Ports.com

TROPICAL CYCLONES AFFECTING SUVA

Specific tropical cyclone data, containing storm tracks, wind and pressure data, etc., are available starting in 1960. However, prior to 1966 satellites were not used operationally and there are likely some storms that are missing from the data set. Table II-1 contains a descriptive history of all the recorded 49 tropical cyclones and hurricanes passing within 180 nmi of Suva during the 42-year period 1960-2002. Unless otherwise indicated, all of the tropical cyclone statistics utilized in this report for storms passing within 180 nmi of Suva are based on the data set used to compile Table II-1. Note that a December 1973 storm has a double name - Natalie/Lottie. This situation occurs when the storm is initially named by the Australian Forecast Center in Brisbane but later moves into the Fiji area of responsibility and is renamed.

The main feature of storms in this area is that the tracks are very erratic compared to other basins and this makes forecasting difficult. Additionally, this is the only basin where storms do not exhibit the classical recurvature pattern. The prevailing approach direction is from the northwest.

Although the tropical cyclone season for Fiji is considered to extend from November through April, Table II-2 shows that Suva's season extends from October through May. However, only two storms were recorded during May and October and three storms were recorded during November throughout the 42-year period 1960-2002. January had the greatest tropical cyclone activity, with 12 of 49 storms (24.5 %) occurring during this month. Of the 49 storms in the data set, 15 were of hurricane strength (greater than or equal to 64 kt) within 180 nmi of Suva. Table II-2 also shows the average direction towards which storms were moving when they were at CPA to Suva. The average movement for all storms is 135° at 12 kt. The average speed by month is fairly consistent and slow, and ranges from 10-13 kt.

Figure II-14 shows in seven-day increments the annual distribution of the 49 tropical cyclones that entered the 180 nmi threat radius around Suva. The period of maximum activity, from October 4th through May 6th, is clearly shown. It is also interesting to note the relatively consistent distribution from November through May.

Figure II-15 depicts the chronology of the 49 tropical cyclones that passed within 180 nmi of Suva during the 42-year period 1960-2002. As shown in the figure, there is an average occurrence rate of 1.2 storms per year for all tropical cyclones, regardless of intensity. The recurrence interval of all tropical cyclones is 0.8 years. There have been 14 years since 1949 when no tropical cyclone entered Suva's 180 nmi threat radius. 1960,1962, 1963, 1967-1969, 1976, 1977, 1989, 1990, 1993, 1995, 1998 and 2002 had no storms passing within 180 nmi of Suva.

Figure II-16 shows the octants from which the 49 tropical cyclones were moving when they were at CPA to Suva. As shown in the figure, the predominant threat direction is from the northwest octant. A total of 29 storms (59.2 %) approached from the northwest. It must be remembered that the figure represents tropical cyclone movement at CPA, and may not represent the tropical cyclone's initial movement towards Suva. Note that both the threat direction and the slow speed of movement are important considerations for evasion at sea or making storm preparations in port.

Figure II-17 shows the tracks of the 49 tropical cyclones over the entire season. As shown in the figure, the movement of the 49 tropical cyclones that entered Suva's 180 nmi radius can be erratic. It is also intersting to note that 20 of the 49 tropical cyclones were not in existence or were not documented 72 hours prior to CPA.

Figure II-18 is the statistical summary of threat probability based on tropical cyclone tracks for the period 1960-2002. For most typhoon haven sites done in the past, it has been possible to stratify the probability charts into two or more groups, depending on the approach direction and translational speed. However, for this site, typical of this Southern Hemisphere location, no such stratification can be established and only one chart, for the entire season, is provided. The summary coincides with the tracks presented in Figure II-17. The data is presented with solid thin lines representing "percent threat" for the 180 nmi radius circle surrounding Suva. The solid dashed lines represent approximate approach times to Suva based on the climatological speeds of movement. In Figure II-18 for example, a tropical cyclone located near 12°S 172.5°E has approximately a 40 % probability of passing within 180 nmi of Suva during the October to May period. In addition, this tropical cyclone would reach Suva in 3-4 days if the speed remains close to the climatological normal for tropical cyclones passing within 180 nmi of Suva.

Wind

Wind measuring instruments are located on the top of the Port Control Building close to King's Wharf (Figure II-3). The measured winds are representative of those experienced by ships in the Port. The Deputy Port Master stated during the site visit in November 2002 from his experience that winds from tropical cyclones gusting from 40-60 kt have been experienced in the past. The Deputy Port Master and Pilots further stated that small ships are relocated to anchorages or mooring buoys and visiting merchant ships and Navy ships would normally sortie at least 48 hours in advance of an approaching tropical cyclone.

As reported in the List of Tropical Cyclones affecting Fiji from 1830 to 2000 Seasons, a severe hurricane on the 20th of February 1941 put "ocean-going vessels ashore in Suva Harbor with a recorded maximum wind of 100 mph."

The Port is protected from the prevailing east wind. However strong easterly and southeasterly winds are the most dangerous to shipping in the channel. The Pilots indicated during the Site Visit in November 2002 that vessels would probably not be permitted to enter or leave the Port with sustained winds greater than 19 kt.

Wind speeds in Fiji are reported as 10-minute averages, i.e. the average wind speed observed over a 10-minute period of time. Wind speeds in the United States are reported as 1-minute averages, i.e. the average wind speed observed over a 1-minute period of time. The difference resulting from these two measurement periods is that the United States' wind speeds will be greater by a factor of 1.1 or 1.2. Consequently, a 10 kt wind as used in Fiji observations, forecasts, and warnings would convert to a speed of 11 or 12 kt in United States' observations, forecasts, and warnings.

Waves

Pilots indicated during the Site Visit in November 2002 that the maximum wave height experienced inside the Harbor during a hurricane passage was 6.6 ft (2 m). Open ocean wave motion is negligible as the Port is almost surrounded by land and reef breaks. 9.8-16.4 ft (3-5 m) waves have been experienced outside the Harbor during hurricane passages but Pilots as well as data from meteorological reports indicate that higher waves are possible.

Storm Surge

Long island coastlines with many bays and inlets and a gradual sloping seabed are more vulnerable to surge conditions because there is less opportunity for the surge waters to evacuate around the sides of the island. In the Southern Hemisphere the full impact of a surge is experienced to the left of the cyclone track in the direction of approach to an island as a result of the violent onshore winds. Hurricane June in May 1997 had a large storm surge impact, inundating many northern coastal areas of Viti Levu. Another example in 1997 was the surge caused by Hurricane Gavin. A sea level monitoring station at Lautoka Wharf, operated by the National Tidal Facility of Flinders University South Australia, recorded a storm surge of 9.2 ft (2.8 m) on 7 March. At the same time a sharp drop in atmospheric pressure at nearby Nadi confirmed this surge.

Because of the extra rise in sea level caused by storm surges, coral reefs that usually afford protection around island coastlines become well submerged at high tide. This can allow the large waves driven by high winds to attack exposed locations, leading to the removal or redistribution of sandy, low reef islands such as cays and the erosion of beach materials. The Yasawa and Mamanuca Islands of Fiji experience the full brunt of storm waves and many beaches suffer degradation as a result.

As the tropical cyclones generally approach Fiji from the northwest, Lautoka on the northeast coast is more susceptible from surge than Suva on the southwest side of Viti Levu Island. The Pilots reported that the maximum surge they had experienced in Suva was 3.3 ft (1 m). However, a characteristic of storms in this area is erratic motion and this is obvious from the track chart provided. Thus, even though Suva is on the lee side of the island insofar as storm approach is concerned, it is possible for storms to make landfall on the "back" side of the island.

"Most notable accounts of storm surge in the area date back to 1886 when it was reported by government officials that Gau Island had been inundated by the sea. It was estimated from the marks on the trees that sea level had risen by 5 meters" (Campbell, 1951).

On the 30th of December, 1959 a moderate hurricane produced high seas in southern Viti Levu that were described as being in the nature of tidal waves that swept inland and destroyed many coastal villages. Oscar in March of 1983 was classified as the worst natural disaster to strike Fiji since the cyclone of 1931. High winds, flooding due to heavy rainfall, and storm surge caused widespread devastation in the west and south of Viti Levu and many other islands. The Fiji Meteorological Service report stated in part that storm surge was a particular notable feature of the cyclone. Surge inundated and severely damaged a number of establishments and crops in coastal farmlands and caused considerable erosion in the shoreline. Near Nandi Airport (not far from the Port of Lautoka) storm surge was estimated to have been about 6.6 ft (2m) and eyewitness reports suggested that it was about 9.8-13.1 ft (3-4 m) at Momi, about 14 nmi southwest of Nandi Airport.

Source: http://www.nrlmry.navy.mil/port_studies/thh-nc/fiji/suva/text/sect7.htm