Calibration and validation of APISM-Wheat Model in Mediterranean areas

Abstract

The Agricultural Production Systems sIMulator-Wheat (APSIM-Wheat) model is one of the most widely used agricultural models. It is a powerful simulator that has been successfully calibrated and tested for many locations in the world, especially in Western Australia (WA). However, there is a noticeable lack of a standard guide for realizing the calibration validation of APSIM-Wheat that could be applied in areas with a Mediterranean climate similar to that of WA. Therefore, this study aims to examine crop simulations reported in published articles and to provide a detailed description of input data and statistical assessment, which represent the two main components of the calibration-validation protocols. The PRISMA (PREFERRED Reporting Items for Systematic Reviews and Meta-Analyses) method was used to identify and select relevant papers for this review. Following the analysis of 31 calibration protocols extracted from selected eligible articles, it was found that regardless of the objective of using APSIM-Wheat, the same category of data is required for calibration. As far as meteorological data is concerned, the information essential to this study was daily maximum and minimum air temperatures, rainfall (mm), and solar radiation. In the case of soil data, information about the texture and hydraulic characteristics, especially PAWC, DUL and LL was required. Regarding agricultural management data, this pertains to cultivated crops, Nitrogen fertilization (rate and time of application) and sowing (date and density). For the statistical evaluation, it was observed that 90 percent of studies analyzed in this review revealed the use of RMSE.