Downscaling transpiration rate from field to tree scale

tEstimation of field spatial variability of tree actual evapotranspiration (ETa) in orchards is key whenquantifying water and associated nutrient leaching at the field scale. Though ETais often measured atthe field scale, spatial variations between individual trees are likely due to local differences in soil wateravailability and canopy cover. It is therefore that we propose seeking a statistical relation between fieldETa, tree midday stem water potential (MSWP), soil water storage (WS), and tree potential evapotrans-piration (ETc) with relative tree canopy cover (Crel). Four years of soil and almond trees water status datawere used to optimize an artificial neural network (ANN), to predict field scale ETafirst, followed bydownscaling to the individual tree scale. ANN’s using two hidden neurons (11 parameters) proved to bethe most accurate (RMSE = 0.0246 mm/h, R2= 0.944), seemingly because adding more neurons generatedoverfitting of noise in the training dataset. Crelwas the main source of variability of ETa, while MSWP wasthe controlling factor for the tree-scale relative ET. At a given soil WS, almond trees of the drip-irrigatedblock were less affected by root zone water stress than the fanjet micro-sprinklers block, likely becauseof soil textural differences between the two main experimental blocks. In wet conditions, the predictedtree ETafollowed a normal distribution (with relative standard deviation of about 5%), which was close tothe Creldistribution. However, standard deviation values increased (7.6% for the whole orchard) duringperiods of water stress.