A digital twin for arable crops and for grass

Samenvatting

There is an opportunity to use process-based cropping systems models (CSMs) to support tactical farm management decisions, by monitoring the status of the farm, by predicting what will happen in the next few weeks, and by exploring scenarios. In practice, the responses of a CSM will deviate more and more from reality as time progresses because the model is an abstraction of the real system and only approximates the responses of the real system. This limitation may be overcome by using the CSM as a digital twin. A digital twin (DT) is a model of a specific physical object, that is kept synchronized by using real-time observations on that object. In this paper we present the Digital Future Farm (DFF), a digital twin for arable and dairy farming. The DFF comprises access to data sources (e.g. weather, soils, farm management, remote sensing), a suite of models, and utilities for data assimilation and visualization of simulation results. The working of the DFF is demonstrated with examples from a multi-year experiment and from a commercial potato farm. In addition to a CSM, the DFF is also demonstrated to work with a summary model for potato growth. Initial experiences indicate that the DFF produces information that is helpful to farmers but it is difficult to evaluate the performance of the DFF in quantitative terms because of variability between years, fields, and the lack of availability of on-farm data. The most immediate contribution of the DFF is to provide farmers with a ranking of their fields according to how urgently they need an intervention. Experiences with the DFF have helped to formulate further research questions.