Chris Nunn 1*, Jaclyn N. Brown 2, Peter McIntosh 3, Kavina Dayal 2, Michael P. Bange 1
1CSIRO Agriculture and Food, The Australian Cotton Research Institute, Locked Bag 59, Narrabri, New South Wales, 2390, Australia.
2CSIRO Agriculture and Food, Private Bag 12, Hobart, Tasmania, 7001, Australia.
3CSIRO Oceans and Atmosphere, Retired.
*Corresponding author (firstname.lastname@example.org).
Australian cotton production is characterised as a high-value, quality product, earning in excess of $2.5B in exports annually. High year-to-year variability in rainfall and temperature influence productivity which could be managed more effectively with appropriate foresight. Seasonal climate forecasts have the potential to equip growers with upcoming season knowledge to mitigate risk but also to implement less risk averse behaviour in good years. Here we outline the key decision points where a seasonal climate forecast of temperature and rainfall could provide actionable insights. We further test the skill and reliability of Australia’s current seasonal climate forecast system to provide this information at four key locations: Moree, Griffith, Dalby and Emerald compared to the SILO record of weather from 1981-2017. Model predictions were assessed in terms of correctness and also how emphatic the prediction was. In many cases the range of predictions offered across the POAMA ensembles is so large that variation in the median is overshadowed, leading to a flattened response across years. This offers little advantage over the current approaches that use climatology as a guide to the future. Advances in climate modelling need to address both the predictive skill and the broad prediction envelope to offer useful advances in this field.