Investigating the effects of APSIM model configuration on model outputs across different environments

Introduction Soil type plays a major role in nutrient dynamics and soil water which impacts crop growth yield. The influence of characteristics on is usually evaluated through field experimentation (in the short term) crop-soil modelling long-term). However, there has been limited research looked at effect model structural uncertainty outputs different types. Methods To analyze impact inputs uncertainty, we developed eight structures (a combination two models, models irrigation models) within Agricultural Production Systems sIMulator (APSIM) across three types (Ferralsols, Alisols Chernozems). By decomposing mean proportion variance simulated values (yield, irrigation, drainage, nitrogen leaching partial gross margin) identified magnitude uncertainty. Results For all types, was most significant source contributing >60% to variability for modelled variables, except demand dominated by choice applied. Relative first order interactions, were minimal (<12%) contributions from second interactions (i.e., inter-model components). We found that higher does not necessarily imply high actual values. Despite yield PGM (contributing over 90%), small standard deviations (ranging 0.2 1 t ha -1 ) 50.6 374.4 USD compared (yield: 14.6 , PGM: 4901 indicate relatively low Similarly, had contribution 45% variance, but ranging 11 33.3 mm overall 500 suggest In contrast, environmental variables- drainage leaching, more than 60% 70% respectively, yet large 7.1 30.6 0.6 7.7 kg (44.4 mm) (3.2 ), Discussion need include only fractional also measured units (e.g. mm, assessments provide useful agronomic or policy decision-making information. findings this study highlight sensitivity agricultural moisture availability, suggesting it important give attention when dry/wet conditions depending output analyzed.