The declaration of a La Niña event by the Bureau of Meteorology (BoM) in early October alerted many viticulturists and winemakers of an increased chance of a wet summer. Many news articles have pointed to 2011 as the last La Niña evoking memories of extremely wet conditions, widespread disease, shortages of chemical treatments and even flooded vineyards.
However, there have been 30 La Niña summers since 1900 and while most have been wetter than average, very few have been as wet as the 2011 vintage in most Australian wine regions. So, this prompts the question, what are the odds of a wet summer and an extremely wet summer in different regions given we are in a La Niña?
SARDI Climate Applications researchers Dane Thomas and Peter Hayman have delved into detailed climate data from 36 sites – from Kingaroy to Margaret River – to examine the shift in the rainfall odds.
The message from the research is to be alert but not alarmed, as the data shows that only about half of La Niña summers are excessively wet. The 2020–21 growing season won’t necessarily be like 2010–11.
| The 2020–21 season is shaping up to be a good one to check BoM forecasts for rain events in the coming week but also for the outlook over coming months and seasons. The BoM is now issuing forecasts for the coming weeks and months that are updated more often, visit www.bom.gov.au/climate/outlooks/#/overview/summary |
‘A grape grower in the Hunter Valley and Riverland would have different thresholds for an extremely wet summer’, Dane said. ‘Rather than use a rainfall amount, we have used a relative ranking. We took the 120 summers since 1900 and ranked them from driest to the wettest and then divided the ranked years into 10 groups called deciles. Decile 1 is the 1 in 10 driest summer and decile 10 is the 1 in 10 wettest. We then determined which decile of the full 120-year record each of the 30 La Niña summers corresponded to’.
Dane and Peter have represented the deciles as a pie chart with 10 slices coloured with warm colours (red) for the driest (decile 1 and 2) and cool colours (blue) for the wettest (decile 9 and 10). They used orange for drier than average (decile 3 and 4), yellow for the mid-range (decile 5 and 6) and green for above average (decile 7 and 8). The black line in the yellow slice of the pie is the median and is the point that half the years have been wetter and half drier.
If a grapegrower was to ignore the fact that we are in a La Niña, they should use the 120-year historical record to plan for a coming summer to have an equal chance of being wetter or drier. The median line at 6 o’clock represents 50 per cent of years wetter than average and 50 per cent drier. When we determine which decile each of the 30 La Niña summers corresponds to, we find more summers that are wetter than average. An increased liklihood of the seasons being wetter will see the black line move anti-clockwise towards 3 o’clock (a 75 per cent chance of being wetter than median) or even 1 o’clock (more than 90 per cent chance of being wetter than the median).
Before looking at individual sites, Dane and Peter used broader climate regions and split summer into two halves – October to December and January to March – to enable analysis of the timing of impact of La Niña (see Figure 1).
Figure 1: Pie charts showing which deciles of the 120-year record corresponded to the 30 La Niña summers in three BoM regions for October to December and January to March. The numbers in mm indicate the amount of rainfall in the long term record for decile 2, 5 (median) and 8. The percent represents the proportion of years in La Niña summers as dry as decile 2, wetter than the median and wetter than decile 8.
The pies show that for all regions, past La Niña summers are associated with an increased chance of blue and decreased chance of red and orange. The swing to increased chance of wetter conditions is greater in the October to December period than in the January to March period.
‘While there is a shift in the odds in south western Australia, it is less than the more eastern regions. For south eastern Australia, the chance of being wetter than the median in October to December has shifted from 50 per cent to 90 per cent, the chance of being wetter than decile 8 has doubled from 20 per cent to 40 per cent and the chance of being in the driest 20 per cent of years has halved to 10 per cent’, Peter said.
Dane and Peter developed maps (see Figures 2 and 3) to show the pies for individual sites, and specific information on the 36 sites is available here. The largest shift in the odds at any location is 80 per cent; which is less than the 90 per cent for the South Eastern Australia region as a whole, but still represents 24 of the past 30 years above median rainfall. The locations that have shown this shift are Canberra, Rutherglen, Heathcote, Mornington and Launceston.
Figure 2 and 3: Likelihood of different deciles of rainfall for three months October – December (upper panel) and January – March
Although the pies show increased chance of blue, the orange and red are not eliminated.
‘Across the periods and regions, the chance of being wetter than the median ranges from about 60 per cent to 80 per cent. Another way of expressing this is that the chance of being drier than median in La Niña years ranges from 20 per cent to 40 per cent’, Dane said.
Knowing that we are in a La Niña increases the chances of being wetter than the median, but there is no guarantee.
‘If the wettest 20 per cent of years (dark blue) are taken as problematic for disease management, there are many sites where the chance has doubled from 20 per cent to 40 per cent and some sites where this is as high as 47 per cent’, Peter said.
‘This is a major revision of the odds. However, when reading the media discussion about the coming wet summer it is important to note that more than 50 per cent of past La Niña summers have not been excessively wet.’
Want to know more about La Niña? Check out the Climate Kelpie website here.