Home | About Us | Contacts | Help | Feedback | SEARCH Global | Australia | NSW | Vic. | Qld | WA | SA | Tas. | ACT | NT | Ant. | Weather & Warnings | Hydrology | Climate | Numerical Prediction | About Services | Learn About Meteorology | Registered User Services |

Global Guide
Main Menu

Global Guide to Tropical Cyclone Forecasting:
CHAPTER 5: SEASONAL FORECASTING

5.2 ENSO RELATIONSHIPS WITH SEASONAL TROPICAL CYCLONE ACTIVITY

5.2.1 North Atlantic Basin

Tropical-cyclone activity in the North Atlantic is more sensitive to El Nino influences than in any other ocean basin. Gray (1984a) and Fig. 1 indicate that comparing a non-El Nino year to one with moderate to strong El Nino, the North Atlantic basin experiences:

1. A substantial reduction in cyclone numbers, especially at low latitudes;
2. A 60% reduction in numbers of hurricane days¹;
3. An overall reduction in system intensity.

Gray et al. (1992) attribute this reduction of tropical cyclone activity to the anomalously strong westerly winds that develop in the western North Atlantic and Caribbean region during El Nino years.

5.2.2 Eastern North Pacific Basin

Satellite estimates of tropical cyclone intensity for this region are available only for the 24 years since 1966, during which five strong El Nino events occurred (Landsea and Gray, 1989). Although the satellite intensity estimates are less accurate than aircraft, there seems to be a strong signal for increased frequency of hurricane force systems during El Nino years (Table 5.1). This finding is physically consistent with the known association of maximum potential tropical cyclone intensity with sea surface temperature (Emanuel, 1986; Merrill, 1988).

Figure 5.1: Tracks of all hurricane intensity systems in the North Atlantic for the period 1900-1976 grouped relative to El Nino years: a)one year before, b) during, and c) one year after (Gray 1984a).

5.2.3 Central and Western Pacific Regions

The total amount of tropical cyclone activity in the western Pacific varies little during ENSO cycles. But recent work by Hastings (1990) and others has clearly demonstrated an eastward displacement of the primary centres of tropical cyclone activity during El Nino years. As shown in Figs. 5.2 and 5.3, tropical cyclones extend a considerable distance eastward during strong El Nino episodes and there is a reduced frequency in the Coral Sea and Eastern-Australian region. These results agree with the earlier findings of Dong (1988) and Revell and Goulter (1986a,b).

Tropical cyclone activity during the intense 1982-83 El Nino provided an extreme example of this eastward shift in tropical cyclone occurrence. Many more tropical cyclones (eight systems) formed east of 180o in the South Pacific in 1982-83 than in any previous year on record, and activity in the Australia region was much reduced. Hastings (1990) also shows that a later than normal start of the South Pacific tropical cyclone season typically occurs during the year following an El Nino.

A similar but less dramatic change occurs in the western North Pacific basin. Chan (1985, 1990a, 1990b) shows that the frequency of tropical cyclones in the North Pacific between 140-160^oE is increased during El Nino years (Fig. 5.4). South China Sea activity follows an inverse relationship, experiencing decreased activity in El Nino years and increased activity in anti-El Nino years. Chan also shows that western North Pacific tropical cyclone activity typically decreases in years following an El Nino, except in the South China Sea (Fig. 5.5).

5.2.4 North and South Indian Ocean Regions

Statistical analysis of the records for the last 40 years of tropical cyclones in the Indian Ocean indicates no obvious systematic, ENSO related variations of seasonal tropical cyclone frequency or location in the North and South Indian Oceans. However, more careful studies of Indian Ocean cyclones are needed. It is likely that meaningful seasonal influences are present and may be elucidated in more detailed analyses.

Figure 5.2: The western South Pacific region tropical cyclone tracks for the El Nino composite (Hastings, 1990).

Figure 5.3: The western South Pacific region tropical cyclone tracks for the anti-El Nino composite (Hastings, 1990).

Figure 5.4: Seasonal anomalies of tropical cyclone numbers in the eastern part of the western North Pacific basin (5-15oN, 140-160oE; Chan, 1990). Letters indicate a weak (W), moderate (M), or strong (S) El Nino.

Figure 5.5: Average seasonal anomaly in tropical cyclone numbers per 5o Marsden square for years following an El Nino (Chan, 1990).

5.2.5 Australian Region

Nicholls (1979) was the first to report that the austral winter to spring anomalies of sea level pressure at Darwin, which is closely associated with the SOI, is highly correlated with early season Australian region tropical cyclone activity and to a lesser extent with total seasonal cyclone activity. Subsequent research and operational testing (Nicholls, 1984, 1985, 1992; Drosdowski and Woodcock, 1991; Ready and Woodcock, 1992) well verifies these results (Fig. 5.6).

Figure 5.6: Scatter diagram of the seasonal number of Australian region tropical cyclones versus the mean September to November SOI (Nicholls, 1992). Seasons after 1986 are indicated by the full dots, from 1959-1986 by open dots; the line is the regression using all data.

There are notable differences in the physical causes of ENSO induced variations of tropical cyclone activity between the Australian and North Atlantic regions. Whereas the strength of the upper level westerly wind and the vertical shear mechanism appears to account for tropical cyclone reduction in the Atlantic in El Nino years, differences in sea surface temperatures and surface pressure seem to provide the primary physical linkage in the Australian region. Here, cool sea surface temperature anomalies and associated high barometric pressure accompany El Nino events and are associated with the diminished frequency of Australian Coral Sea area cyclones. By comparison, the El Nino appears to cause no significant alteration in Atlantic Sea surface temperatures or sea level pressure.

5.2.6 Summary of ENSO Relationships

The ENSO modulation of total tropical cyclone frequency and intensity is strongest in the North Atlantic basin. Strong zonal displacements in tropical cyclone location occur in the central and western Pacific and a later start of the tropical cyclone season typically occurs in the year following an El Nino. Table 5.2 summarises the ENSO influences by region, and provides a basic forecast relationship. If a significant El Nino (or anti-El Nino) event is occurring or is expected to occur, then the anticipated seasonal tropical cyclone activity and intensity can be qualitatively altered as specified in this table. Care needs to be taken, however, as it is to be expected that some El Nino and anti-El Nino years will not fit these guidelines. The forecasts should be used for general guidance only.

1. Defined on the number of days that a hurricane was present in the basin.

Contents Chapter5.3