Estimation of evapotranspiration in maize fields with ground mulching with plastic film in semi-arid areas using revised Shuttleworth-Wallace model

Evapotranspiration (ET), which is comprised of evaporation from soil surface (E) and transpiration from vegetation (T), plays an important role in maintaining global energy balance and regulating climate. Quantifying partitioning of ET is particularly important for accurate prediction of climate response to ecosystem carbon, water and energy budgets. Using eddy covariance measurements in maize fields for the growing season at the Experiment Station of Agro-ecosystem in Semiarid Area (ESASA) of Lanzhou University, we ran the revised Shuttleworth-Wallace model (S-W model), partitioned evapotranspiration in maize fields under plastic film mulch conditions into evaporation and transpiration, validated the performance of the model for different time scales and under different weather conditions with measured eddy covariance values, analyzed the driving factors, and determined parameter sensitivity of ET and its components. The results suggested that the simulated ET in the study area was in good agreement with the measurements in both sunny and cloudy days, but the model performed badly in rainy days. In diurnal timescale, the modified model performed well when ET was larger than 2 mm.d^-1 in both sunny and mostly cloudy days, and the ratios of simulated values by S-W model to measured values were close to the 1∶1 line. But the model slightly overestimated ET in rainy days. Solar radiation and temperature were key environmental factors influencing ET in maize fields under plastic film mulch that led to seasonal variations. In general, T accounted for a small fraction of ET in maize fields under plastic film mulching. Diurnal variation in E/ET followed a single-peak curve, the low point was observed at night. At seasonal timescale, E/ET decreased from 18% to 8% at jointing stage, and kept at 8% level at tasseling stage and filling stage. We compared our result with others and found that both film mulch and environmental factors affected the value of E/ET. The E/ET was controlled by canopy stomatal conductance at diurnal timescale, while at seasonal timescale, it was mainly controlled by leaf area index (LAI) and soil moisture content (θ) which regulated transpired water from leaf stomata and evaporated water from bare soils. The sensitivity analysis showed that ET and its components were most sensitive to aerodynamic resistance from canopy to reference height (r_a^a) and bulk resistance of boundary layer (r_a^c), and moderately sensitive to bulk resistance of canopy stomatal (r_s^c), and insensitive to aerodynamic resistance from soil to canopy (r_a^s) and soil surface resistance (r_s^s). It was suggested that ET and its components were more sensitive to parameters related to canopy. Therefore, it was much more important to determine resistance parameters of raa, rac and rsc when simulating ET in maize fields under plastic film mulch using S-W model.