CHAPTER 9 CLIMATIC INFLUENCES ON FLOODING IN KWAZULU-NATAL
9. Introduction
9.2. Discussion
Examination of the temporal distribution of regional storm events indicates two factors that have to be accounted for namely, what controls the number of regional storm event events in a year and what controls the discharge magnitude for the regional storm events. As expected the correlation between mean annual precipitation (wet cycles) and the regional storm event count are generally positive with the result that wetter periods are more likely to produce a greater number of regional storm events (wet cycle = 62%, dry cycles = 38%). However, the dry cycle periods from 1897 – 1915, 1963 – 1971 and 1992 – 1995 which show higher regional storm event counts than other dry cycle periods (Fig. 9.4), indicate that mean annual precipitation is not the only climatic factor influencing regional storm event occurrences.
The gravitational effect of the moon on the earth’s atmosphere and oceans by the Lunar Nodal Cycle have been recognised by a number of researchers (e.g. Tyson 1986; Yndestad 2006; Yasuda 2009; Malherbe et al. 2014). Currie (1984) linked the Lunar Nodal Cycle with the wet and dry rainfall cycles, Yasuda (2009) detected the 18.6 years Lunar Nodal Cycle’s influence on the Pacific Decadal Oscillation and Malherbe et al. (2014) found that it influenced the Antarctic Oscillation. The Lunar Nodal Cycle influence has also been related to coastal erosion (e.g. Gratiot et al. 2008; Smith et al. 2010) and tidal sedimentary deposits (e.g. Ray 2007; Mazumder & Arima 2005). While the influence of the Lunar Nodal Cycle on climatic cycles are well documented there appears to be limited correlations with the regional storm events and extreme storm events. There is a correlation between the Lunar Nodal Cycle and Wet/Dry Cycles up to 1985, thereafter they occur out of phase. This indicates that the Wet/Dry Cycles may not necessarily only be affected by the Lunar Nodal Cycle. Cook et al. (2004) suggests that wet cycles may not necessarily consist of wetter periods but rather longer intense periods of wetness.
Sunspot cycles, global temperatures, sea surface temperature and atmospheric CO2 appear to have no impact on the number of regional storm event annual counts or discharges. Comparison of climate cycles show that neither the Indian Ocean Dipole, Atlantic Multi-decadal Oscillation, El Niño Southern Oscillation, Antarctic Oscillation nor Pacific Decadal Oscillation have any correlation with the regional storm events annual count. The best correlation between the annual occurrences of regional storm events is with the Southern Oscillation Index. Apart from an anomalous period around 1940, the mean annual precipitation, wet/dry cycles, and the Southern Oscillation Index have a positive correlation. The
anomalous periods identified from the mean annual precipitation and wet/dry cycles 1963 – 1971 and 1992 – 1995 are also present in the Southern Oscillation Index but not the 1897 – 1915 anomaly. From the available data the most likely driver of the amount of the regional storm events is the teleconnection from the Southern Oscillation Index when sea surface temperature increases in the Pacific Ocean causing a shift in the Walker Circulation to positive as noted by Preston-Whyte & Tyson (1997).
While correlation with El Niño Southern Oscillation is not conclusive as the driver for regional storm event discharge magnitude, the correlation for regional storm event discharges and the Pacific Decadal Oscillation cycle is very good. In Chapter 3, Table 3.1 it was shown that the easterly wave systems dominate the summer rainfall (December – March) coinciding with the highest regional storm event count (Chapter 8). The easterly waves transport moist air from the south west Indian Ocean over the land and when sea surface temperatures increase over the south west Indian Ocean, these waves transport moister air resulting in wetter years (Washington & Preston 2006).
The decadal fluctuations of the Pacific Decadal Oscillation warming the south west Indian Ocean (Alexander & Bladé 2002) (Fig. 9.25) can account for the correlation of wet/dry cycles as a function of wetter and dryer years and the magnitude of regional storm event and extreme flood event discharges. An investigation of two periods which resulted in high magnitude regional storm event discharges during 1974 - 1976 (Washington & Preston 2006) and 1995 - 1996 (Rautenbach 1998) attribute these anomalous wet periods to increased sea surface temperature in the south west Indian Ocean and they correlation with the positive phase of the Pacific Decadal Oscillation. Ramsay & Cohen (1997) investigated a coral core taken from a site off the northeast coast of the study area, and identified fluorescent banding which they attributed to extreme flood discharges (1886, 1891, 1925, 1939, 1984, 1985, and 1987). There thus appears to be a strong correlation with the coral proxy dates and that of the Pacific Decadal Oscillation cycle peaks (Fig 9.26).
Figure 9.26 Annual regional storm event and extreme storm event discharge comparison with the Pacific Decadal Oscillation (PDO) and coral proxy data (Ramsay & Cohen 1997). Note the periods of increased sediment incorporation
into the corals around 1885, 1937 and 1985. Also note the increased extreme storm event discharge magnitude correlation with the coral proxy periods.
Figure 9.25. Map showing a positive Pacific Decadal Oscillation sea surface temperature distribution. Note the warm water in the eastern Pacific Ocean and associated cold water in the northern western Pacific Ocean and warmer water in
the western Indian Ocean (Image: After Wikipedia Commons based on data from Hadley Met Office (HMO 2014)
According to Alexander & Bladé (2002) the sea - air interaction of the Southern Oscillation Index is the driver for El Niño Southern Oscillation and creates global teleconnections that can influence sea surface temperature around the world. El Niño Southern Oscillation is the forcing agent of the Pacific Decadal Oscillation causing variability over a decadal scale (Newman et al. 2003). The relationships between Southern Oscillation Index, El Niño Southern Oscillation and Pacific Decadal Oscillation and the teleconnections that result are complex, as can be seen from Figures 9.27 and 9.28. According to Preston- Whyte & Tyson (1997), the Southern Oscillation Index plays a major role in determining the southern African climate. The Southern Oscillation Index forcing of El Niño lags so that the Southern Oscillation Index and the El Niño Southern Oscillation are generally 180° out of phase. The El Niño Southern Oscillation forcing of the Pacific Decadal Oscillation has the same general trend. With El Niño causing dryer conditions and La Niña wetter conditions (Tyson et al. 2002). However warmer water in the eastern Pacific Ocean results in El Niño conditions which leads to less rainfall over southern Africa and Pacific Decadal Oscillation conditions that correlate with regional storm event discharges. This suggests that the El Niño Southern Oscillation forcing of the Pacific Decadal Oscillation creates a lag in the Pacific Decadal Oscillation or that the teleconnections from Pacific Decadal Oscillation are stronger than those of the El Niño Southern Oscillation.
Figure 9.27. Annual regional storm event and extreme storm event annual count comparison with the El Niño Southern Oscillation , Southern Oscillation Index , and Pacific Decadal Oscillation. Note that the Southern Oscillation Index , which is seen as the driver for the El Niño Southern Oscillation, is generally 180° out of
phase with the El Niño Southern Oscillation.