THE CLIMATE KELPIE BLOG: Consensus on forecasts informs decision making

Posted by BCG on 6th April 2021

Climate models can help farmers get a feel for climate directions, but they don’t always agree and this can be a real challenge for on-farm planning.

As NSW grower John Hamparsum says, “Climate is one of the main factors impacting a farming business. A farm is basically a factory without a roof on it and any climate — whether it be hot, cold, wet, dry, or windy — will have an impact on your business.”

“The real challenge is in deciding how much of an influence the models should have on your decisions at any point in time.”

John crops at Breeza on the Liverpool plains in New South Wales with his sister Juanita, wife Nicole and children Ben and Sarah. In summer they grow cotton and sorghum, and in winter, durum and bread wheat and canola.  

John Hamparsum in a canola crop on the Liverpool Plains in NSW.

Looking for consistency

John’s approach is to look for consensus between the different climate models.

“I like to look at several different models. If they are divergent then I have low confidence and they are less likely to influence my decision making. If the model outlooks converge then I have a high confidence in making changes to my on-farm plans.”

John uses both short and long-term models to influence decision making.

In the short term he likes the Meteologixs forecast because of the ability to compare temperature and rain forecasts generated by eight or so different models and also uses the Australian Bureau of Meteorology (BOM) Land and Water forecasts. While they all go through various stages of accuracy, he thinks that the European model, the ECMWF, seems to be the most accurate in recent times and is an excellent website showing three different models including the ECMWF.

“There are lots of short-term decisions where climate information is useful. For instance, we want to know whether there is enough rain coming to risk dry sowing of crops. If we dry sow and the follow-up rain does not arrive, we could lose the seed.

Irrigation timing to counteract the impact of heat and wind on crops is another decision that can be delayed if models are consistently indicating coming rains or cooler conditions.

“We often see one model predicting 60 mm of rain in the coming week, but over time this disappears,” said John. “We really need to see the same consensus from multiple models. For instance, if there is only one or two models predicting heavy rain in the coming week, then I will still schedule my irrigation water. But if they are all consistently forecasting rain then I’m more inclined to wait and see.”

Short-term forecasts are also useful for timing fertiliser or chemical application.

The longer-term

In recent seasons John has lost some faith in the longer-term models.

“They seem to have become less useful than they were in the past. I am not sure if that is because I’m not in direct contact with the BOM forecasters that I used to know and so I’m not aware of the nuances in the system or whether a changing climate is making the forecasts less accurate,” he said.

“Some of the models work some of the time. I’m inclined to believe that each model has its strengths and weaknesses, and the data sets are calibrated to work well with particular climate drivers, but not with others.”

While he does look at the BOM Climate Outlook he is more inclined to put his trust in stored soil moisture. “The summer forecasts have let us down too many times in recent years.”

For instance, after the dry winter in 2020 he chose to fallow 70 per cent of his summer crop as the irrigation demand was likely to exceed supply.

John has two sources of irrigation water – the more reliable groundwater allocation and a licence to pump from the river when it is high flow. In this area single-skip cotton (skipping one row of three) is typically sown on dryland areas and solid cotton (plant every row) on irrigated country. Tweaking the mix is an important part of his risk mitigation strategy.

Skip-row cotton, where one in three rows is not sown, is used in preference to solid cotton irrigation allocations are less reliable

“If we are confident about our water supply we will plant solid cotton on irrigation, but if the forecast is less reliable we’ll increase the area of single-skip just in case we are unable to pump from the river. We might also shift away from cotton and into sorghum, which has lower water requirements.”

John has joined Farmers for Climate Action because he believes that our climate is being impacted by climate change and that we need to be more proactive in this space. He has also installed solar panels to subsidise on-farm energy costs.

John has installed solar panels to subsidise on-farm energy costs.

“I’d like to see more support for research to improve our forecasts. The autumn forecast has never been overly reliable, but we used to be more confident in the September-October forecasts. I think our weather systems are changing and our models need to be changed to improve the predictability of into the future.”


John Hamparsum, 0429 445 899,

Leave a Reply