Understanding frost risk by monitoring on-farm temperatureStudy results
GroundCover™ Issue: 49
By Kirrily Condon, Manager – Research & communication, FarmLink Research
Frost is an ever-present problem for graingrowers in southern NSW, and one which is difficult to measure in terms of actual lost productivity.
To try and measure the problem more accurately, and find possible management options, a temperature study was established as part of the GRDC funded project ‘Tools to reduce the impact of climate variability in south-eastern Australia’ (involving NSW Agriculture and DPI Victoria).
The study examined temperature variation across paddocks and how these temperatures compared with the nearest Bureau of Meteorology (BoM) station. If a comparison could be made, historical temperature data could be used to look at frost risk over time.
Ten data loggers were placed in each of four paddocks at Junee Reefs, Burrumbuttock, Ardlethan and Rankins Springs. Elevations within the paddock ranged by 33 metres at Burrumbuttock to just 1.5m at Ardlethan. Temperatures were downloaded monthly from July to November 2003.
Unfortunately, it was a really good year to do the monitoring – because there were severe late frosts. Temperatures varied markedly within paddocks according to elevation, soil type and cold air drainage.
At Junee Reefs, minimum temperatures on 10 August varied in the paddock by 5°C from the coldest point (-7.8°C) to the warmest point (-2.7°C), over a change in elevation of 25m. The difference in flowering time between these two points was six days.
A common result from the monitoring was that at three of the sites, the median difference in temperature between the coldest point in the paddock and the nearest BoM station, was three degrees (for frosts of at least -3.5°C in Sept/Oct).
If we assume a significant paddock frost is -3°C (the temperature at which damage is likely to occur in wheat is -3.5°C), this would equate to 0°C recorded at the nearest BoM station.
The number of times temperatures of 0°C or below have been recorded is very infrequent, suggesting 2003 was a very unkind anomaly.
For more information:
Kirrily Condon, 02 6978 0428; firstname.lastname@example.org
GRDC Research Code: DAV 00006, program 4
Based on historical temperature records, the frequency of significant frost events is low in the areas studied. However when they do occur, as the study revealed, the impact is significant. Late sowing to delay flowering has been a common management option used to avoid damage from late frosts.
However, optimum flowering time is a compromise between flowering too early and suffering yield loss from frost, or flowering late and suffering yield loss from heat stress and high evaporation.
Based on the historical frequency of frost events, it is suggested that the number of years a yield advantage could be achieved by sowing early (in the optimum sowing window) would outweigh the number of years an advantage could be made by sowing late to avoid frost. As the severe frost events of 1998, 2001 and 2003 showed, sowing time had little influence on the amount of damage.
Management practices for high frost areas In frost-prone paddocks (or parts of paddocks) where the frequency is likely to be much higher, alternative crops and special management practices should be considered, including:
Region South, National, North, West