Estimating the water use efficiency of spring barley using crop models. (14th February 2018)

• Record Type: Journal Article
• Authors: Pohanková, E.; Hlavinka, P.; Orság, M.; Takáč, J.; Kersebaum, K. C.; Gobin, A.; Trnka, M.
• Is Part Of: Journal of agricultural science. Volume 156:Number 5(2018)
• Page Start: 628
• Page End: 644
• Publication Date: 2018-02-14
• Subjects: Crop model, -- evapotranspiration, -- field experiment, -- spring barley, -- water use efficiency; Agriculture -- Periodicals; 630.5
• DOI: 10.1017/S0021859618000060
• Languages: English
• ISSNs: 0021-8596

Abstract

Abstract: In the current study, simulations by five crop models (WOFOST, CERES-Barley, HERMES, DAISY and AQUACROP) were compared for 7–12 growing seasons of spring barley ( Hordeum vulgare ) at three sites in the Czech Republic. The aims were to compare how various process-based crop models with different calculation approaches simulate different values of transpiration (Ta) and evapotranspiration (ET) based on the same input data and compare the outputs of these simulations with reference data. From the outputs of each model, the water use efficiency (WUE) from Ta (WUETa ) and from actual ET (WUEETa ) was calculated for grain yields and above-ground biomass yield. The results of the first part of the study show that the model with the Penman approach for calculating ET simulates lower actual ET (ETa ) sums, at an average of 250 mm during the growing season, than other models, which use the Penman–Monteith approach and simulate 330 mm on average during the growing season. In the second part of the current study, WUE reference values in the range 1.9–2.4 kg/m 3 were calculated for spring barley and grain yield. Values of WUETa /WUEETa calculated from the outputs of individual models for grain yields and above-ground biomass yields ranged from 2.0/1.0 to 5.9/3.8 kg/m 3 with an average value of 3.2/2.0 kg/m 3 and from 3.9/2.1 to 10.5/6.8 kg/m 3 with an average value of 6.5/4.0 kg/m 3, respectively. The results confirm that the average values of all models are nearest to actual values.