Data requirements for agricultural systems research and applications

Abstract

The quantitative assessment of agricultural systems requires a balance between simulation and data to support the simulations. Data for systems research is needed as inputs or as constraints for simulation models and as validation of simulations to establish the credibility of the systems approach. If agricultural and land-management simulation models are to be useful for solving problems, they need to include in the detail of the simulation representation of processes in several subject-matter areas. The level of detail required as output in a simulation determines the amount of data necessary as inputs and for validation. This paper suggests five levels of modelling detail, ranging from local experience to very detailed mechanistic simulation models. Biophysical models that require daily weather data have evolved into the most useful form of models for the transfer of description based on agrotechnology information. However, some elements of the required daily weather data, especially global radiation, are not always available. In such instances, substitutes such as hours of sunshine or a weather generator can be used.

Data that describe the biophysical characteristics of a system are required as an input for the simulation process. They include data on weather, soils, crop management, the characteristics of the cultivar, and the incidence of pest. Information on the crop is needed for model validation. Socioeconomic data required include input costs and selling prices, attitudes toward risk, financial and human resource constraints, and data on several social factors such as land tenure, education and wealth.

Scientists within the ICASA group are committed to the development of practical and balanced systems approaches that will assist in many important aspects of decision-making processes in agriculture. A systems approach can increase the efficiency of agricultural research and provide more realistic information to policymakers.