Quantifying the value of historical climate knowledge and climate forecasts using agricultural systems modelling

A well tested agricultural systems model was used together with 114 years of historical climate data to study the performance of a dryland wheat-fallow system as impacted by climate variations and nitrogen input levels in southeast Australia, and to investigate the value of: (1) historical climate knowledge, (2) a perfect climate forecast, and (3) various forecasts of targeted variables. The potential value of historical climate records increases exponentially with the number of years of data. In order to confidently quantify the long term optimal nitrogen application rate at the study site at least 30 years of climate data are required. For nitrogen management only, the potential value of a perfect climate forecast is about $54/ha/year with a reduction of excess nitrogen application of 20 kg N/ha/year. The value of an ENSO based forecast system is $2/ha/year. Perfect forecasting of three or six categories of growing season rainfall would have a value of $10-12/ha/year. Perfect forecasts of three or six categories of simulated crop yield would bring about $33-34/ha/year. Choosing integrated variables as a forecasting target, for example crop yield derived from agricultural modelling, has the potential to significantly increase the value of forecasts.