta Piedrita, Punta las Marias, Punta El Medio, Punta Cangrejos, and Punta Maldonado received extensive damage. Several walls were undermined, sidewalks collapsed, and minor structural damage was caused by heavy waves associated with Hugo. (See Figure 6-11, Figure 6-13, Figure 6-14, Figure 6-15, Figure 6-16, and Figure 6-17.) In the Condado, the Oasis Restaurant's front wall was destroyed (Figure 6-11). Figure 6-12 demonstrates the power of storm waves. The boulder in the photo was lifted from the beach and over a 1-m wall into this park, which is immediately west of the Oasis Restaurant. Shoreline retreat during the storm, in addition to collapsing walls, exposed tree roots, as shown in Figure 6-13 and Figure 6-14 from Punta las Marias. Shoreline erosion also exhumed infrastructure, as shown in Figure 6-15 at Punta el Medio in Isla Verde. The shoreline at Punta Uvero east of Loiza was heavily armored with a boulder revetment, but only in front of some of the houses. Over the years, the shoreline has continued to erode past the revetted section. Hugo destroyed the boulder revetments and most of the first row of structures (Figure 6-16). A mobile home park on the beach in Loiza was destroyed, mostly by wind force, though much wave damage to the dunes in the area was noted. This is one of a very few mobile home parks known in Puerto Rico. USGS and PRDNR have been carrying out a beach-profile monitoring program for several years in the San Juan metropolitan area. They use, among other things, stone walls and sides of buildings for their beginning benchmarks (the points from which they run their profile lines seaward). Several of the benchmarks were destroyed in the storm, making correlation of pre- and post-storm data difficult. As of 1991, several of the benchmarks were re-established. Comparison of beach profiles shows little or no long-term effect on beach shape resulting from Hugo.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-17 Balneario Isla Verde shoreline showing tree stumps on beach and toppled light poles. There was little shoreline damage south of the Roosevelt Roads Naval Station. Some minor overwash on the east coast and removal of sand at the Caribe Playa Sunbeach Resort near Cabo Mala Pascua on the south coast was noted. Wave damage on the off-shore islands was similar to that observed on the main island of Puerto Rico. DEGRADATION OF RECREATIONAL BEACH RESOURCES A major resource for Puerto Rico is its beautiful beaches. These are important not only to the tourism industry but also to the local population, which swarms to the beaches during the calm, summer months. The public swimming beaches of Isla Verde and Luquillo were severely damaged. Luquillo's beach is not typically prone to erosion. Isla Verde, however, is an erosive beach; it was exhibiting such effects even before Hugo, and a 1-m scarp is maintained there year round when storms are frequent. Trees were fallen, light poles that had been at the edge of the erosion scarp in January 1988 were toppled, and the scarp itself was eroded back about 15 ft during Hugo, undermining more of the parking lot (Figure 6-17). The Barbosa Park beach in Ocean Park is a very popular one in the San Juan area. The shoreline here retreated about 10 m during Hugo. Sidewalks were undermined and utility poles were toppled (Figure 6-18). This was also a site of extensive sand overwash. The beach in front of a gabion wall at Playa de Humacao on the east coast disappeared during Hugo (Figure 6-19). The wall was built after Hurricane David in 1979 to protect mostly single-story homes and businesses. The structures were largely

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-18 Barbosa Park shoreline after Hugo. FIGURE 6-19 The gabion wall at Playa de Humacao. The beach is gone, and waves will now begin to attack the wall directly. Some beach rebuilding is expected naturally, though how much is impossible to predict.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA spared, but the beach is now gone. The beach will build out again, but that is only temporary. Storms will eventually remove all the sand in front of the wall and waves will attack it with increasing power in the future without any beach for protection. This is an excellent example of the all-too-common situation of a wall being built to protect structures behind it, but leading to degradation of the recreational beach. The privately owned Caribe Playa Sunbeach Resort, near Cabo Mala Pascua on the south coast, historically suffers from beach erosion during tropical cyclones, and Hugo was no exception. The owners replenish the beach as necessary with sand from other parts of their property to provide an aesthetically pleasing shorefront. Figure 6-20 shows downed palm trees, as well as a seawall and gravel beach (normally covered) that were exhumed during Hugo. New sand already emplaced only 3 days after Hugo can be seen in the right-hand side of the photo. Interestingly, some stretches of sandy beaches (i.e., Taino Park, east of Punta Piedrita) actually accreted during Hugo. “SETTING UP” THE SHORELINE FOR WINTER STORM DAMAGE Perhaps the biggest problem from a coastal erosion standpoint is that, even though the beaches will rebuild somewhat naturally, it will not occur until summer, when long-period waves gently push sand landward. Winter storm waves will thus attack an already degraded shoreline. This is called “setting up” the shoreline for FIGURE 6-20 Shorefront of Caribe Playa Sunbeach Resort near Cabo Mala Pascua. Seawall and gravel beach are kept covered by owners ' efforts of replenishing the beach with sand taken from elsewhere on their property.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA more intense damage from subsequent storms, because much of the natural protection offered by the beach and dune systems has been removed, putting developments in and behind this zone in even greater danger. The winter storm season is the time of the largest wave activity on the northern coast of Puerto Rico. Far-travelled swells from North Atlantic storms (the so-called northeasters) traverse the ocean and impinge violently upon the shoreline of Puerto Rico. SOME UNIQUE UNDERSEA WATER DATA NEAR ST. CROIX An unusual set of observations were made off the north shore of St. Croix near Christiansted. The National Oceanic and Atmospheric Administration (NOAA) sponsored an underwater research habitat 60 ft below the surface on a sand plain of the Salt River Submarine Canyon (Kalvaitis, 1989). The work is performed by NOAA's National Undersea Research Center at Farleigh Dickinson University (NURC-FDU). Prior to the passage of Hurricane Hugo, personnel from NURC-FDU deployed a S4 current meter, manufactured by InterOcean Systems, Inc., to collect data on the storm passage (Taylor and Trageser, 1990). The S4 current meter is housed in a 25-cm diameter sphere. Water flows through the electromagnetic field created by the sphere generating a voltage proportional to the water velocity. The meter does not have any vanes or propellers. Because of limited data storage capacity, the meter recorded 1 sec averages of the .5 sec sample rate for 18 min every 2 hours from 1633 AST on September 16 to 0851 AST on September 19. This provided 33 bursts of data from the time Hurricane Hugo was about 400 miles east-southeast of St. Croix to about 400 miles northwest of St. Croix. Current velocity, current direction, depth, and conductivity were collected during this period. The NOAA habitat and the S4 current meter are about .25 miles (.4 km) north of the mouth of the Salt River (Figure 6-21). The current meter was situated about 12 ft from the base of the west canyon wall under 60 ft of water, and the habitat is located in about 55 ft of water. Figure 6-22a presents the speed and direction of the current, and the depth of the water over the meter from burst 17, 18, and 19, which began at 2233 AST on September 17, and ended at 0251 AST on September 18 (Taylor and Trageser, 1990). The eye of Hurricane Hugo was south of St. Croix and the low-level winds were coming from the north or north-northeast in the vicinity of the Salt River Canyon. The current speeds under approximately 60 ft of water in bursts 17 and 18, the first two-thirds of Figure 6-22b, were all less than 100 cm/sec, except for an occasional spike. The last one-third of the figure, burst 19, shows a dramatic increase in the current speeds (a maximum of 351 cm/sec). The average current speed was 52.09 cm/sec for these three bursts. In contrast, the average current speed of bursts 2, 3, and 4—measured between 1633 and 2051 AST on September 16—was 5.67 cm/sec. Changes in depth from the crest to the base of the

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-21 NOAA habitat and the S4 current meter approximately .4 km north of the mouth of the Salt River. Note: Habitat was since moved in 1990 and redeployed in the upper Florida Keys Marine Sanctuary in 1993. Courtesy: NOAA/NURP, 1990. wave during this period exceeded 3 m on a number of occasions. For bursts 2, 3, and 4, difference in the depth from the crest to the base were all less than .8 m. The currents in the vicinity of the habitat were sufficient to wash away the underlying sand and cause the habitat to shift from level. Figure 6-23 shows some of the shifts in the habitat caused by the currents. As much as 6 ft of sand was washed away from under some of the pods that provide the base for the habitat. SUMMARY The Duke University Department of Geology's Program for the Study of Developed Shorelines has been studying the response of developed shorelines to storms for several years. A variety of shoreline settings have been studied, ranging from the East Coast of the United States to the Yucatan Peninsula after Hurricane Gilbert in 1988, to Puerto Rico and South Carolina after Hurricane Hugo. Observations have been made before and after winter storms as well as hurricanes. Observations have produced a list of “lessons learned” that are applicable across the range of shoreline and storm types. The list of lessons learned first appeared in Bush et al. (1988) and has been published in Thieler et al. (1989). There are few surprises learned from Hugo. The response of developed shorelines is already well known, as are many of the “safe” practices for living with

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-22a S4 current meter showing speed and direction of current. Courtesy: Taylor and Trageser, 1990. FIGURE 6-22b Depth of water overcurrent meter for bursts 17, 18, and 19. Courtesy: Taylor and Trageser, 1990.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 6-23 Pod on NOAA Hydrolab showing the sand scouring. the damage potential of coastal hazards. Hurricane Hugo illustrated many of the poor development practices that frequently increase the potential for damage from coastal storms. Hopefully, the following list of lessons learned will lead to better coastal zone management policies in Puerto Rico: Storms cause shorelines to retreat. Shoreline migration is a natural geologic process caused by a sea level rising over a sloping land surface. Storms are one of the driving forces that move beaches landward. Wide beaches protect. The more beach there is to absorb and dissipate storm-wave energy, the better the possibilities for mitigating damage to structures. This helps in understanding why structures situated low on rocky shorelines, which typically have little or no sandy beach in front of them, sustain so much damage even though they are on a relatively nonerosive shoreline. 3. Dunes protect. Where dunes, rather than buildings, are present to absorb the impact of waves and storm surge, post-storm beaches are markedly wider. In addition, structures located behind the frontal line of dunes are usually not damaged. Forests Protect. Overwash penetration and storm damage is typically an order of magnitude greater where coastal forest is removed for development. This is illustrated well in the Pinones area, where overwash sand penetrated much farther inland at locations that had been denuded of trees in order to make way for roads and houses.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA Rocky shorelines protect. Rocky shorelines erode slowly and add may elevation, thus protecting structures located on them. No better example exists than the longevity of Old San Juan, situated high atop the rocky shoreline. Building setbacks protect. Structures built farther from the water stand a better chance of not being damaged. Shore-perpendicular roads become overwash passes. Elevating and curving the roads so that they approach the beach at an angle would reduce the extent and amount of overwash. Simply putting a hump in the road at its beach terminus would probably help dramatically. Roadbed material should not be obtained from beach or dunes. This was not much of a problem during Hugo in Puerto Rico because the surge was low. Notches in dunes create overwash passes. Without exception, notches cut in dunes for beach access, view, or development are naturally exploited by waves and storm surge. Storm-surge ebb is intensified when funneled by structure development. This is more a problem in the South Carolina-type setting, where surges were higher than in Puerto Rico. Seawalls and storm rubble cause beach narrowing. Beaches in front of hard structures are invariably narrower than natural sandy beaches. Storm response is affected by pre-storm shoreline engineering. Pre-storm shoreline-stabilization structures have dramatic impacts on storm response, some beneficial (such as wide beaches on the upstream side of jetties), but most detrimental. Studies through the years have important implications for the large-scale decisions society will be making concerning response to rising sea levels. Trying to fight an encroaching sea with hard shoreline stabilization is not only expensive; but sacrifices beaches for buildings and is also doomed to long-term failure. For example, the gabion wall at Playa de Humacao will eventually lead to loss of the beach. In addition, several seawalled stretches of shoreline in the metropolitan San Juan area (and indeed, around the entire island) already have no recreational beach at all. Beach replenishment saves the beach but is very expensive. Cost-benefit studies need to be done so that informed decisions can be made by appropriate authorities. A large-scale management approach to rising sea level that is gaining popularity is retreat or relocation. Prohibition of the rebuilding in place of destroyed shorefront buildings is one method of relocation. Damage to the first row of buildings is almost always more severe than damage to subsequent rows. If rebuilding of totally destroyed structures was prohibited, fewer structures would be in the highest-risk zone. A number of years of relatively low-economic-risk development would be provided while shoreline retreat gradually narrows the distance between the beach and the first row of development. Observations over the years by the Duke University program have shown that in Puerto Rico, response to shoreline erosion has largely been on a crisis-response basis. Individual structures or small segments of shoreline have been walled,

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA seemingly without regard to the effect it would have on neighboring properties. The Puerto Rico Planning Board oversees shoreline construction islandwide. They understand that a more controlled, unified shoreline-management program is needed, incorporating sound practices for new development, relocation where possible of structures now in danger, beach replenishment for tourist beaches, and even abandonment of some shoreline development in places. RECOMMENDATIONS In January 1988, Orrin H. Pilkey, Jr., director of the Duke University program, accompanied David Bush on a trip around the island to determine the “State of the Shoreline” of Puerto Rico. Since then, aerial videos have been shot to record the shoreline situation, and studies of historical erosion and the effect of seawalls on beach width have been done. The following list of recommendations is derived from those studies and additional information from observations of the impact of Hugo. Initiate a long-term study of shoreline erosion in Puerto Rico. Such a study would assess the islandwide and community-by-community erosion situation, start a continuous beach-profiling program, make islandwide and communitywide recommendations of shoreline-management alternatives, and begin planning for the sea-level rise. The USGS Marine Geology Office in San Juan, cooperatively with the PRDNR, has an ongoing study of the beaches of the San Juan metropolitan area that has been under way for several years. That work should continue and expand to include the rest of the island. Quite a bit of baseline-type research on shoreline setting and state of the shoreline of Puerto Rico has already been done by the Duke University program. Prevent the mining of sand from Puerto Rico's beaches. The PRDNR should be assisted in clamping down on illegal mining and should issue no more permits for “legal” taking of sand from beaches or rivers. Control hard stabilization of the shoreline, perhaps through a permitting process. To support this initiative, beach “surveillance,” including monitoring of all shoreline engineering projects through aerial photography and beach profiling, is needed. Again, the USGS and PRDNR should undertake such studies. Evaluate alternatives to hard stabilization. Options include development setbacks, relocation or demolition of low-cost shorefront buildings, and beach replenishment. The problem is that building of walls and revetments is usually done on a crisis basis, allowing no time to consider alternatives. At least two communities, Luquillo and San Juan (Isla Verde west to the Condado), could benefit immediately from beach replenishment and could probably build a case for federal financial participation. A concerted effort is needed to halt beach degradation in Puerto Rico. Many miles of Puerto Rico's recreational beaches have been seriously degraded and even destroyed by seawalls built to protect buildings. Preservation of recreational beaches

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA should be given high priority, in many cases higher priority than preservation of shorefront buildings. Continue research on improved storm-surge models. Reliable surge predictions will allow an inventory of property at risk to be made. The result will be a scientific basis for coastal-zone management and shoreline building restrictions. As part of the research, more tide gage stations should be installed around the island. Form a “Beach Watchdog” committee, with representatives from the USGS Marine Geology Office, the PRDNR, the Puerto Rico Planning Board, and other public and private groups (i.e., the Tourism Company, Hotel Owners Association, environmental groups, etc.). This group should oversee a unified, scientific-based, coastal-zone management policy that puts preservation of beaches ahead of preservation of buildings. These recommendations apply equally to the Virgin Islands.

OCR for page 130
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA REFERENCES Bush, D. M., E. R. Thieler, and O. H. Pilkey, Jr. 1988. Impact of Hurricane Gilbert on the Shoreline of the Yucatan Peninsula, Mexico. Report to the National Academy of Sciences. Fields, F. K., and D. G. Jordan. 1972. Storm-Wave Swash Along the North Coast of Puerto Rico. United States Geological Survey Hydrologic Investigations, Atlas HA-430, 2 sheets. A. Kalvaitis, 1989. Personal communication to Dr. Joseph H. Golden. Leonard, J. 1989. Video of Hurricane Hugo, Luquillo, Puerto Rico. September 18, 1989. G. Taylor, and J. Trageser. 1990. Directional wave and current measurements during Hurricane Hugo. Pp. 118-140 in Proceedings of Marine Instrumentation 1990 Conference, February 27-March 1, 1990. Spring Valley, California: West Star Productions. Thieler, E. R., D. M. Bush, and O. H. Pilkey, Jr. 1989. Shoreline response to Hurricane Gilbert: Lessons for coastal management . Pp. 765-775 in Coastal Zone '89, Proceedings of the Sixth Symposium on Coastal and Ocean Management . New York, New York: American Society of Civil Engineers.