Can seasonal forecast minimise the threats of climate variability to achieve prof- itable crop-livestock productions in NSW

Reliable seasonal rainfall forecasts with sufficient lead time can play an important role in designing responses to rainfall variability. In this study, seasonal climate forecasts from POAMA (Predictive Ocean Atmosphere Model for Australia) are used to drive the AusFarm (agricultural systems analysis model) biophysical model at two contrasting locations of NSW. The POAMA2 was able to provide good rainfall forecasts started from 1 March at Wagga Wagga for the target months of March-August with statistically significant rainfall anomaly correlation skill (0.30 - 0.44) and 65-72% hit rate. At Narrabri for target months of July-December started from 1 July, the statistically significant rainfall anomaly correlation skill ranges from 0.15 to 0.46 with a corresponding hit rate of 57 to 73%. The good skill of POAMA2 in forecasting the station rainfall variability (even without downscaling) suggests that the forecast may provide value to strategic decision making in a mixed-farming systems. At Wagga Wagga, the forecast distribution of crop yields (wheat, barley, and canola), lamb sale weight and fleece weight for target months starting from March to August show comparable median productivity compared to the observed ones. Likewise at Narrabri, forecast distribution of sorghum, wheat and barley yield is similar to the distribution of observed ones, implying sufficient ability of the POAMA forecast during pre-planting period to make forward farming decisions. Work of this nature, particularly improvement of accuracy of seasonal forecast at a long lead time, has the potential to benefit production and financial performance in a given agricultural system.