Modelling soil water dynamics in winter wheat using different estimates ofcanopy development

Three years of soil water dynamics within plots of winter wheat were measur ed and simulated by the SOIL model. The winter wheat was cultivated at thre e nitrogen fertilization levels (0, 60 and 120 kg N ha(-1)), resulting in d ifferences in canopy development and evaporative demand between plots. The soil properties were assumed to be uniform within the field. The SOIL model was calibrated using soil water content measured during 1990 in a plot fer tilized at the highest level. Standard meteorological variables together wi th measured parameters describing soil and plant properties were used as in puts to the model. Additional parameters were obtained from the literature. The parameter set resulting from the calibration was applied for the years 1990-1992 and for all three fertilization levels. Measurements of leaf-are a index and estimated rooting depth were specified for the individual years and plots. Time series of canopy resistance are usually supplied to the mo del as a driving variable, but in our application, the model alternatively calculated canopy resistance using the Lohammar equation. The equation has mostly been applied to forests but was here used for winter wheat with good results using parameter values obtained from the literature. Three estimat es of canopy development were used as input to the SOIL model: green leaf-a rea index and green leaf + stem area index measured on plant samples in the laboratory and leaf area derived indirectly from measurements of spectral reflectance. The agreement between model predictions and measurements of so il water dynamics was generally good when green leaf-area index or leaf are a derived from spectral-reflectance measurements were used as input. Genera lly, spectrally derived leaf area was found suitable for replacing laborato ry measurements. Spectral-reflectance measurements are non-destructive, fas t and inexpensive compared to standard destructive measurements. Model pred ictions were most sensitive to the methods used for measuring leaf area ind ex and for estimating canopy resistance during the early season when evapot ranspiration was limited by canopy size and under drought and nitrogen limi ted conditions. (C) 2000 Elsevier Science B.V. All rights reserved.