Australian rainfall during El Niño and La Niña events

Summary

Following the breakdown of the 2020–2023 'triple' La Niña in March 2023, there was a brief period of ENSO-neutral that transitioned to an El Niño. Dry conditions emerged across most of Australia during late-winter and spring 2023, followed by a wetter than average summer during 2023 to 2024.

In June 2023, sea surface temperatures (SSTs) reached El Niño thresholds in the Niño3.4 region and remained above the thresholds until March 2024. Globally, SSTs were at record warm levels in 2023, with the pattern of warmth differing to historical patterns of SSTs associated with ENSO. Atmospheric coupling did not occur until September 2023, when the Southern Oscillation Index (SOI) reached values consistent with El Niño. There was a weak atmospheric response for much of the 2023–2024 El Niño, evident in fluctuating cloudiness at the International Date Line and near-average Pacific trade wind strength.

Atmospheric indicators of El Niño returned to neutral from mid-December 2023 while SSTs in the tropical Pacific persisted at El Niño levels until April 2024.

Rainfall

The beginning of the 2023 El Niño was characterised by dry conditions across Australia. This coincided with a positive Indian Ocean Dipole (IOD) developing in the Indian Ocean and becoming active in late August. August to October was the driest 3-month period on record since 1900 (Figure 1), aligning with the development of the El Niño and the positive IOD. September was the second-driest month on record, behind April 1902. Spring was drier than average across large parts of eastern and southern Australia, the period when El Niño typically has the greatest influence on rainfall.

In spring 2023, a relatively large and long-lasting Antarctic ozone hole contributed to a strong polar vortex, which in turn supported a positive phase of the Southern Annular Mode (SAM), particularly during November and December. The positive SAM favoured onshore flow for eastern Australia, which, in addition to warmer than average coastal SSTs, generated widespread thunderstorms, many severe, during this period. The concurrence of El Niño and a positive SAM is rare.

Across most of Australia, except Tasmania and the southern mainland coast, rainfall was above average from November 2023 to March 2024 (Figure 2). This above average rainfall was more likely due to SAM than El Niño as the atmospheric component the El Niño weakened over the 2023–2024 summer. Furthermore, El Niño typically has less influence on Australian rainfall in summer than it does in spring.

While climate influences such as ENSO and IOD can contribute to large-scale shifts in Australia's monthly to seasonal rainfall and temperature patterns, at times, short-term variability, such as those from major weather events, have a greater influence on Australian weather, as was the case with a range of major weather events between December 2023 and March 2024. These included severe storms impacting parts of eastern Queensland and New South Wales in December 2023, and from January to March 2024, three tropical cyclones (Kirrily, Lincoln and Megan) and their associated monsoon troughs that brought widespread heavy rain across much of northern Australia. Nationally, it was the ninth-wettest January on record and the third-wettest March on record (Figure 3).

Other impacts

Conditions were generally warm across Australia during the 2023–2024 El Niño. 2023 was Australia's warmest winter–spring since records began in 1910, and the third-warmest summer on record (behind 2018–2019 and 2019–2020). The 2023–2024 El Niño coincided with record-warm oceans globally and warm SSTs that persisted across the western Pacific. Globally, 2023 was the warmest year for land and ocean temperatures since global records began in 1850. These unique conditions, unlike many strong El Niño events in the past, likely contributed to the above-average rainfall across much of Australia during November 2023 and March 2024. Record-warm ocean temperatures during the summer and autumn months led to extensive coral bleaching in the Great Barrier Reef, with the south of the reef most severely impacted. The impact of future El Niño events on Australian rainfall as the climate continues to warm is the subject of ongoing research.

The warm and dry conditions in spring 2023, combined with stronger than average south-easterly trade winds, saw major bushfires across northern Australia, particularly in the Northern Territory and Queensland. These regions had seen significant vegetation growth during the 2020–2023 La Niña, increasing available fuel.

The overall effect of this El Niño on Australia was weak to moderate, with the 13 months from April 2015 to April 2016 (Figure 1) resulting in widespread areas of below average rainfall across central to southern Queensland, southeast South Australia, Victoria, and Tasmania. Dry conditions were already in place in Queensland and southeastern Australia in the lead up to this El Niño.

After sea surface temperatures in the central tropical Pacific Ocean came close to El Niño levels in 2014, the central tropical Pacific Ocean was primed for El Niño in 2015. Westerly wind events in January and March led to rapid ocean warming during autumn 2015, reaching El Niño levels during mid-April. Further westerly wind bursts consolidated the event in May and June. Rainfall was below average over parts of eastern Australia during April to August 2015 (Figure 2), although very warm sea surface temperatures (SSTs) in the eastern Indian Ocean may have offset some of the drying effect of the El Niño pattern in the Pacific. Warmer waters in the eastern Indian Ocean may provide extra moisture for rain-bearing systems as they cross Australia.

Between late August and mid-November a positive Indian Ocean Dipole (IOD) developed, further suppressing rainfall over much of Australia. The combination of the positive IOD and El Niño saw widespread below average rainfall during September and October 2015 (Figure 3). Rainfall was the equal fourth-lowest on record for Australia during September, and Tasmania had its driest spring on record. Mean temperatures for Australia were also highest on record for October to December 2015.

Below average rainfall persisted over much of eastern and northern Australia during the first four months of 2016 (Figure 4), with large parts of the northern coastline recording rainfall in the lowest 10% records. The event finally broke down during May 2016.

Peak values of central tropical Pacific SSTs and NINO indices exceeded +2 °C, placing this event alongside 1982 and 1998 as one of the strongest events on record. NINO3.4 reached +2.5 °C in November 2015.

This very strong El Niño contributed to an early start to the 2015–16 southen fire season, a markedly drier than average northern wet season, the weakest and least active tropical cyclone season on record (record low 3 TCs in 2015–16 season, none reaching category 3 strength), and prolonged heatwaves across Australia during autumn 2016. As the event started to break down very warm ocean water was pushed towards Australia. Combined with a delayed monsoon and cloudless skies this resulted in the worst coral bleaching event on record for the Great Barrier Reef. For more information on the effects of this El Niño see Climate Update El Niño is over, but has left its mark across the world.

The overall effect of this El Niño was weak, with the 11 months from May 2009 to March 2010 (Figure 1) resulting in widespread areas of below average rainfall across WA alone. In fact, the Australian rainfall anomaly pattern was close to a complete reversal of that usually observed.

The initial phase of the El Niño had the largest negative impact, when the SOI signal was rather weak. Consistent with the typical El Niño impact on Australia, May to October 2009 (Figure 2) was rather dry over much of the country, with Queensland and the NT having large areas with rainfall in the driest 10% of the historical record. Further south in areas that expect winter and spring rains, the impact was not quite so marked. Eastern Victoria and most of NSW, except the northeast and north central areas had below average rainfall for the period, while western Victoria, southern SA, Tasmania and northeast and north central NSW had average to above average rainfall.

By November, the SOI signal strengthened and finally came into line with other major ENSO indicators, most notably the SST. Paradoxically this ushered in a wet period over the eastern half of the country. For the 5 months from November 2009 to March 2010 (Figure 3), the NT, SA, Queensland, NSW and Victoria all had areas of rainfall in the top 10% (decile 10). Being generally located inland of the Great Dividing Range, these decile 10 regions included some small parts of record high falls. Particularly active monsoon conditions occurred in February and March when a southward surge of tropical air brought heavy to flood rains across central and southern parts of the NT, southern Queensland and northeast NSW. Remnant moisture also produced heavy rains across southern NSW and Victoria. This period included the Melbourne Hailstorm on the 6th of March 2010.

So Australia's rainfall patterns had switched from being typical of El Niño well before the main ENSO indicators had shown strong signs of retreating to neutral values, although SSTs had been in slow decline since the end of December. A more emphatic sign was the 25.8 rise in the SOI from March to April 2010 which heralded the end of the event as far as broadscale indicators were concerned.

During May 2006 to December 2006 (Figure 1), most of Australia was strongly affected by this weak El Niño, with large regions in the southern and coastal areas being in the lowest 10% of rainfall. Particularly, southern Victoria and northern Tasmania were very dry, receiving lowest on record falls.

These 8 dry months had a very strong impact when factoring in the lack of relief from the 2002/03 El Niño. This was especially obvious in the 2006/07 fire season, with the Great Divide Fires being the longest running bushfires in Victoria's fire history. They began by lightning strikes in December 2006, and ended in February 2007. These fires caused the worst bushfire smoke since records began, affecting those with respiratory conditions such as asthma. Furthermore there were widespread fires in New South Wales during this time.

A change in conditions occurred around January 2007, with January to May 2007 (Figure 2) having a return to average to above average rainfall for most areas. However the Melbourne region and southeast Queensland were exceptions, being plagued by ongoing dry conditions.

This weak to moderate El Niño event had a very strong impact in Australia. The major 2002-03 drought had rainfall deficiencies over the period March 2002 to January 2003 (Figure 1) that ranked in severity and areal extent with the extreme droughts of 1902 and 1982-83. Practically all parts of the country were affected, and in southern areas this exacerbated the effects of several preceding years of dry conditions. The extreme dryness coincided with exceptionally warm conditions: maximum temperatures averaged Australia established new records in each of the seasons autumn, winter and spring by a wide margin for the post-1950 era. Severe bushfires in eastern NSW, Canberra, and the mountains of southeast NSW and eastern Victoria, and widespread water shortages, were some of the main effects. Widespread above to very much above average falls occurred in February 2003 raising hopes of a consistent period of wet weather to erase the effects of severe drought. However, this was not to be. Totals for the remainder of 2003 were insufficient in many areas to overcome existing rainfall deficiencies - especially in parts of Queensland and southeast Victoria where 2003 was another rather dry year (Figure 2).

Over the 12 months from April 1997 to March 1998 (Figure 1), below average totals occurred across Victoria, east and north Tasmania, central & east NSW, eastern sub-tropical Queensland & southwest WA. Overall the impact would be classed as weak with crops benefiting from widespread falls in both May and September. Southern and eastern Victoria were the worst affected with much of the region having falls among the driest 10% on record for the 12 months. Rainfall was consistently above average in eastern Australia from April 1998 onwards.

Over the 10 months from March to December 1994 (Figure 1), rainfall was in the lowest 10% of totals across most of Queensland, NSW, Victoria, SA, NE Tasmania and southern WA. Falls were also below average over the southern half of the NT and much of WA. So the impact would be described as strong. Average to above average falls fell in parts of eastern Australia (especially NSW) in November and December, and then very heavy rain and flooding occurred over inland NSW, SW Queensland and northern Victoria in January effectively ending the event. However, there was another period of below average rain from February to April 1995 (Figure 2) before higher falls became re-established.

Not a "classic" event by any stretch of the imagination. Despite the SOI staying in moderate negative values until October, the ocean signal was weak except for April and May 1993. The effect was very weak with above average falls occurring over eastern Australia in July, October, November, and December 1993, and then again in February 1994. In fact, it was so wet that large parts of NSW and Victoria recorded totals in the highest 10% of records for the 6 months from July to December 1993 (Figure 1). Against this though, it was a poor wet season in eastern and northern Queensland, particularly the 3 months from December to February (Figure 2).

A strong effect over Queensland with about three-quarters of the state recording decile 1 rainfall totals for the 9 months from March to November 1991 (Figure 1). It was the driest such period on record in parts of the Darling Downs. The northern half of NSW was also seriously affected with about half this region in decile 1 for the nine months. Further south totals were average to above, although the seasonal distribution was very uneven. In Victoria, southern NSW, and much of Tasmania, a very dry autumn was followed by a wet June-September period and then more below average falls in October-November. Above to very much above average rain fell in December over SE Queensland and the eastern half of NSW marking the beginning of the end of this event in these regions. Although falls were patchy in January 1992, heavy rain fell again in February, particularly in the southeast of Queensland and NSW. In the tropics though, totals were consistently below average from November through to April (Figure 2) with many areas of WA, the NT, and north Queensland recording decile 1 falls for the 6 months.

Generally weak with mainly average to above average falls in Victoria, NSW and Queensland during the 9 months from May 1987 to January 1988 (Figure 1). Northern and eastern Tasmania, pockets of Gippsland and the southwest corner of WA fared rather worse with totals in the lowest 10% for the 7 months from April to October 1987 (Figure 2). It was also a poor wet season over Cape York and northwest Queensland with decile 1 totals for the 6 months from October 1987 to March 1988 (Figure 3).

A very strong impact on Australia with drought conditions widepsread across eastern and southern Australia. Below average rainfall patterns were established in April 1982 and continued almost unabated up to and including February 1983 when southern Australia experienced heat-wave conditions and bushfires that culminated in the Ash Wednesday disaster. For the 11 months from April 82 to February 83 (Figure 1) virtually all of the eastern two-thirds of the country recorded rainfall totals within the driest 10%. In fact, the vast bulk of Victoria, the southern halves of both NSW and South Australia, together with large tracts of central and western Queensland recorded record low falls for this particular 11-month period.

The pattern changed abruptly in March 1983 when flood rains in central and southern Australia heralded several months of above average rainfall across much of the country.

A moderate to strong impact, but of fairly short duration. For the six months from June to November (Figure 1), most of NSW, the southern half of Queensland, patches in northern Victoria and northern Tasmania, and scattered regions in both WA and SA recorded rainfall totals in the driest 10% for that 6-month period. Most remaining areas in these states, with the exception of SA, registered below average falls. Above average rain in January eased the situation in eastern Australia, but for the nine months from June 77 to February 78 (Figure 2), decile 1 totals were registered in northern Victoria, southern and far western NSW together with the northeast corner of that state, and SE Queensland.

Drier than average weather became established in the southern areas in March 1972, and apart from a brief respite in August, continued to the end of the year. Overall the impact would be classed as strong. For the ten months from March to December 1972 (Figure 1), rainfall totals were in the driest 10% of records (decile 1) over the majority of Victoria, SA, southern WA, north and east Tasmania, western and southern NSW, and far southwest and central Queensland.

Above average falls occurred in January, followed by widepsread above average to record falls in February, thereby ending the drought and heralding the wettest two-year period in Australia's history.

In this case, weak indicators lead to a weak effect through the region of eastern Australia most commonly affected. There were some dry months between June 1969 and February 1970, Figure 1, (notably June and October in Victoria), but for this 9-month period decile 1 totals were mainly confined to some of the border regions between SA and each of Victoria, NSW and Queensland. Parts of the Eyre Peninsula and southwest WA also had totals in the driest 10% of falls.

However there was a strong impact across most remaining parts of WA together with some adjacent regions in eastern NT. These areas recorded decile 1 or record low falls for the nine months from June '69 to Feb '70, but it is more than likely that an additional influence unrelated to El Niño was crucial to the development of this rainfall anomaly.

The overall impact was moderate. Over the 9-month period from March to November 1965 (Figure 1), rainfall was below average across most of NSW and SA, the southern half of Queensland, southern NT and patches in both Victoria and Tasmania. The most seriously affected region was a large area straddling the NSW/Queensland border stretching from near Longreach in the north to Dubbo in the south. Totals here were in decile 1 for the nine months with a few patches of lowest on record.

Above average totals occurred in August over SA and southern NSW, but it wasn't until December 1965 that some relief came for the worst affected area with above to very much above average falls. However, the relief was short-lived as conditions were generally drier than average right up until July 1966 causing a large part of NSW and southern Queensland to be in decile 1 for the 17 months (March 1965 to July 1966), Figure 2. This was depsite a rise in the SOI and a return to neutral Pacific sea-surface temperatures, thus suggesting a mechanism apart from El Niño.

This event seemed to develop a little later than is typical, with the main period of dryness being over the relatively short period from September 1963 to March 1964 (Figure 1). During this 7-month period the impact was weak across the Murray-Darling Basin, the most commonly affected area, but strong over regions further west. However, because this affected region was unusual, there is a strong probability that other influences helped to cause the anomalously low rainfall.

These seven months were exceptionally dry across SA, the southern half of the NT, eastern WA and the western fringes of Victoria and NSW. Large parts of western SA and southeast WA had record low falls for the period.

The overall impact was moderate to strong with most of the southern half of the continent being drier than average, including extensive parts of NSW, SA and WA that were in decile 1 for the nine months from March to November 1957 (Figure 1). Parts of central NSW and WA registered their lowest rainfall on record for this 9-month period.

May and the spring months were particularly dry, but falls were more patchy in winter with some above average regions in NSW and Victoria. Tasmania and western Victoria though, were drier than average in winter with record low falls in the Apple-Isle.

The overall impact was strong with most of the northern two-thirds of the country recording below to very much below average rainfall. For the 12 months from March 1951 to February 1952 (Figure 1), rainfall totals were in decile 1 over nearly all of Queensland, the northern half of the NT, as well as scattered parts of northern NSW, SA and WA. Large areas in northern Queensland and the NT had record low falls for this particular 12-month period.

Rainfall was average to above average in Victoria and southern SA, particularly between April and August (Figure 2) when 5-month totals were in decile 10 or the highest on record.

A strong impact over eastern and northern Australia with most of NSW, Queensland, and the north of the NT registering decile 1 falls for the 10 months from April 1946 to January 1947 (Figure 1). A large region in the central-west of Queensland experienced record low totals for this particular 10-month period. Large parts of SA, Victoria, and WA were also drier than average.

Areas of below average rainfall first became evident over Queensland in autumn and gradually spread southwards. However, there were some months when the dry spell was eased by useful rains - most notably across Victoria in July and the NSW/Queensland border areas in September. Average to above average falls were also prominent in November and December.

The pattern was broken in February 1947 when widespread heavy rain fell in the worst affected areas. Good follow-up falls also occurred in March.

The overall impact was moderate to strong, particularly in NSW which was mainly in decile 1 for the 10 months from April 1941 to January 1942 (Figure 1). The southern tier of Queensland was similarly affected, as was Cape York in the far north of the State. The east and north of both Victoria and Tasmania were drier than average during this 10-month period with decile 1 falls in some areas.

The pattern began to change in February 1942 with a return to average falls or higher for the period up to mid-winter, although April 1942 was a dry month across the eastern half of NSW.

The overall impact was strong, although the period that was characterised by anomalously low rainfall was a little shorter than that which occurred with many other events. For the nine months from March to November 1940 (Figure 1) inclusive, most of the country recorded below average rainfall, the main exceptions being parts of eastern and northern Queensland and some areas in the NT. There were vast areas where falls were in decile 1, and this includes the above average rainfall that fell in April in NSW, Victoria and central Australia. For the shorter 7-month period beginning in May (Figure 2), one-half to two-thirds of the country recorded decile 1 rainfall totals.

Above average rainfall in eastern NSW and southeast Queensland in December heralded a change in the pattern. January 1941 was an excpetionally wet month over eastern Australia with decile 10 falls covering nearly all of Victoria and NSW, much of the southern half of Queensland and the eastern half of SA. Highest on record totals occurred in each state. March 1941 was another wet month and above average falls continued in the tropics until June.

The impact would be classed as strong, but the main area affected was somewhat to the west of that most commonly influenced. For the 12 months from March 1925 to February 1926 (Figure 1), most of eastern and central Australia had below average falls with decile 1 to lowest on record values across most of Tasmania, Victoria, northern SA, southern NT and far western Queensland.

There was quite a deal of month to month variation in the rainfall patterns with May and June 1925 being wet or very wet over parts of eastern and central NSW. Other useful falls to occur during the period occurred as follows: July 1925 in East Gippsland, August 1925 in southeast Queensland, September 1925 in southeast SA and November 1925 across northern NSW. October, December and February were particularly dry over large areas.

With the exception of Tasmania and western Victoria, the dry weather pattern was broken by widespread and heavy rain in March and April 1926.

A classic event in terms of impact with strong rainfall deficits over the Murray-Darling Basin, Tasmania, western Victoria and the southeast of SA for the 15 months from March 1919 to May 1920 (Figure 1). Most of these areas recorded falls in decile 1 with areas of lowest on record featuring in central NSW. Although there were pockets that received decent falls from time to time during this period, the only widespread falls occurred in May and December. The driest months were June, July, October, November and February.

The SOI had been negative since late 1918 rising to around zero during the early part of 1920. However, it wasn't until the SOI rose to around +10 in June and July that the drought was eased or broken with widepsread above average falls.

This was the second year of a double El Niño event and the effect was strong. The SOI was generally in the −15 to −20 range from April 1914 to May 1915, and apart from coastal NSW virtually all of eastern Australia, together with the east of SA and southwest WA recorded below average rainfall. For the 12 months from May 1914 to April 1915 (Figure 1) decile 1 or lowest on record totals were particularly prominent in Tasmania, Victoria, southeast SA and southern NSW.

Dry weather set in with a vengeance in southeastern Australia in June and continued almost unabated until the end of October. The pattern spread northward into Queensland in August, although much of Victoria, NSW and Queensland had some relief in November and December with average to above average rains. The central and north coasts of NSW were very wet in September and October during which time record dry conditions occurred over Victoria and Tasmania.

The pattern of below average rainfall returned in January 1915 and continued to affect most of eastern Australia until the end of April. From May onwards the pattern reverted to one of average to above average falls that eventually led into the double La Niña event of 1916/1917.

A weak to moderate effect with most of southeastern Australia and parts of southern Queensland recording below average rainfall for the June 1913 to February 1914 period (Figure 1). Areas of decile 1 totals tended to be rather patchy, the largest covering parts of Gippsland, northeast Victoria and adjacent areas of the southern inland of NSW. The below average rainfall pattern took a while to become established with wet conditions prevailing over much of NSW between March and June 1913. However, June 1913 was very dry in far western NSW, SA and western Victoria and this pattern spread eastward in July. There were average to above average falls in September and October 1913, but rainfall tended to be below normal for the four months following this.

This event took a little while to get going in terms of rainfall anomalies, but for the 14 months from April 1911 to May 1912 (Figure 1) the overall impact was moderate to strong. For this period falls were in decile 1 over much of eastern Queensland, northern NSW, northern Victoria and the southern border areas of NSW, western WA and parts of both southeast SA and Tasmania. In parts of central Queensland, it was the driest such period on record.

Occasional good falls happened during the early and middle parts of the event with November and December 1911 being particularly wet across central NSW. However, January, April and May 1912 were dry over large areas. This pattern was broken in June and July 1912 with widespread above average rainfall.

Despite very strongly negative SOI values in autumn, it wasn't until winter 1905 that drier than average areas began to emerge in eastern Australia. The effect was relatively short-lived with a moderate impact over the 8 months from June 1905 to January 1906 (Figure 1). During this period most parts of eastern and northern Australia recorded below average rainfall, with decile 1 areas scattered throughout. Below average falls also occurred across much of SA and the west of WA.

The driest period during the event over eastern and southern Australia was from November to January, but the pattern changed in February with above average to record falls occurring over southern Queensland and western NSW. High totals were widespread in the east of the country in March, although the monsoon remained weak over the far north. There was a brief return to widespread dry conditions in April.

This was the culmination of the Federation Drought with a very strong impact over the eastern half of the country where most rainfall totals for the 12 months from November 1901 to October 1902 (Figure 1) were in the lowest 10% of records. Vast areas of Queensland, NSW and central Australia had record low falls. The southwest of WA was also drier than average.

Negative SOI values occurred from September to December 1901, but apart from February 1902, the next negative monthly SOI value wasn't observed until August 1902. So it was an odd event in this respect.

Areas of above average rainfall were observed in November 1902 over parts of SA, southern NT, western Queensland and the far northwest of NSW heralding further widespread falls in December across southern and eastern Australia. The relief was shortlived though with more dry conditions in January and February 1903. It wasn't until March 1903 that a consistent pattern of above average falls was established.