Predicting bean yields, and the optimal harvest date, is as important as producing prime quality beans.
Supply chain management is critical if market expectations for quality and price are to be met by Australian producers.
Reductions in both yield and quality can occur due to several biological factors and/or or environmental events.
Such events, for example a pest outbreak or diseases outbreak, require a management response to mitigate the decrease in yield and/or quality.
The management responses to events are expert agronomic decisions. These may constitute an event that may explain an outlier in the predicted yield or may reduce the packaged yield or proportion of beans in a quality category.
Data management was a key part of this project with both historic data from Mulgowie and downloaded or purchased meteorological data contributing to the data for model development.
The model took an agro-ecological approach using planting date, phenology, historic growing data and expert agronomic inputs along with meteorological data to predict harvest date and yield.
The predictive use of the model is likely to require forecast (3 monthly) data for application to current growing conditions to predict future harvest dates.
This feasibility study was extended into a full project: Predicting Green Bean Harvest & Yield.
