Penman-Monteith approaches for estimating crop evapotranspiration in screenhouses--a case study with table-grape

In arid and semi-arid regions many crops are grown under screens or in screenhouses to protect them from excessive radiation, strong winds, hailstorms and insects, and to reduce crop water requirements. Screens modify the crop microclimate, which means that it is necessary to accurately estimate crop water use under screens in order to improve the irrigation management and thereby increase water-use efficiency. The goal of the present study was to develop a set of calibrated relationships between inside and outside climatic variables, which would enable growers to predict crop water use under screens, based on standard external meteorological measurements and evapotranspiration (ET) models. Experiments were carried out in the Jordan Valley region of eastern Israel in a table-grape vineyard that was covered with a transparent screen providing 10% shading. An eddy covariance system was deployed in the middle of the vineyard and meteorological variables were measured inside and outside the screenhouse. Two ET models were evaluated: a classical Penman-Monteith model (PM) and a Penman-Monteith model modified for screenhouse conditions by the inclusion of an additional boundary-layer resistance (PMsc). Energy-balance closure analysis, presented as a linear relation between half-hourly values of available and consumed energy (1,344 data points), yielded the regression Y=1.05X-9.93 (W m(-2)), in which Y=sum of latent and sensible heat fluxes, and X=net radiation minus soil heat flux, with R2=0.81. To compensate for overestimation of the eddy fluxes, ET was corrected by forcing the energy balance closure. Average daily ET under the screen was 5.4±0.54 mm day(-1), in general agreement with the model estimates and the applied irrigation. The results showed that measured ET under the screen was, on average, 34% lower than that estimated outside, indicating significant potential water saving through screening irrigated vineyards. The PM model was somewhat more accurate than the PMsc for estimating ET under the screen. A model sensitivity analysis illustrates how changes in certain climatic conditions or screen properties would affect evapotranspiration.