Programme OS1a Mediterranean
Hydrological modelling in the High Atlas mountains with the help of
remote-sensing data: milestones of the SudMed project
Author(s): G. Boulet(1), A. Boudhar(2), L. Hanich(2), B.
Duchemin(1), V. Simonneaux(1), P. Maisongrande(1), A. Chehbouni(1), S. Thomas(1), A. Chaponnière(1)
(1) CESBIO (UMR 5126 CNES-CNRS-UPS-IRD), Toulouse,
(2) Faculté des Sciences et Techniques, Marrakech, Maroc
Keyword(s): Morocco, Remote-Sensing, Snowmelt, Climate Change,
Session: OS1a Mediterranean
The sustainable management of
water resources worldwide is one of the most important challenges of the 21st century. In the Tensift-Haouz plain,
located in central Morocco and characterized by a semi-arid climate, 85% of available water is used for agriculture.
Precipitation, which is concentrated over the High-Atlas mountain range, falls in a significant proportion as snow
leading to storage until the snowmelt period and a significant contribution to baseflow in the summer.
In order to
understand the hydrological cycle of the Tensift basin, the sudmed project has developed a research program based
on modelling, experimental and remote-sensing monitoring of the High Atlas subbasins. A top-down and a bottom-
up strategy have been combined in a modelling approach of medium-range complexity. This approach simulates the
largely unobservable subsurface flow processes using input data from the sparse network of climate stations while
incorporating as much ancillary available information as possible: noisy streamflow observations, geochemical
elements tracing, remotely-sensed land cover and snow area. The bottom-up approach is based on the distributed
hydrological model SWAT and uses a simplified land cover map, while the top-down approach uses two global
semi-empirical models, GR4J and SRM.
In parallel, complex energy-balance models are used to ensure realistic
outputs for the simple degree day snowmelt models embedded in the three models. Time series of snow cover
derived from SPOT-VEGETATION and MODIS are assimilated into all models to improve the contribution of
snowmelt to runoff. The improvement is significant given the lack of distributed climate observations, but its impact
on the runoff is relatively minor. The spatialisation of the climatological forcing data remains a pitfall for all studies. An
improved spatialisation is developed on the basis of remote-sensing data in the Thermal Infra Red spectrum and
outputs of current mesoscale meteorological models. Kriging methods based on a digital elevation model and the
available high-resolution TIR data at selected dates are combined with the very low resolution TIR images from
geostationary satellites and the outputs of the meteorological models.
The climate-change scenarios provided by
these models are eventually used to assess the impact of climate change on the hydrological regime of the High Atlas
and the availability of water in the foot plain.