Modeling water needs and total irrigation depths of maize crop in the south west of France using high spatial and temporal resolution satellite imagery

Climate change is projected to increase water resources limitation and to impact significantly agricultural production. A big challenge for agriculture will be to reduce the amount of water used to fit the environmental constraints, while maintaining a level of production that ensure food security. In this context, we develop a methodology based on high spatial and temporal resolution remote sensing data combined with a semi-empirical crop model coupling the Simple Algorithm For Yield estimates (SAFY, Duchemin et al., 2008, 2015) with the new formulation (Battude et al., 2016) to a water balance model adapted from the FAO-56 method (Allen et al., 1998). A module was added to automatically simulate irrigation. The model was used to assess the dynamics of actual Evapotranspiration (ETca) and water supplies of maize crop over large areas and during contrasted climatic years in the south west of France. It was first calibrated and evaluated over an experimental field using four years of ETca measurements. The validation was done over 18 maize fields and larger irrigated zones (135 ha to 450 ha) using total irrigation depths. This work permitted to quantify the ability of different methods to estimate the storage capacity (soil map vs in situ data) and the basal crop coefficient Kcb (standard vs remotely sensed values) and their impact on total irrigation depths. Good estimations were obtained for ETca (R=0.88; RRMSE=20%). The model also reproduced correctly the total irrigation depth over the 18 maize fields (R=0.79; RRMSE=18.8%) and three larger irrigated zones (R=0.8; RRMSE=42%). The underestimation (Bias = -93 mm) is due to several reasons such as errors in soil water storage capacity estimates, but also to an overestimation of water needs by water managers or a potential over-irrigation carried out by farmers. Finally, the work demonstrates the high potential of combining a simple agro-meteorological model using only few parameters with satellite imagery for a large-scale monitoring of total irrigation depth. (C) 2017 Elsevier B.V. All rights reserved.