THE CLIMATE KELPIE BLOG: What do forecasts really mean?

Posted by BCG on 17th December 2020

Forecasting weather and climate is a tricky business. The four main climate drivers — the El Niño-Southern Oscillation (ENSO), the Indian Ocean Dipole (IOD), the Madden-Julian Oscillation (MJO) and the Southern Annular Mode (SAM) — interact in complex ways. Their influence drives local climate and weather outcomes in different ways across Australia.

Because of the many, variable ways these drivers can interact to create the climatic patterns and day-to-day weather we experience, forecasting the exact outcome can be difficult. That is why forecasts provide for a range of possible outcomes.

Dr Luke Shelley from the Bureau of Meteorology Agriculture Program said that while today’s forecasts aren’t perfect, they have improved substantially over the last 40 years.

“Today, the Bureau’s weather forecasts for the next four days are as good as the one-day forecasts generated in the 1980s, and they’re improving all the time.

Photo by Michael Weidner on Unsplash.

Weather forecasts

The Bureau describes weather forecasts as being for the next seven days, whereas climate outlooks predict the next one to six months.

“Our weather forecasts may be deterministic — for example, defining temperatures for the coming week — but they may also present a range of possibilities, such as a rainfall range,” Luke said.

“Say you have a weather forecast that says ‘90 per cent chance of rain, 3-8mm’. That forecast has two pieces of information (see Figure 1). It is saying there is a 90 per cent chance that there will be measurable rain, and on our equipment ‘measurable rain’ means 0.2mm or above;. The other numbers given – the 3-8mm – are also based on probability. The first amount – in this case the 3mm – is the 50th percentile. That is, there is a 50 per cent chance that there will be at least 3mm rain. The second number – the 8mm – is the 25th percentile. That is, there is a 25 per cent chance there will be 8mm rain”.

“Putting all that together, ‘90 per cent chance of rain, 3-8mm’ means there is a 90 per cent chance that there will be measurable rain (0.2mm or more), a 50 per cent chance there will be 3mm or more, and a 25 per cent chance there will be 8mm or more”.

“But there is also a 10 per cent chance that there will be no rain at all. If it doesn’t rain, it doesn’t mean the forecast was wrong, but we do measure how reliable our forecasts are over time. It’s all about probability. When we forecast, we forecast patterns, and the rainfall that can result from those patterns is highly variable over time and space, making it very difficult to make a perfect prediction. That’s why we present probabilities, so that our customers can understand the range of possibilities they can expect.”

Figure 1. The Bureau of Meteorology rainfall forecasts provide information on both the chance of any rain, and the probabilities of how much rain. Source: BOM Weather app.

Climate outlooks

The climate outlooks are presented as probabilities of certain events based on observed and modelled patterns.

“From today forward, it could be drier or wetter. If you think of the red squares on a roulette wheel as representing it being wetter and the black ones as being drier, it’s pretty obvious that you throw the ball onto the wheel and it could land on red or black; it could be wetter or drier. However, specific patterns that influence Australia’s climate can push the odds one way or the other. If the conditions are such that it is more likely to rain than not, you can think of that like there being more red squares. But it doesn’t mean that there’s no chance of it being dry – there are still black squares there and the ball could land on them; it’s just less likely. That’s what seasonal climate forecasting is like.

“The forecasting at BOM is very sophisticated, and it is a leader in international best practice in communicating forecasts as probabilities. Forecasts always have uncertainties — our challenge is to help communicate our forecasts in more meaningful ways to support decision making.”

The BoM provides a customisable tool to assess the past accuracy of seasonal outlook forecasts for minimum and maximum temperature and rainfall in different locations at different times of the year.

Boosting agriculture

The BOM is totally independent with no commercial interest in the forecasts it provides, but is working to increase the profitability of the agriculture sector.

“Our strategy in the Bureau’s Agriculture Program is to make agricultural businesses more profitable,” Luke said. “We aim to deliver $300 million worth of value by 2022; that’s not about investing $300 million in our products, it’s about making Australian agriculture that much more profitable by providing better information to help improve decision making.

“We are having conversations with farmers and strengthening our relationships with industry bodies because that is how we will get better at differentiating our products to make them more useful to the agriculture sector.”

Luke said that while the trend is towards providing more and more information the real challenge was to package it up in meaningful ways that can support on-farm decision making.

“The best way farmers can use weather and climate outlooks to their advantage is in combination with a whole range of factors in their business,” he said. “The key is to understanding your own sense of what a forecast means to you in the context of your own business situation. It’s about understanding risk and impact and using the information to make a decision that best fits with your business plan.”

On the ground

Sam Simons, an agronomist with Poole Ag based at Moree in NSW, says farmers in the area can make significant changes to their cropping program based on seasonal outlooks.

“The much-anticipated development of a La Niña in late 2020 gave some growers the confidence to increase their summer crop in both rainfed dryland and irrigated systems,” he said. “The positive rainfall outlook feeds that appetite for risk. For instance, some growers were willing to plant sorghum and cotton on less stored soil moisture than they normally would given the optimistic La Niña outlook.”

“We had an unexpected dry spell in November that was ideal for the winter crop harvest but created a real challenge for summer crops like sorghum that were at a critical stage trying to fill grain during heat waves with a drying soil profile.”

“We need to remember that the outlooks are probabilities, not guarantees. Nothing is certain in climate.”

More information

Bureau of Meteorology resources

This article was adapted from the Meat & Livestock Australia’s Feedback Magazine.

Leave a Reply