Using long-term field experiment data to prepare a crop simulation model for climate impact studies

Abstract

This study presents the performance of the 4M crop model when simulating the effects of different fertilizer levels. Calculated biomass formation data were compared to observed data collected in a long-term field experiment launched in 1958 at Martonvásár, Hungary. After calibration, the model adequately simulated the differences between the effects of the different fertilization levels, as well as the long-term trends and averages of biomass productions. On the other hand, the model underestimated the inter-annual variability of the yield revealing a fundamental problem of crop modeling. A new formula was proposed to enable the model to simulate the increased leaf decay rate induced by heat and water stresses. The calibrated model was used to estimate the yields of 294 representatively selected Hungarian farms. After validation, the model was used for projecting the effects of fertilization in a future climate scenario. 4M predicted a considerable decrease of maize yields as well as yield safety for the future. The comparison of the observed and simulated data suggested two ways of model improvements which facilitate a more accurate projection of the impact of environmental stress factors.