Hurricane impact on beaches in the eastern Caribbean Islands 1989 - 1995

7.  EFFECTS OF THE 1995 HURRICANES ON THE ISLANDS FROM ANGUILLA TO  DOMINICA continued

7.4  St. Kitts

Tropical Storm Iris passed to the east of St. Kitts and caused little damage.  The centre of Hurricane Luis passed 70 km (44 miles) east of St. Kitts between 4th and 6th September, 1995.  There was considerable infrastructural damage and the reclamation that was in progress in Basseterre Harbour was severely impacted.  All the coasts of St. Kitts were impacted by the hurricane force waves.

St. Kitts is a volcanic island in the Leeward Islands.  The highest peak is Mount Misery, the upper slopes of which are forested, while the lower slopes are cultivated for sugar cane.  The main part of St. Kitts is joined by a narrow neck of land to the Southeast Peninsula which is little developed.  The coastline of St. Kitts can be divided into two main sections: the southwest coast from Basseterre to Dieppe Bay which consists of black sand and boulder beaches and sections of cliffs; and the rest of the coast which consists of a mixture of silica, coral and volcanic sand beaches.  Particularly on the north coast and in the Peninsula the beaches are backed by low sand dunes and often enclose salt ponds.  The coastline in the Peninsula is indented while the rest of the coastline is fairly straight.

A beach monitoring programme was established in St. Kitts in 1991.  Nineteen beaches (35 profile sites) are monitored on a regular basis.  Appendix VII shows the location of the monitored beaches.  Prior to Hurricane Luis in 1995, 21 of the 35 profile sites were showing erosion and the average erosion rate for all the beaches was 0.27 m/yr (Cambers et al., 1995).  Erosion was more widespread on the Peninsula than on the mainland.

In 1989 Hurricane Hugo passed close to St. Kitts and damaged the coastline and beaches.  Following this hurricane several stretches of coastline were protected with boulder revetments, these sites included a section of the Basseterre waterfront east of the town, the northern end of Old Road, a cliffed section near Godwin Ghut and the road at Pump Bay.  There are no measurements of beach changes caused by Hurricane Hugo.

Hurricane Luis caused considerable damage to the beaches in St. Kitts.  Beach measurements were taken over the period March to August 1995 (the mainland sites in March and the Peninsula in August) and again in the middle of September, 1995.  Profile area before and after the hurricane were compared and the difference computed as a percentage change.  The change in profile width was calculated as an absolute value in metres.  Finally the change in position of the edge of the land or the base of the sand dune was determined from the profile plots.  The data for each site at each beach are shown in Appendix VII.  The data are summarized in Table 13, the second column shows the overall average for the island, the third column relates to the Mainland beaches only and the fourth column to the Peninsula beaches.

 

Photograph 7.  South Friars Bay, St. Kitts, May 1996 The sand beach used to extend  seaward of the rock stack. The beach  had eroded by 33 m in Hurricane Luis  leaving the rock stack out in the sea.

There was considerable variation in the beach changes in St. Kitts resulting from Hurricane Luis.  The erosion was less severe in the Mainland than in the Peninsula.  When the Mainland sites were averaged there was overall accretion.  The most severe erosion on the Mainland was at the western end of Basseterre Bay and at Dieppe Bay near the spit.  On the Peninsula, however, the erosion was more severe, particularly at South Friars Bay, Majors Bay and Banana Bay.  (There was also severe erosion at North Friars Bay, but this beach usually shows very dramatic seasonal changes, Cambers et al., 1995).  At the northern end of South Friars Bay, the sand beach used to extend almost out to the headland, surrounding a rock stack with sand.  After Hurricane Luis, the beach had narrowed by 33 m (108 ft) leaving the large isolated rock stack surrounded by water, see Photograph 7.  At other beaches the sand had been moved from one end of the beach to the other, e.g. Cockleshell Bay where the western end showed erosion and the central and eastern profile sites showed accretion.

When the data for all the sites were averaged, the profile area decreased by 6%, the profile width narrowed by 3.2 m and the dune/land edge retreated inland 4 m.  this latter change is regarded as a permanent loss, since although the beaches recover after a hurricane, the position of the beach will be further inland, see also Section 3.1.

In order to try and assess the extent of beach recovery, the pre-hurricane data for August 1995, have been compared with the data for April 1996 for the sites in the Peninsula.  (1996 data were not available for the Mainland sites at the time of writing).  The profile area for April 1996 has been calculated as a percentage of the value for August 1995.  Similarly the profile width for April 1996 has been calculated as a percentage of the value for August 1995.  Thus these values show the amount of recovery as a percentage of the pre-hurricane values.  The profile width has also been calculated in absolute terms.  The data are summarized below in Table 14 and the individual data for each beach are contained in Appendix VII.

Seven months after the hurricane the beaches in the Peninsula had recovered to 94% of their pre-hurricane values.  However, the beaches were still on average 4 m narrower.  Reference to Appendix VII shows that there was considerable variation between beaches.  Recovery was lowest at South Friars Bay.  At Banana Bay West the recovery was also slow and it appears that morphological changes may be impeding the recovery here.  The hurricane waves had exposed a beachrock ledge in the inter-tidal zone.  This may be acting as a wall and reflecting the wave energy thereby causing more turbulence and impeding the beach recovery at this site.  Figure 12 shows the beach profile at this site before and after the hurricane and Photograph 8 shows the beachrock ledge.  (A similar phenomenon was recorded at Morris Bay in Antigua, see Section 7.2).

7.5  Montserrat

Tropical Storm Iris passed west of Montserrat causing only slight damage.  The centre of Hurricane Luis passed 90 km (56 miles) northeast of Montserrat between 4-6th September, 1995 causing significant damage to the coast.

Montserrat is a volcanic island located in the Leeward Islands.  The terrain is generally rugged and there is only a narrow coastal plain.  Most of the beaches are located on the west coast of the island and they vary between volcanic sand and boulders.  There is one coral sand beach at the northern end of the west coast, at Rendezvous Bay.  The coastline is generally fairly straight with few significant embayments.

Since July 1995 the Soufriere Hills volcano in Montserrat has resumed activity and the population has had to endure periodic evacuations to the northern end of the island.  This has had a significant impact on all aspects of the island’s economic and social activity and is likely to influence future planning in both the short and medium term.

Montserrat was severely impacted by Hurricane David in 1979.  In 1989 the centre of Hurricane Hugo passed directly over Montserrat.  The damage from this hurricane was catastrophic, both to the island’s infrastructure and the environment.  Most of the island’s housing stock was damaged or destroyed, the port was badly damaged together with the hospital.  The beaches were severely eroded, and as post hurricane reconstruction began, the erosion continued as the beaches were mined for construction sand.  However, since 1992 the government has taken several steps to manage the problem.  These have included the purchase of a new stone crusher, the substitution of quarry dust for beach sand in all government construction contracts, and the closure of all the beaches to sand mining except for Farms Bay.

A beach monitoring programme was established in Montserrat in 1990, immediately after Hurricane Hugo.  Ten beaches are monitored, nine of which are on the west coast.   In July 1995, five additional profiles were established, making a total of fifteen profile sites.  Between 1990 and 1993 most of the beaches were accreting as they recovered from Hurricane Hugo (Cambers & Christopher, 1995).  The main exception was Farms Bay on the east coast which has remained open to sand mining.  During 1994 an erosion trend replaced the previous accretion trend.

Beach measurements were taken in July 1995 and in the middle of September, 1995.  Profile areas before and after the hurricane were compared and the difference computed as a percentage change.  The change in profile width was calculated as an absolute value in metres.  Finally the change in position of the edge of the land or the base of the sand dune was determined from the profile plots.  The data for each site at each beach are shown in Appendix VIII, the data are summarized in Table 15.

 

The hurricane caused extensive erosion along the west coast of Montserrat, especially at the northern ends of Sturge Park Beach and Foxes Bay; Woodlands Bay, Bunkum Bay and Little Bay.  At Foxes Bay the seaward side of the Bird Sanctuary was extensively damaged and the channel to the sea was opened up.  Figure 13 shows the beach profile at Sturge Park North (in front of the Montserrat Springs Hotel) before and after the hurricane.  The low grassy dune area behind the beach had been totally eroded.  This was one of the most severely eroded beaches.

When the data for all the sites were averaged, the profile area had decreased by 31% and the profile width narrowed by 11 m (36 ft).  There are no significant sand dunes in Montserrat, but the land edge retreated inland an average of 3.5 m.  This latter change is regarded as a permanent loss, since although the beaches recover after a hurricane, the position of the beach will be further inland, see also Section 3.1. 

A report (Cambers, 1990) showed that at Sturge Park Beach the edge of the land had retreated inland between 10 and 16 m between 1966 and 1990, see Photograph 9.  This erosion reflects the influence of two major hurricanes (David in 1979 and Hugo in 1989) as well as major sand mining activities and port construction.  Severe hurricanes, such as David, Hugo and Luis, are undoubtedly a major cause of the retreat of the coastal land edge.

In order to try and assess the extent of beach recovery, the pre hurricane data for July 1995, have been compared with the data for January 1996 (more recent data were not available at the time of writing).  The profile area for January 1996 has been calculated as a percentage of the value for July 1995.  Similarly the profile width for January 1996 has been calculated as a percentage of the value for July 1995.  Thus these values show the amount of recovery as a percentage of the pre-hurricane values.  The profile width has also been calculated in absolute terms.  The data are summarized below in Table 16 and the individual data for each beach are contained in Appendix VIII.

Four months after the hurricane the beaches had only recovered to 79% of their pre-hurricane levels.  On average the beaches were still 7 m narrower than before the hurricane.  There was considerable variation from beach to beach.  Some beaches such as Sugar Bay had shown considerable recovery, others such as Sturge Park North, Woodlands Beach and Bunkum Bay were much slower to recover.  Undoubtedly four months is too short a time to assess beach recovery, especially since the data record shows that after Hurricane Hugo the beaches continued to recover over a four year period (Cambers and Christopher, 1995).

7.6  Dominica

Tropical Storm Iris passed 30 km west of Dominica, Hurricane Luis passed 180 km north of Dominica and Hurricane Marilyn passed less than 20 km east of the island.  All three storms impacted Dominica, but Hurricane Marilyn was the strongest and closest.  Following the storms a report was prepared (James, 1995) and much of the information relating to the beach damage contained in this section is derived from that report.

The combined damage from the three storm systems was estimated at US$ 184 million.  Tropical Storm Iris destroyed about one half of the island’s banana crop, damaged the west coast road and cut road access between Soufriere and Scotts Head.  Hurricane Luis damaged coastal structures - hotels, roads, utilities, jetties and fish landing sites.  Damage was also sustained in agriculture, forestry and shipping sectors.  Hurricane Marilyn caused similar damage, but particularly in the north and eastern parts of the island.

As discussed in Section 6.1, beach monitoring was started in Dominica in 1987, the monitoring continued until 1992 when it was discontinued due to a shortage of manpower.  The programme was resumed in 1994 and the list of sites was modified.  Appendix IX shows the location of the sites where monitoring was resumed/started in 1994.

There was severe damage to the beaches, especially on the west coast.  In particular the northwest beaches from Prince Rupert Bay to Toucarie were heavily impacted.  The road was washed out and the hotel damaged at Coconut Beach, similar damage occurred at Mero.  Trees were undermined and washed away, and in some cases - Toucarie, Belle Hall and Mero - the sand was replaced with stones and boulders.  At some sites such as Batalie, debris from the reefs was washed up onto the beach.  New beachrock formations were exposed at some of the beaches e.g. Woodford Hill.

A set of measurements were taken at most beach sites in January 1995, these were used as the pre-hurricane data set.  Some beaches were measured at the end of August, after Tropical Storm Iris, others were measured in September after Hurricane Luis and then others after Hurricane Marilyn.  Thus the post hurricane database does not necessarily reflect the impacts of the same event or combinations of events.  Nevertheless the two data sets have been compared, although it should be realised that the combined impacts of the three storms were probably more severe than reflected in this comparison.

Profile area before and after the hurricane were compared and the difference computed as a percentage change.  The change in profile width was calculated as an absolute value in metres.  Finally the change in position of the edge of the land were determined from the profile plots.  The data for each site at each beach are shown in Appendix IX, the data are summarized below in Table 17.

The data show that the erosion was most severe on the west coast beaches particularly at Rockaway Beach, Mero Beach, Coconut Beach and Purple Turtle.  The geographical distribution of the beach erosion is very similar to that recorded during Hurricane Hugo in 1989, see Section 6.1.  For instance the three beaches showing the most severe erosion were the same in 1989 as in 1995, although the order varied slightly.  Similarly Batalie Beach was fairly stable during both storms, it accreted slightly in 1989 and eroded slightly in 1995.  Comparison of the average figures showed that the 1995 erosion was slightly more severe than in 1989.

The north and east coast beaches showed some variation after the 1995 storms.  For instance at Woodford Hill on the north coast, the central and western part of the bay showed erosion while the eastern end showed accretion.  At Bout Sable on the east coast, there was slight accretion.  In contrast there was severe erosion at Pagua, as was the case during Hurricane Hugo in 1989.

When the data for all the sites were averaged, the profile area decreased by 24%, the beaches narrowed by 6.7 m and the land edge retreated inland 2.4 m.  This latter change is regarded as a permanent loss for although the beaches recover after a hurricane, the position of the beach will be further inland, see also Section 3.1.  Photographs 10 and 11 illustrate how the land edge has moved inland at Coconut Beach.  In 1987 before Hurricane Hugo there was a wide beach and a tree lined area between the beach and the road.  Four months after the 1995 storms the road was adjacent to the water.  Figure 14 compares the pre and post hurricane profiles at this site, some data extrapolation was necessary in this comparison.

In order to try and assess the extent of beach recovery, the pre hurricane data for January 1995, have been compared with the data for January/April 1996.  The profile area for January/April 1996 has been calculated as a percentage of the value for January 1995.  Similarly the profile width for January/April 1996 has been calculated as a percentage of the value for January 1995.  Thus these values show the amount of recovery as a percentage of the pre-hurricane values.  The profile width has also been calculated in absolute terms.  The data are summarized below in Table 18 and the individual data for each beach are contained in Appendix IX.

Seven months after the hurricane the beaches had recovered to 89% of their pre-hurricane levels on average.  However, recovery data was only available for three of the seven west coast beach sites measured, thus this cannot be considered a complete picture of the recovery.  The beaches were still on average 3 m narrower.

7.7  Comparative Assessment of the Effects of the 1995 Storms

The data base is not sufficiently detailed to determine the individual effects of each storm/hurricane.  In nearly all islands the post hurricane measurements were taken after Hurricane Marilyn so the impacts of the three storms are grouped together.  With the possible exception of Dominica, it appears that Hurricane Luis was the major event as regards beach changes.

Table 19 summarizes the beach change data for each island as well as the change in the land/dune edge and the proximity to the centre of Hurricane Luis.

 

Figure 15 shows profile area change plotted against the distance from the centre of Hurricane Luis.  As was expected, the erosion increased as the distance to the centre of the storm decreased.  However, there is a considerable scatter of the data points, so it is obvious that other factors come into play.  These factors include variation in the strength of the storm’s four quadrants; effects of previous storms on the beaches (both in 1995 and further back such as 1989); and geographical variations including existence of coral reefs, degree of beach exposure and width of the offshore shelf.

There was no clear relationship between the change in profile width and the proximity to the storm centre.  For instance the least change in beach width was experienced in Barbuda, yet the centre of Hurricane Luis passed over this island.  As explained in Section 7.2. the extensive sand dunes behind the beaches in Barbuda were eroded and much of that reservoir of sand was transferred to the lower part of the beach.

Figure 16 shows the change in position of the edge of the land/dune plotted against the distance from the centre of Hurricane Luis.  This displays the expected relationship - the closer the hurricane, the greater the land/dune edge retreat.  As discussed previously, the retreat of the land/dune edge inland is regarded as a "permanent" effect of the hurricanes.  Figure 16 also shows that there is a threshold around 40 km from the storm centre.  Between the storm centre and 40 km, erosion rates varied between 18 m and 5 m.  At greater distances from the storm centre and up to 180 km, erosion rates were generally less than 4 m.