Fixed and variable light extinction coefficients for estimating plant transpiration and soil evaporation under irrigated maize

A glasshouse experiment was carried out to estimate plant transpiration and soil evaporation from maize using fixed and Uchijima's variable light extinction coefficients at different plant densities. Plant transpiration for different plant densities was determined as the difference between evapotranspiration measured with weighing lysimeters and soil evaporation measured with micro-lysimeters. At all plant densities, the fixed light extinction coefficient gave a less precise estimation of plant transpiration (root mean square error, RMSE about 1 mm day−1). The low precision was due to an underestimation of plant transpiration during early crop growth period and overestimation after the mid-crop growth period. The Uchijima's variable light extinction coefficient gave a relatively more precise estimation of plant transpiration at all plant densities (RMSE < 1 mm day−1). A modified variable light extinction coefficient derived from our experimental data was proposed. Using this coefficient resulted in better estimations of plant transpiration and soil evaporation and relative errors were within 20% of measured values at all plant densities.