Researchers probe warming climate frost puzzle

GroundCover Live and online, stay up to date with daily grains industry news online, click here to read more

An increase in the number of frosts and a later frost season means more and more growers are having to reassess their sowing dates and varieties

Image of a man standing in field taking temperature readings

CSIRO’s Dr Steven Crimp downloads temperature
and relative humidity data from loggers on a farm near
Kyabram in northern Victoria.

It is a cruel paradox that although Australia’s climate is warming, the number of frost days and the length of the frost season are increasing across much of the Australian grainbelt.

CSIRO climate application scientist Dr Steven Crimp has been studying the incidence of frosts across the grainbelt between 1960 and 2011. “Observational data shows that the incidence of frosts has increased in Australia and there have been shifts in frost occurrence to later in the year,” he says.

“So even though we have a warming trend, we have this unexplained change in frost risk. In the east, the window of frost occurrence has broadened, so frosts are occurring earlier in the season and much later in the season. As we move to the west there is less occurrence of earlier frosts and it is more of a shift to frosts later into the season.”

The frost window has lengthened by three weeks in the Victorian grainbelt and by two weeks in the New South Wales grainbelt. Western Australia has, statistically, remained the same, while eastern South Australian sites are similar to Victoria, and sites in the west of SA are more like WA.

Dr Crimp says frost risk is estimated by using Bureau of Meteorology (BoM) temperature data, using a threshold of 2°C. “A measurement of 2°C, registered in the standard Stevenson screen, which is at head height, amounts to a surface temperature of around zero, or freezing, at ground level.”

His team found the frost window over much of northern Victoria had lengthened considerably in the decade to 2011. “If you look at the risk of experiencing a 2°C minimum temperature event, the 10 per cent risk now occurs 46 days later than in any of the previous decades. The frost window over this past decade has been much wider than farmers have experienced before.”

The consequences of a late frost for grain growers can, of course, be devastating. “A grain crop is particularly sensitive at grain fill, and there has been anything from a 40 to 80 per cent loss when frosts occur.”

Dr Crimp and his colleagues conducted a detailed simulation of frost risk in 17 Victorian locations.

“Northern Victoria seems to be the epicentre of this change in frost occurrence. We used our model to calculate across the entire record of climate that we have, from 1960 to the present.”

Figure 1 Frost days index per time period August to November
trend; 1961 to 2010

Meteorological map showing frost days index per time period August to November trend; 1961 to 2010

Trends in the number of frost events (temperatures below 2°C) for
the period August to November (1961 to 2010) based on the BoM
high-quality gridded minimum temperature dataset.
Regions coloured in blue represent areas of increasing frost number
and areas depicted in red indicate declining frost number.
The legend values represent the change in number of events per
month, per year.


Because growers can manage frost risk by changing the planting date and using faster or slower growing varieties, the researchers have sought to inform these decisions by examining the interaction between sowing and flowering dates for three widely used wheat varieties in northern Victoria: the early flowering AxePBR logo, mid-season variety YitpiPBR logo and the late-maturing Rosella.

Using APSIM (the crop model behind Yield Prophet®) they ‘planted’ each variety each week from 1 April (day 90) to 15 July for the period 1960 to 2011. They calculated a mean flowering date for each sowing date for the three varieties across the 17 locations. The time of flowering was then graphed against the risk of frost.

For the Birchip site, Dr Crimp said 90 per cent of the frosts occurred at, or before, day 240 (27 August).

“Ten per cent of frosts occur after day 290 (16 October). If I am a risk-averse farmer, and I want to avoid all but 10 per cent of the frosts that may occur, I would have to plant on day 170 (18 June) for early maturing, day 160 (8 June) for medium, and day 150 (29 May) for late.”

Dr Crimp notes that many growers prefer to dry sow on a certain calendar date, which he says could make those crops more vulnerable to frost risk if sowing dates do not take into consideration the changing frost risk.

His team found decadal differences. For example in 2001 to 2011 frost days occurred between 20 to 46 days later than in the 1950s.

Dr Crimp says many people thought the increase in frosts was due to dry conditions, but the past decade has included some very wet years.

To try to find an explanation, the research project has used hourly, three-hourly and daily climate data to look for links between synoptic patterns, or for large-scale climate patterns that can be associated with frost events.

Dr Crimp says researchers’ attention is focused on the subtropical ridge*, its change in intensity and displacement south, as well as the increase in the frequency and intensity of blocking highs.

“As these synoptic features are displaced further south, they bring masses of very, very cold air over the country.” He says this so-called ‘southward displacement’ is what researchers think is driving these changes, particularly for the later frost occurrences.

Next, Dr Crimp and his colleagues want to look at whether the southward displacement is a temporary change or whether it is likely to continue. They also hope to be able to build a forecasting system.

More information:

Dr Steven Crimp
02 6246 4095

Steven Crimp, Mike Pook, David Gobbet, Nirav Khimashia, Alison Laing and Uday Nidumolu, Understanding frost risk in a variable and changing climate

*The subtropical ridge is a belt of high pressure situated around 30 degrees latitude. It is responsible for trade winds and the westerlies. It can be displaced by the ENSO climate cycle, with a subsequent impact on monsoon regimes.

Next: Pre-breeding taken to the paddock

GRDC Project Code MCV00010

Region National, South, West, North