Journal article Open Access
Popova, Zornitsa; Gabrielle, Benoit; Leviel, Bruno; Kercheva, Milena
This paper reports on CERES-maize model stepwise calibration, modification and validation using independent data from specific experiments carried out in a chromic luvisol (Chelopechene), Sofia field,1997-1999. Laboratory based water conductivity curve WCC (eq.1) and water retention curve WRC (eq.2) are calibrated (step i) and adjusted to data measured in field (step ii) (Fig.1; Table 3). Field observations of water dynamics under maize hybrid Kn 509 relative to optimal nitrogen supply N200 (kg ha-1 ) and variable water stress (irrigated A1 plots/lysimeters №1 and №2) are used to modify the function of Root Water Absorption (eq.3) (step iii). The ability of calibrated biological module of CERES-NC-Maize model to make predictions without any further adjustments of parameters is tested against independent three-year dataset (1997-1999) collected in the same field (Chelopechene) but in lysimeter №1 (step iv). Observed water contents in the 1.30 m root zone PESW and soil layers SW (Figs.3a, 3b, 3c, 3d), actual crop water uptake (evapotranspiration ET) (Figs. 4 and 5) and water fluxes prove acceptable agreement with validated model outputs. Model predictions of crop growth and dry weights of maize are acceptably precise in most of the tested situations (Fig.6). The results support the use of tested CERES-NC model in prediction of aboveground dry matter, water balance and storage under maize on Chromic Luvisols in Sofia field.