Skip to Main content Skip to Navigation

Suivi de l'évapotranspiration des cultures irriguées du Sud de la Méditerranée par télédétection multi-capteurs et modélisation globale

Abstract : With the increasing pressure on water resources accentuated by the climate change threat, irrigated agriculture, especially in semi-arid zone, faced more challenges. Optimal management of water resources is essential in irrigated areas in order to avoid both excessive irrigation and water stress that is harmful to the growth of crops. In this context, proposing a simple, parsimonious and robust, easyto- use tool for monitoring the crop water requirement at different spatial and temporal scales would provide a tangible indicator of irrigation efficiency over irrigated perimeters. This thesis aims to evaluate and adapt several tools for monitoring the evapotranspiration of irrigated crops from the plot scale to the regional scale through the use of multi-sensor and multi-resolution remote sensing observations. Our study area is the Tensift watershed in Morocco, studied by LMI TREMA (Jarlan et al., 2015) and supported by the Tensift observatory. We identified 3 tools: a surface energy balance model (TSEB; Norman et al., 1995) driven by a radiometric temperature observation that provides indirect information on surface water status; the FAO-56 dual crop coefficient approach predicts the hydric status by computing the water balance but requires the know precisely the water input, very difficult to obtain on the plot scale of irrigated perimeters ; global modeling which, as opposed to the first two, is an autonomous approach requiring no external forcing. Firstly, the comparison of the TSEB model simulations with a set of experimental measurements collected on a plot scale during 2 consecutive agricultural seasons and with the results of the calibrated FAO-56 dual crop coefficient approach has shown that : - the TSEB model is very robust and offers acceptable performance for the prediction of evapotranspiration (RMSE <1mm / day during the two agricultural seasons); - a good capabilities of TSEB model for detection water stress; - a good ability also to partition the evapotranspiration between evaporation of the soil and transpiration of the plant. In the second step, we evaluated the predictive capacities of a global model that we developed on the basis of a time series of evapotranspiration observed in the field. Global modeling is based on the theory of the nonlinear dynamic systems theory. If it does not have the explanatory capabilities of the mechanistic models mentioned above, it may be a good alternative for predicting evapotranspiration. The analysis of the horizon of predictability of the global model that we obtained shows a limited interest for the farmers or the managers since this horizon does not exceed 3h. Nevertheless, this approach, very original in this context, remains particularly attractive and opens perspectives. Finally, we have developed a prototype, based on the TSEB model and using only MODIS free products and ERAInterim reanalyses, for monitoring evapotranspiration at Tensift watershed scale and over long periods. After an evaluation of the ERA-Interim products, we evaluated the prototype performance by (1) comparing the predictions of the energy balance components to the experimental data from Tensift observatory and (2) comparing the predicted monthly consumption at the main irrigated perimeters of the region with monthly water supplies provided by the regional office which manages these perimeters. Thus, there is a strong overestimation (almost a factor of 2) of the water consumed compared to the inflow of water, which could be related to the high number of boreholes in the region. This work and the tools developed opens perspectives for piloting and managing agricultural water in semi-arid regions.
Document type :
Complete list of metadatas

Cited literature [258 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Thursday, May 31, 2018 - 10:30:26 AM
Last modification on : Wednesday, October 14, 2020 - 4:10:52 AM
Long-term archiving on: : Saturday, September 1, 2018 - 12:43:34 PM


Version validated by the jury (STAR)


  • HAL Id : tel-01803940, version 1


Alhousseine Diarra. Suivi de l'évapotranspiration des cultures irriguées du Sud de la Méditerranée par télédétection multi-capteurs et modélisation globale. Sciences de la Terre. Université Pierre et Marie Curie - Paris VI, 2017. Français. ⟨NNT : 2017PA066488⟩. ⟨tel-01803940⟩



Record views


Files downloads