2

Hydrology

John L. Vogel, National Weather Service, National Oceanic and Atmospheric Administration

INTRODUCTION

Puerto Rico and the U.S. Virgin Islands are embedded in a warm tropical air mass with only occasional excursions of cold air from cold fronts invading the island during the winter months. Precipitation is maximized during the summer months as a result of thunderstorms, easterly waves, and an occasional tropical storm or hurricane. The maximum rainfall amounts for durations of 6 through 48 hours are primarily associated with hurricanes and tropical storms, and on some occasions with cold fronts and easterly waves (U.S. Weather Bureau, 1961). The maximum 24-hour rainfall is 24 inches on the north side of the mountains of Puerto Rico associated with the San Ciriaco Hurricane of August 1899. For durations of 6 hours or less, localized thunderstorms are responsible for the most intense rainfall events.

Hurricanes and tropical storms have passed within 100 NM (190 km) of Puerto Rico or the U.S. Virgin Islands 91 times from 1871 through 1985, and have passed over Puerto Rico 26 times, according to the tropical storm tracks generated by Neumann et al. (1985). From 1985 through 1988 no hurricane or tropical storm was observed within 2 degrees of the Virgin Islands or Puerto Rico (Goodman, 1988). From 1871 through 1989 a total of 93 hurricanes or tropical storms passed within 100 NM, and 27 moved across Puerto Rico and the Virgin Islands, including both Hugo and Inez in 1989. Goodman also points out that the frequency of hurricanes or tropical storms that passed within 2 degrees of Puerto Rico or the Virgin Islands dropped from 2.1 per 5-year period from 1886 through 1935 to 1.0 per 5-year period from 1936 through 1985. This is likely due to a short-term climatological shift in the tracks of hurricanes, which could shift again. At this time, it is not known which 50-year period—from 1886 to 1935 or from 1936 to 1985—represents a more “normal” period.

Climatologically, the wet season in Puerto Rico extends from May through January, when monthly precipitation ranges from 4.5 to almost 7.0 inches. May, July, August, and September are generally the rainiest months. September has the most occurrences of maximum 24-hour rain amounts (U.S. Weather Bureau, 1961), and



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 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA 2 Hydrology John L. Vogel, National Weather Service, National Oceanic and Atmospheric Administration INTRODUCTION Puerto Rico and the U.S. Virgin Islands are embedded in a warm tropical air mass with only occasional excursions of cold air from cold fronts invading the island during the winter months. Precipitation is maximized during the summer months as a result of thunderstorms, easterly waves, and an occasional tropical storm or hurricane. The maximum rainfall amounts for durations of 6 through 48 hours are primarily associated with hurricanes and tropical storms, and on some occasions with cold fronts and easterly waves (U.S. Weather Bureau, 1961). The maximum 24-hour rainfall is 24 inches on the north side of the mountains of Puerto Rico associated with the San Ciriaco Hurricane of August 1899. For durations of 6 hours or less, localized thunderstorms are responsible for the most intense rainfall events. Hurricanes and tropical storms have passed within 100 NM (190 km) of Puerto Rico or the U.S. Virgin Islands 91 times from 1871 through 1985, and have passed over Puerto Rico 26 times, according to the tropical storm tracks generated by Neumann et al. (1985). From 1985 through 1988 no hurricane or tropical storm was observed within 2 degrees of the Virgin Islands or Puerto Rico (Goodman, 1988). From 1871 through 1989 a total of 93 hurricanes or tropical storms passed within 100 NM, and 27 moved across Puerto Rico and the Virgin Islands, including both Hugo and Inez in 1989. Goodman also points out that the frequency of hurricanes or tropical storms that passed within 2 degrees of Puerto Rico or the Virgin Islands dropped from 2.1 per 5-year period from 1886 through 1935 to 1.0 per 5-year period from 1936 through 1985. This is likely due to a short-term climatological shift in the tracks of hurricanes, which could shift again. At this time, it is not known which 50-year period—from 1886 to 1935 or from 1936 to 1985—represents a more “normal” period. Climatologically, the wet season in Puerto Rico extends from May through January, when monthly precipitation ranges from 4.5 to almost 7.0 inches. May, July, August, and September are generally the rainiest months. September has the most occurrences of maximum 24-hour rain amounts (U.S. Weather Bureau, 1961), and

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA is the month when most hurricanes and tropical storms pass within 100 NM of Puerto Rico and the U.S. Virgin Islands. Rivers in Puerto Rico and the Virgin Islands are driven by geography, length, and climate. Most originate on either the north or south slopes of the ridge that moves mostly east to west across the islands. Figure 2-1 shows the relation between topography and some of the rivers in Puerto Rico. The spine along the center of the island is anchored on the northeast by El Yunque (1,065 m), and in the central part of the island by Cerro de Punta (1,338 m), the highest point on the island. Between these two peaks, the terrain lowers into a saddle. During high flow periods induced by heavy rains, the rivers flow from the source to the seas in 12 hours or less. The rivers on the Virgin Islands are similar. Flooding in Puerto Rico and the Virgin Islands is usually localized, and only during hurricanes, tropical disturbances, or rare cold frontal events, which cause widespread intense rains across the islands, is there a chance for flooding of all the major streams. PRECIPITATION Radar observations from the NWS 10-cm WSR-74S at San Juan, Puerto Rico, provide a general overview of the sequence of precipitation events over Puerto Rico and the Virgin Islands. For most of the Virgin Islands, it is the only record of the precipitation events from Hurricane Hugo, since only a few rain gages remained FIGURE 2-1 Rivers on Puerto Rico and relation to topography.

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA available for measurements. This situation is attributable to the extreme winds that blew many rain gages over or damaged them by creating flying debris. Moreover, many observers in the Virgin Islands had to abandon their observation posts before Hurricane Hugo arrived. Radar was the primary tool used to examine the precipitation over the Virgin Islands. The data were preserved on film. Clock times were available on the film until 0626 AST on September 18; after that, only frame counts were available until 1946 AST on September 18. During this period, times are only approximate. Eye positions, the onset of precipitation, and other information were used to time the radar. St. Croix By 1615 AST on September 17 some scattered showers and thunderstorms were observed about 250 km in advance of Hugo, organized in an outer rainband over Puerto Rico; the rainshield associated with the hurricane eye was just beginning to move into radar view approximately 50 km east of St. Croix. The leading edge of the rainshield moved from east-southeast at about 24 km per hour, and at 1945 AST it was situated just north of the island of St. Croix. According to radar, the precipitation did not begin over the island until 2030 AST. However, the radar beam from San Juan was probably unable to discern low-level or stratiform precipitation over St. Croix because, at that range, it would be unable to see the lower 3,000 to 5,000 ft of the atmosphere. The leading edge of the rainband did not completely cover St. Croix until 2145 AST, so it took about 2 hours to encompass the island, a distance of less than 20 km. Between 2145 and 2215 AST on September 17, new eyewall convection began to propagate or redevelop from about 90 km south through about 50 km west-southwest of St. Croix. Figure 1-14a, Figure 1-14b, Figure 1-14c, Figure 1-14d and Figure 1-14e show this development. At 2109 AST there were only some light rainfall intensities visible to the south of the heavy rainband 120 to 160 km southeast of San Juan. By 2233 AST more intense rainfall was being observed in the same area relative to the rainband. This convective activity continued to grow through the night and continued to have about the same position relative to the eye of Hurricane Hugo through 0400 AST on September 18. During this period, intense precipitation continued over St. Croix. The forward progress of the hurricane appeared to slow between 0400 and 0500 AST southwest of St. Croix and then accelerated in the next hour. By 0555 AST (Figure 1-14d) the intense precipitation passed over St. Croix as the eye's center passed west and moved northwest of St. Croix. The radar echoes indicate that little or no precipitation was associated with large convective clouds or rainbands to the rear of Hugo. As Hugo moved northwest toward Puerto Rico, it developed a radar pattern similar to that of Hurricane Frederic as it moved onto the shores of Louisiana and Mississippi in September 1979 (Parrish et al., 1982). The lack of forward progress of Hugo between 0400 and 0500 AST could be associated with the deep convective activity to the south and southwest

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA or unobserved changes in the larger-scale steering flow. Willoughby et al. (1984) speculated that the mesoscale convection south and southwest of Hurricane David in 1979 may have slowed its advance. There may be some similarity between these two hurricanes. However, Hurricane David had an extensive rainband structure in all four quadrants. The radar images of Hurricane Hugo do not show such an extensive rainband structure, especially in the rear quadrants. Only two rain gage reports are available from the NWS network on St. Croix: Ham Bluff Light House Station on the northwest coast, which recorded 11.2 inches (284 mm) for the 3-day period from 0700 AST on September 16 to 0700 AST on September 19; and Fountain in the northwest, which recorded 9.2 inches (234 mm). At Ham Bluff, climate records for September prior to Hurricane Hugo were lost, and the rain gage at Fountain could not be read before the morning of September 19 because downed trees made the road impassable. Considering the high winds experienced over St. Croix, it is likely that the rainfall totals at Ham Bluff and Fountain are conservative estimates of the total precipitation. With peak wind gusts in excess of 100 knots (115 mph), between 30 and 50 percent more rain may have been blown over the top of the raingage. This means that another 2.75 to 3.40 inches might have fallen, if one assumes an undercatch of 30 percent, or a total estimated true rainfall of 12 to 15 inches for September 17 and 18. The radar analysis indicated that most of the rain fell in a 12-hour period from 2000 AST on September 17 through 0800 AST on September 18. If one assumes that 75 percent of the total rains fell during this period, then the 12-hour rainfall totals would be between 9 and 11.25 inches. This is equivalent to a return frequency of more than 100 years for a 12-hour period (U.S. Weather Bureau, 1961). St. John and St. Thomas Hurricane Hugo's first major rainband moved across the British Virgin Islands about 1830 AST on September 17 and over St. John and St. Thomas between 1845 and 2000 AST. The large rainshield moved onto St. John by 2100 AST and covered St. Thomas by 2245 AST. Initially, the showers and thunderstorms associated with the rain mass were moving from the northwest. By 0100 AST on September 18 the storm cells were moving from the east-northeast, and shifted from the east by 0500 AST. This was in response to the movement of Hurricane Hugo as it passed near the southwest coast of St. Croix and moved southwest of St. John and St. Thomas. As in the rainfall over St. Croix, the heavy precipitation remained over the islands for about 12 to 13 hours. Heavy rains were observed over St. John from 2100 AST on September 17 through about 0900 AST on September 18; on St. Thomas the rains were observed from about 2230 AST on September 17 through about 0930 AST on September 18. Rainfall observations were available from only five locations over St. John. The best estimates for daily rainfall values at these stations are shown in Table 2-1. Some of the measurements were accumulations for the 3 days, and

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA TABLE 2-1 Daily Estimate of Precipitation (in inches) Over St. John and St. Thomas   9/17 9/18 9/19 Total St. John Caneel Bay Plantation 0.00 8.50 0.58 9.08 Catherineburg 0.27 4.90 0.45 5.62 Coral Bay 3.00 6.60 0.06 9.66 East End 0.02 6.24* 1.11* 7.37 Cruz Bay 0.36* 6.46* 0.59* 7.41 St. Thomas Water Isle 0.02 5.24 0.15 5.41 *estimated daily amount estimates of the daily totals were made from other daily rain gages and the recording rain gage at Caneel Bay that functioned throughout Hurricane Hugo. The greatest 3-day rainfall total recorded by daily gages was 9.69 inches (246 mm) at Coral Bay on the eastern part of St. John. The total rains observed at the other stations on St. John ranged from 5.62 inches (143 mm) at Catherineburg to 9.08 inches (231 mm) at Caneel Bay Plantation. In addition to a daily rain gage, Caneel Bay had a recording rain gage; its record is shown in Figure 2-2. The record at this rain gage confirms the timing of the heavy rains over St. John, from 2000 AST on September 17 through 0900 AST on September 18. During this 13-hour period, 9.1 inches (231 mm) of rain was recorded. The maximum 1-, 2-, 3-, 6-, and 12-hour rains, their amounts, and their return frequencies from the U.S. Weather Bureau (1961) are given in Table 2-2. The maximum clock-hour rain of 2.3 inches (58 mm) represents about a 5-year return. However, as the duration of the maximum rainfall increased to 6 hours, the return frequency was greater than 100 years. Again, it is likely that the rain gage measurements are an underestimate by at least 30 percent of the total rains that fell, because of the strong winds that can deflect the rains across the top of the rain gage. Thus, the maximum 12-hour total could have been as high as 11.7 inches (297 mm). The 11.7 inches (297 mm) in 12 hours would exceed the 100-year return value by 2.9 inches (74 mm). The 12.5 inches (318 mm) in 24 hours would exceed the 100-year return for St. John by 2 inches (51 mm). Only one rain gage observation was available on St. Thomas: Water Isle (Table 2-1), which measured a 3-day total of 5.37 inches and a 1-day maximum of 5.20 inches. Considering the evidence from radar, it is believed that considerably larger

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 2-2 Mass curve at Caneel Bay Plantation, September 17 to 19. TABLE 2-2 Duration, Amount, and Return Frequency of Precipitation at Caneel Bay Duration Frequency (hours) Amount (inches) Return-(years) 1 2.3 5 2 4.1 50 3 4.7 50 6 7.4 > 100 12 9.0 > 100

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA amounts of precipitation fell over St. Thomas but were not measured. Rainfall totals similar to those observed on St. John can be expected at St. Thomas. Puerto Rico A few showers and thunderstorms in an outer rainband from Hurricane Hugo moved across Puerto Rico, Vieques, and Culebra from 1700 to 2300 AST on September 17. The major rainshield moved onto the eastern edges of Culebra and Vieques about 2300 AST and onto Puerto Rico about midnight on September 17. At the time that the rainshield moved onto Puerto Rico, the eye of Hurricane Hugo was about 230 km southeast. Radar shows the leading edge of the rainshield moving very slowly to the west across the island. In addition, the shield was apparently retarded from moving south by the main east-west spine of the island. The winds at this time were from the northeast and were blowing against the east-west mountain spine. The main shield, according to radar, did not reach the extreme western part of Puerto Rico until about 1300 AST, and then did not cover the southwest tip of the island. The radar data are corroborated by data from recording rain gages on Puerto Rico. Figure 2-3 gives the timing of the beginning of the intense rains across Puerto Rico and confirms the timing of the radar observations. The rainshield moved eastward at about 20 mph from midnight September 17 to 0200 AST September 18 over northern Puerto Rico, but after the first hour the forward movement was retarded in the southern parts. This much slower advance over the southern parts of Puerto Rico continued throughout the morning and early afternoon hours; the major part of the rainshield never covered the southwest tip of Puerto Rico, and the northwest tip of the island barely came under the influence of the main rainshield. Throughout the radar observations of Hurricane Hugo, the rainshield configuration relative to the eye generally remained constant. As a result, the western edge of the rainshield continued a more northerly track and did not move to the western parts of Puerto Rico until the center of Hugo moved near the northeast tip. Throughout the time that Hurricane Hugo moved from St. Croix to north of Puerto Rico, there was little evidence of convective precipitation in the two rear quadrants of the hurricane, and the recording rain gage data on Puerto Rico and St. John verified this observation. Once the eye of Hugo move past the islands, little or no additional precipitation was observed. As on St. John, St. Thomas, and St. Croix, rain gage data from daily and hourly recording stations on Puerto Rico were missing or data were lost because the severe winds blew the stations over or flying debris damaged them. However, since the winds were not as strong over Puerto Rico, and most of the severe winds were confined to the adjacent islands, an ample amount of precipitation data is available. The closest point of approach of the hurricane was the northeast tip of the island, and this is where the most rain fell from 0700 AST on September 17 to 0700 AST on September 19, aided by the orographic effects of El Yunque mountain (Figure 2-4). The recording rain gages indicated that most of the rains fell from about midnight

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 2-3 Timing of beginning of intense rains across Puerto Rico on September 18. FIGURE 2-4 Hurricane Hugo rainfall (inches).

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA September 17 to about 1600 or 1700 AST on September 18. There were some scattered showers and thunderstorms as one of the outer rainbands moved across Puerto Rico on September 17, and some isolated rain showers and thunderstorms moving from the southwest as part of a trailing outer rainband of Hurricane Hugo from 1700 AST on September 18 through September 19. The maximum recorded rainfall for the 48-hour period beginning on 0700 AST September 17 was 13.25 inches (337 mm) at Rio Blanco Lower on the south side of El Yunque. A secondary maximum, which was collected on a roof at the official rain gage, was observed at the San Juan Airport. Such an exposure makes a poor rainfall measurement point (Vogel, 1988). However, the actual rainfall catch at San Juan Airport was corrected by local NWS officials to compensate for the effects of the winds. From this secondary maximum a general maximum runs west with a tight gradient of rainfall amounts along the southern slopes of mountains that run east-west across central Puerto Rico. The rainfall totals along the southern, western, and northwestern coasts are at a minimum. This agrees with the radar observations, which show that the main rainshield did not penetrate south of the northern slopes of Puerto Rico or to the west until Hugo was already moving well north of Puerto Rico. The heaviest rainfalls were generally observed from 0000 to 1200 AST on September 18 over the northeast part of Puerto Rico, and the accumulated rainfall for two recording rain gages—Cubuy and Pena Pobre-Naguabo—are shown in Figure 2-5. The two stations are very close and show little difference in the shape of the curves. The rains began about an hour earlier at Pena Pobre-Naguabo and remained intense through 1200 AST at both stations. As shown in Table 2-3, the maximum 1-hour rains have return frequencies of less than 1 year. The return frequencies increase to 25 to 27 years with increasing duration through 12 hours. However, both of these stations were on the periphery of the main maximum rainfall. In the vicinity of the precipitation center, which is estimated to be in excess of 14 inches (356 mm), approximately 95 percent of the precipitation fell in 12 hours over the 2-day period from 0700 AST on September 17 to 0700 AST on September 19. This means that about 13.1 inches (333 mm) of rain fell in 12 hours at the maximum, which is more than the 100-year return frequency for 12 hours (U.S. Weather Bureau, 1961). For each of the recording rain gages, the percentage of rainfall during the maximum 12-hour period was obtained. This map was then applied to the total 2-day rainfall to determine the return frequencies of the 12-hour rainfall. Figure 2-6 depicts the spatial analysis of the return frequencies for 12-hour durations from Hurricane Hugo. Return frequencies of 2 years or greater were observed in the northwest corner of the island, peaking at greater than 100 years in the vicinity of El Yunque. A secondary maximum, with return frequencies greater than 25 years, was observed over eastern portions of Greater San Juan. The peak wind speeds over Puerto Rico were not as high as those over the Virgin Islands; however, it is estimated that at least 30 percent of the rain catch was missed, especially in the northwest quadrant of the island. This means that for the 2-day period, rains of up to 18 inches (457 mm) could have been concentrated over El Yunque, with perhaps more than 16 inches (406 mm) occurring in 12 hours. Thus,

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA FIGURE 2-5 Mass curves for Cubuy and Pena Pobre-Naguabo. an even larger area of the island may have experienced rainfall greater than the 100-year return value for a 12-hour duration. FLOODING Flooding was limited. In the Virgin Islands, there was flash flooding, and sheet flow must have existed over many of the roads because of the intense rainfall for at least 12 hours over the area. However, no major flooding was noted. Over Puerto Rico, flooding was primarily confined to the northeast corner of the island and in urban areas of San Juan. Historical peaks were exceeded on Rio Fajardo and Rio

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA TABLE 2-3 Durations, Amounts, and Return Frequencies for Cubuy and Pena Pobre-Naguabo Duration Frequency (hours) Amount (inches) Return (years) Cubuy 1 1.6 < 1 2 2.9 1 3 4.0 3 6 6.8 16 12 9.1 25 Pena Pobre-Naguabo 1 1.8 < 1 2 3.5 2 3 4.9 6 6 7.7 22 12 9.7 27 FIGURE 2-6 Return frequencies of 12-hour precipitation from Hurricane Hugo.

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA Mameyes. Other streams that neared their peak discharges were Rio Espiritu Santo, Rio Sabena, and Rio Ocacos. In addition, El Carraizo dam on the Rio Grande de Loiza, which impounds water for the San Juan water supply, was overtopped. Figure 2-7 shows some of the erosion that was sustained at this dam. In San Juan there was flooding in low-lying parts of the urban area. These are the same areas that frequently flood in any intense rainfall. The flooding in these areas was more severe for two reasons. First, the pumps used to move the water from low-lying areas lost power and were no longer able to remove water. In some instances, proper maintenance was not being performed, and the pumps were not functional at any time during Hurricane Hugo. Second, the strong winds in San Juan caused many trees to fall into the streets and, for those trees that remained standing, many lost all of their leaves. As a result, the leaves and twigs blocked the drainage grates and acted as local obstacles. LANDSLIDES Larson (1989) reported on the occurrence of 200 landslides in the northeast mountains of Puerto Rico during Hurricane Hugo. Most of these landslides were FIGURE 2-7 Damage at El Carraizo Dam.

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA shallow, with half associated with highway construction or road cuts. According to Larson's observations, 50 percent of the landslides were earthslides, 25 percent were debris flows or slides, and the remaining 25 percent were slumps or rockfalls. The largest slide was 130 m long and 21 m wide and is estimated to have moved 40,000 m3 of soil and rock into a river. SUMMARY AND RECOMMENDATIONS Intense precipitation fell for approximately 12 hours over the Virgin Islands and Puerto Rico during the passage of Hurricane Hugo. Over all of the islands, the 12-hour return precipitation was in excess of the 100-year return frequency over relatively small areas, and the precipitation did not approach the record values. There is, however, an urgent need to be better prepared hydrologically for hurricanes similar to San Ciriaco of 1899 and San Felipe II of 1928. Both of these moved across the center of Puerto Rico, produced heavier and more extensive rains, and caused considerable flooding. Backup power and proper maintenance of pumps would help alleviate some of the local flooding experienced during Hugo. Water supplies must be better protected, and plans should be in place to evacuate people in rural areas. Rain gages in Puerto Rico and the Virgin Islands suffered much the same fate as rain gages during San Ciriaco and San Felipe II. They were destroyed or blown over in heavy winds (Mitchell, 1928). This points to the need for more stable rain gage mounts that can sustain the heavy winds. ALERT rain gages were available for Hurricane Hugo, and they performed well where the winds were not excessively strong (below 57 knots [65 mph]). In regions where the winds and the rains were intense, the antennas of the ALERT rain gages were blown off. New designs need to be initiated to ensure that antennas can sustain intense hurricane winds. A primary tool that was available during Hugo was the WSR-74S radar at Puerto Rico. This radar was able to supply valuable qualitative information about the precipitation field during and after the hurricane 's passage over Puerto Rico and the Virgin Islands. It will be replaced in the future by a WSR-88D or the NEXRAD. It is important that real-time rain gage data, such as that supplied by the ALERT system, be available to this radar. During hurricanes or major convective storms, this new radar will be able to supply quantitative information about the precipitation fields (Hudlow et al., 1989), and will prove invaluable in pinpointing troubled areas where flooding is occurring or where landslides are most likely to occur. Thus, it is important that a site be obtained that can interrogate all of the Virgin Islands and Puerto Rico. In addition, the ALERT system of rain gages and stream gages should be upgraded over Puerto Rico and be extended to the Virgin Islands. Major hydraulic structures need to be better prepared to sustain the intense rains of a hurricane. During Hurricane Hugo, the dam on the Rio Grande de Loiza could not operate its floodgates because of a lack of emergency power and, subsequent to the hurricane, it had no spare pumps to provide fresh water to San Juan. If the hurricane had moved across Puerto Rico and caused more substantial

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA damage to hydraulic structures, there would have been greater problems. It is important that all hydraulic structures be investigated for the ability to sustain hurricane rains and that proper emergency procedures be in place. In those regions where there is the potential for extensive loss of life, the hydraulic structures should be checked to ensure that they can sustain a probable maximum precipitation event. Rainfall frequency and probable maximum precipitation data have not been updated over Puerto Rico and the Virgin Islands since 1961. It is important that new studies be initiated to determine if there have been any changes in the return frequencies of intense rainfall events and to update probable maximum precipitation data for this area.

OCR for page 48
HURRICANE HUGO: PUERTO RICO, THE U.S. VIRGIN ISLANDS, AND SOUTH CAROLINA REFERENCES FEMA. 1985. Federal Interagency Flood Hazard Mitigation Team Report for Puerto Rico, In Response to the October 10, 1985, Disaster Declaration. FEMA-746-DR-Puerto Rico, FEMA Region II. 1985. Washington, D.C.: Federal Emergency Management Agency. Goodman, R. D. 1989. An Historical Look at Tropical Cyclone Paths in Relation to Puerto Rico and the Virgin Islands. National Weather Service Forecast Office, San Juan, internal paper. Hudlow, M. D., J. A. Smith, M. L. Walton, and R. C. Shedd. 1989. NEXRAD-New Era in Rain Hydrometeorology in the United States. Preprint: International Symposium on Hydrologic Application of Weather Radar Paper B-1. Salford, England: University of Salford. Larson, M. C. 1989. Landslides caused by the intense precipitation of Hurricane Hugo, September 1989, eastern Puerto Rico. Water Resources in Puerto Rico and the U.S. Virgin Islands: A Review 8(5/6): 8. Mitchell, C. L. 1928. The West Indian hurricane of September 10-20, 1928. Monthly Weather Review 56: 347-352. Neumann, C. J., G. W. Cry, E. L. Caso, and B. R. Jarvinen. 1985. Tropical Cyclones of the North Atlantic Ocean, 1871-1980 (with storm track maps updated through 1984). Washington, D.C.: U.S. Government Printing Office. Parrish, J. R., R. W. Burpee, and F. D. Marks. 1982. Rainfall patterns observed by digitized radar during the landfall of Hurricane Frederic (1979). Monthly Weather Review 70: 1933-1941. U.S. Weather Bureau. 1961. Generalized Estimates of Probable Maximum Precipitation and Rainfall Frequency Data for Puerto Rico and Virgin Islands. Technical Paper No. 42. Vogel, J. L. 1988. An Examination of Chicago Precipitation Patterns for Water Year 1984 Illinois State Water Survey Contract Report 449, Champaign, Illinois. Willoughby, H.E., F. D. Marks, Jr., and R. J. Feinberg. 1984. Stationary and moving convective bands in hurricanes. Journal of Atmospheric Sciences 41: 3189-3211.