Irrigation Performance using Hydrological Modeling and Remote Sensing: Case of Semi-arid Region of Doukkala, Morocco
Akdim, Nadia1; Labbassi, Kamal1; Alfieri, Silvia Maria2; Menenti, Massimo2

Semi arid regions such as large parts of Morocco suffer from fresh water shortages, which may take dramatic proportions due to climate change and increasing socioeconomic pressure. Improvement of irrigation efficiency, which does not exceed 50%, is therefore of utmost importance. It can be achieved by carefully tuning the allocation of water to the actual requirements in space and time, in addition to technical measures to reduce losses caused by leakage during transportation and by evaporation from the soil. Remote sensing methods to monitor irrigation efficiency potentially offer unprecedented performance due to the easy access to a variety of high spatial resolution image data. The study area is an early developed irrigated area of the Doukkala region extended over about 96000 ha. We have analyzed SPOT HRV image data for an entire irrigation season in 2000-2001 and in december 1994, november 2005 and november 2008 to assess inter-annual variability of irrigation water requirements. We have selected available SPOT, LANDSAT, KOMPSAT-2, ALOS-AVNIR-2 images at the key phenological stages from 2000 to 2010 to extend such historical analysis and then for the current irrigation season we will make use of data to be acquired during an intensive ESA campaign which will collect SPOT4 data over Morocco with an unprecedented revisit frequency of 5 days. This campaign will be further enhanced by the concurrent acquisition of RapidEye and Landsat 8. The monitoring of a number of pilot plots of land (5 to 10 plots for each crop type) is being used to validate the results of biophysical parameters estimated by satellite images. The main objectives of our study are: - monitoring distributed crop water requirements on the basis of the temporal sampling provided by satellite data; - evaluate the operational viability of methods to compute actual consumptive water use with multispectral radiometric data in the VNIR, SWIR and TIR spectral regions at a spatial and temporal resolution relevant to irrigation water management in the Doukkala scheme; - demonstrate and evaluate indicators of the effectiveness of applied irrigation water. We assessed the consistency of the water allocation with the actual irrigated area and crop water requirements (CWR) by using a combination of multispectral satellite image time series and both physical (Soil Water Atmosphere Plant - SWAP- model) and empirical (FA0 56 Model and Analytical Approach) techniques. The acquisition of satellite images at key stages of crop growth, allowed an accurate estimation of biophysical parameters characterizing the state of the crop: leaf area index (LAI), crop coefficient (Kc), crop height (hc). The results of the analysis for the irrigation season 2000/2001 clearly indicate that the water allocation was much larger than CWR in November, adequate in March and severely insufficient in July. The inter-annual variability of water allocation however was very large, but always much larger than CWR.