El Niño in Australia

E. Linacre, Bart Geerts and Richard Grove


A conference on 27/2/98 at the Australian National University brought together academics and research workers from a wide range of disciplines, mostly from within that university. The following points were made -

1. ENSO and the North Australian monsoon (Robert Wasson)

There is a close correlation between the start of the monsoon in northern Australia (October-December) and the SOI in the previous winter, in June - August. An early start and hence a long wet season happens after a positive Southern Oscillation Index (SOI). The consequent spring rains promote the formation of ‘opal phytoliths and sponge spicules’ on the surface of a freshwater swamp in the Kakadu national Park. This contrasts with the dominance of grasses in El Niño years, when the growing season is shorter. One exception is the '97-'98 El Niño event. The June-August 1997 SOI was at record low values, yet the northern Australian monsoon was slightly wetter than normal. Most of the time, a layer of phytoliths and spicules in a profile of the swamp sediment shows a positive SOI at the time the layer was formed. This allows a rough dating of SOI events over the past 2000 years, including those at about 610 AD, 1300 and 1600, approximately. There is some indication of a periodicity of about 100 years.


2. ENSO during the historical data record (Robert Allan, Neville Nicholls and Ross Cunningham)

Analysis of the rhythms within 179 years of global surface data (i.e. sea-surface temperatures, sea-ice information and global sea-level pressures) shows a tropospheric quasi-biennial oscillation, fluctuations with periods of 2.5 - 7 years (i.e. ENSO) and about ten years. There is no evidence of any connection between ENSO and global warming.

There is a striking correlation between wheat yields in Australia (wheat is harvested in the summer) and the SOI during the previous winter, especially in Queensland. This is due to the relationships between SOI and SST near the continent, between the SST and rainfall, and between rainfall and yield. Wheat harvests are below average after an El Niño. However, the Australian wheat harvest in 1997-'98 was above normal notwithstanding the very strong 1997 El Niño. This is because Australia did not suffer a drought, unlike the last strong El Niño year (1982-'83).

An El Niño is associated with an amplified variation of climate. The consequences of a single strong ENSO event can last a long time, e.g. the present Pillaga Scrub woodlands in southern Queensland, resulting from the El Niño of 1877-8, following a notable La Niña. There have been less obvious La Niña’s since 1980. At the same time SOI values have risen on the whole. The intensity of El Niño’s has become less for any particular SOI. See the following for more details - Vegetatio 91 23--36, Weather 52 66-72, Geophys. Res. Letters 23 3357-60.

Sophisticated statistics applied to data of 1876-1997 show that despite the close correlation between monthly SOI and simultaneous rainfalls at Queanbeyan, near Canberra, for instance, the SOI now is not a useful predictor of the amount of rainfall in the month to come.


3. ENSO during the last millenium (Richard Grove)

There have been remarkable ENSO’s in the past, eg in 1251, 1316-7, 1396, 1405, 1630-1, 1686-8, 1737, 1788-93, 1844-6, 1877-9. They can be related to failures of the monsoons in Madras. (Banks pointed to the bad year for forest fires in Australia in 1790.) The Irish potato famine occurred in 1845. What historians call the ‘17th century crisis’ can be partly explained by ENSO’s, likewise the decline of Great Zimbabwe, collapses of population in Mexico and Java, etc. Such matchings of history and ENSO’s is striking in the southern hemisphere, where there are long records in colonial South America and southern Africa, for example.


4. Corals, ENSO and Climate Change (Michael Gagan, Chantel Alibert, and Linda Ayliffe)

Analysis of the strontium/calcium ratio and the oxygen isotope ratio of a layer of coral gives information about the SST and either the local rainfall or (near the mouth of a river) the amount of estuarine outflow, i.e. the rainfall within the catchment. Fossil corals from the Great Barrier Reef off northeastern Australia allow estimation of temperatures 6000 years ago: conditions were around a degree Celsius warmer than now.

The analysis of corals of the Great Barrier Reef shows the customary cooler winters before the ENSO events of 1965, 1972 & 1982, for instance. There was also an indication of a general warming since 1965 by about 1.3 K, especially since 1979. Such a rise is much larger than the simultaneous rise of southern hemisphere mean temperature, which raises questions about its validity.

Dating a coral layer is like that of tree-rings, but it becomes difficult near the equator, where the annual variation of coral growth rate is less than at higher latitudes. Despite this, analysis of a 30-cm core of coral from 30 km outside the estuary of the Sepik River in Papua New Guinea provides clear evidence of fluctuations of the river’s discharge, i.e. of rainfall in its catchment area inland. There were notable reductions of flow during the El Niño years of 1982, 1987, 1992-4.


5. Tree rings and ENSO (John Banks)

Deducing past climates from tree-rings was first proposed by Leonardo da Vinci. Growth of surviving trees after a forest fire is typically rapid, and the resulting tree ring broad, even when the seasonal rainfall is normal. Such tree ring records imply decades of major fires in the southwestern states of the USA in 1700, 1780, 1790, 1800 and 1840, for instance, which can be correlated with the dates of ENSO events. Extension of tree-ring studies to the southern hemisphere is only just beginning.


6. ENSO in South America and southern Africa (Simon Haberle, Ian Farrington, and Janette Lindesay)

The nature and age of peat deposits near the tree line on a remote island off southern Chile indicate an increase of climate variability some 6000 years ago, after millenia of stability. The variability has been high during the last 2000 years also, notably during the Little Ice Age. Pollen records point to a big change from wet to dry climates about 6500 BP.

Rainfall records from the arid coast of southern Peru indicated that ENSO’s are essenially sporadic and local in their effects. Typically the heavy rains cause landslides. The normal absence of rain leads to a kind of grass/low bush vegetation called ‘lomas’, which relies on moisture from the prevailing fog.

An ENSO brings typically reduces rainfall in the eastern part of South Africa by 24%, while the rainfall increases over Madagascar.