The Naduruloulou Research Station was established in 1975. Since beginning operation several floodings have been reported. The first report dates back to 1980 when the carp brood stock was washed away. Extreme damage caused by floods are again reported in 1983 (Cyclone Oscar), and in 1993 (Cyclone Kina). In 1993 the lower part of the site, where the earthen ponds are located, was completely flooded; with an approximately 3 to 4 meter high water column above ground level being observed. In 1997, Hurricane Gavin flooded the site.
It is likely that flooding of the site occurred before 1980 also, but due to the lack of aquaculture operations, they were not recorded.
Scarcity of data hinders the precise computation of probability of occurrence of flooding. The available data, previous reports and assessment of surface formations and elevations make it, however, possible to draw some reliable conclusions.
The NRS site is located about 7 km north of Nausori Town, in an upstream direction along Rewa River on the left bank about 300 m from the river. In the very close vicinity of NRS there is the village of Kasavu located at the bank. Although there is a small hill between the NRS site and the river, a side valley links the site to the river's inundation basin.
Drainage of NRS is maintained through the side valley, and the drainage canal discharges into the river at Kasavu. Elevation of the site9 varies between 3,5 to 11 m above mean sea level (amsl); the ponds occupy the lower part of the site. Crest levels of ponds are at ground level and do not exceed 6 m amsl.
Continuous series of data are available since 1972 from the Nausori water level gauge (located at the Nausori bridge, Courthouse), but the distance between Nausori and NRS limits the direct transposition of readings of Nausori station to the site of NRS. It should be noted here that the crest level of the river bank at Nausori is around 5 m amsl; above 5 m inundation takes place.
There is also a water level gauge at Drekenikelo about 2 km downstream from NRS, but readings are incomplete and processed data are limited to some peak yearly water levels for the period between 1980 and 1990. This set of data must, therefore, be regarded with due care.
Reliability of the Drekenikelo readings is, however, supported with a measurement taken in respect to Cyclone Kina (1993). Village people of Kasavu have marked the highest flood level, which mark was checked in January 1998. Its reading is 9.29 m amsl. This figure demonstrates that difference between crests of a flood at NRS and at Nausori may be considerable and may reach even 2–4 metres.
The peak yearly water levels measured at the Nausori station, at the Drekenikelo Station and at Kasavu are jointly presented in Table 1.
Correlation coefficient between Nausori and Drekenikelo figures is 0.5, which means a loose correlation between the two sets of data. Correlation is affected by the tide at Nausori, the routing of flood, the likely inundation of a large area between the two gauge stations, attenuation of flood and also by various geographical formations.
According to fisheries staff of NRS the area gets inundated during every major flood which is normally accompanied by a cyclone. Table 5 only includes references to reported flooding.
9 Site survey was carried out by the Dredging Project, Nausori, MAAF on 28 January 1998.
Table 5. Annual peak water levels of Rewa River
|Year||Date||Peak discharge at Nausori station|
|(m3/sec)||(m amsl)||(m amsl)||(m amsl)|
|1980||04 Apr||11,485||6.18||inundation reported||Tia+Vally|
|1983||03 Mar||7,770||5.00||7.35||inundation reported||Oscar|
|1993||03 Jan||13,020||5.49||inundation reported 9.29||Kina|
|1997||09 Mar||6,059||3.75||inundation reported||Gavin|
1. Decimal: dot, thousand separator: comma
2. amsl=above mean see level
Figure 3. Annual peak discharges and corresponding water levels of Rewa River (1972–1997)
Figure 4. Annual peak water levels of Rewa River (1972–1997)
Despite the relatively loose correlation comparison of site elevation at NRS and measurements at Nausori may lead to the conclusion that a prolonged flood at Nausori with the flood crest above 4 m amsl would cause inundation of all those plains within a certain distance of which the elevation is 4 m amsl. The higher the flood crest at Nausori, the greater the affected distance.
Further analysis of data (Figure 4) shows that during the period of 25 years (from 1972 to 1997) the yearly peak water level exceeding 4 m amsl has occurred 13 times (50%), and peak water level exceeding 5 m amsl has occurred 5 times (20%). Taking the latter into consideration it is reliable assumption that a prolonged water level of 5 m amsl at Nausori would cause inundation at NRS. Such situation has 20% probability of occurrence, or in other words, one flooding may occur in every 5 years. It is necessary to note here, that 4 occurrences of inundation of NRS have already been reported during the period of 25 years, that equals 16% probability of occurrence.