Does collaborative farm-scale modelling address current challenges and future opportunities

Abstract

Historically, farm-scale models have tended to be created, owned and maintained by a single person or research organisation. This modus operandi is proving increasingly fragile, when confronted with budgetary constraints and staff turnover. At the same time, rapid developments in sensors and communication technology mean that there are increasingly opportunities for data acquisition relating to farm-scale activities; data that could enable models to be parameterised for individual farms. Collaborative modelling is proving to be a viable alternative that has numerous advantages; it accesses a wider range of expertise – particularly relevant for farms with livestock, manure management systems and a range of crops, it allows costs to be shared, buffers budget and staff changes in individual organisations, increases quality control of model code and extends the biophysical and management dimensions of model testing. However, collaborative modelling itself presents practical and cultural challenges that must be overcome and also imposes some costs. The practical challenges include agreeing the choice of computing language (and often operating system), the need to develop QA/QC procedures and agreeing how costs should be shared. The cultural challenges include the need for research organisations to acknowledge the necessity of joint ownership of a flagship activity and for modellers to agree common quality criteria, and the willingness to accept criticism that this implies. We here reflect on the experience garnered through the development of two modelling platforms and assess their role in determining the future course of farm-scale modelling.