An Integrated DSS for Groundwater Management Based on Remote Sensing. The Case of a Semi-arid Aquifer in Morocco

A Decision Support System has been set up as the result of a fruitful cooperation between several public and research institutions in the framework of a large cooperation program. The DSS aims to compare spatially and temporally sectorial water demands of the Haouz-Mejjate plain (Morocco) in regard to available surface and groundwater resources. It is composed of a tool for satellite estimation of Agricultural Water Demand (SAMIR), a tool for integrated water resources planning (WEAP) and a groundwater model (MODFLOW) each of them relying on a common Geographical Information System not described here. The DSS is operating on a monthly time scale. Agricultural water demand accounts for about 80 % of the total demand. In areas where groundwater abstraction is difficult to quantify by direct methods, multitemporal remote sensing associated to the FAO methodology is a simple and efficient alternative to estimate Evapotranspiration (ET). In this work, a monthly estimate of ET from irrigated areas is derived from freely available MODIS NDVI for the 2001–2009 period. An important part of the paper deals with the validation of these estimates with eddy covariance flux measurements installed on different irrigated crops of the region. Results are satisfactory with a minus 6.5 % error per year on the monthly time scale. This preprocessing allows to dichotomize irrigated versus non-irrigated areas, and then, to estimate groundwater abstraction in subareas distinguishing by their operating modes: traditional, dam or privately irrigated. A dynamic linkage between MODFLOW and WEAP transfers the results of one model as input data to the other. The model restitutes both spatial and temporal variations in head charges and allows the calculation of the ground water balance. After calibration, piezometric validation is acceptable for the majority of the 21 head control points.