Comparing canopy-scale actual transpiration fluxes of a tomato crop as measured in the field and estimated by an agro-hydrological model

Root water uptake, and subsequently transpiration, is a main component of the hydrological cycle and, hence, a main input to hydrological models. Transpiration rates can be either measured in the field at leaf and/or canopy scale or can be estimated using numerical modelling with either microscopic or macroscopic approaches. The main purpose of this study is to compare the transpiration rates measured at the leaf scale and those calculated by a macroscopic approach embedded into the Agro-hydrological model FLOWS under variable soil properties and water availability. For this purpose, sixteen plots were cultivated with tomato crops in Metaponto Area in South Italy. Of those plots, 8 plots were irrigated with 100% of the potential evapotranspiration, ETp, (hereafter, the control group), and 8 plots were irrigated with 80% of ETp (hereafter, the Deficit Irrigation group or DI group). Soil Hydraulic Properties (hereafter, SHP) were collected using a new fast field measurement based on the infiltration from a point source. Leaf-Area Index, LAI, was also measured in situ using a leaf-area meter. The crop coefficients, Kc, were estimated from LAI based on the literature for tomato crops in Southern Italy. The daily macroscopic transpiration rates, Ta,m, were obtained using FLOWS Agro-hydrological model, which is based on solving one-dimensional Richards Equation (RE), using the soil and vegetation data. The leaf-scale stomatal conductance, gs,l, and transpiration rates, Ta,l, were measured in the field using the infrared Gas Analyzers (IRGA). For the sake of comparison with the macroscopic transpiration rates, gs,l was upscaled to canopy scale stomatal conductance, gs,c, by the big-leaf approach using LAI and an extinction factor accounting for radiation attenuation. Then, the canopy-scale transpiration rates, Ta,c, were obtained by the well-known Penman-Monteith equation using the gs,c. Multiple Linear Regression, MLR, was used to find the statistical correlation among transpiration rates (both Ta,m and Ta,c), the SHP and gs,c, The results emphasize the strength of the model as it smooths the spatial variability of transpiration rates reducing the uncertainties resulting from the erratic variabilities coming from leaf-scale measurements as well as the ability of the model to obtain the daily transpiration rates along the whole growth season, which are difficult to obtain from leaf-scale measurements. The results also showed the important role of SHP in transpiration rates. Both Ta,m and Ta,c are strongly affected by the saturated water content, θs, and the slope of the water retention curve, nvG. In addition, a reduction in transpiration rates was observed in the whole DI group and even in a plot in the control group. The stress experienced in the latter plot was due to the SHP and proved that stress periods can occur even when providing the roots with 100% of ETp.