387 - In search of lost water under glaciers
October 2011
Over 99 % of the Earth’s fresh water exists in ice formations or underground. IRD geophysicists, aiming to find ways of detecting this resource, are at the spearhead in the development of an innovatory method based on nuclear magnetic resonance. To date, it is the only technique applicable for detecting liquid water underground or under a glacier from the surface and for estimating the volume.
This method recently found an original application as an aid for warning of glacier hazard. It successfully detected the presence of an immense water pocket of 55 000 m3 sitting under the Tête Rousse glacier in Haute-Savoie. This posed a flooding threat to people living in the valley below. Warning was given and the local authorities conducted a draining operation.
This technique is adaptable to glacier risk management, but it can also help for water supply provision. It can benefit both tropical mountain areas, such as the Andes or the Himalaya where glacial water can be a major threat, given the context of climate change, and semi-arid regions where water resources lie stored deep underground.
In the world’s semi-arid regions water is usually hidden, because it lies mainly underground. Detection and quantification of this resource has always been one of hydrogeologists’ main activities. However, the underground environment is riddled with discontinuities. These cannot be spotted from the surface, as pinpoint exploration techniques by probe or borehole drilling are too costly and insufficient for correctly estimating the volume of water available in an underground water body.
To get round this limitation, geophysicists have developed indirect surface-based exploration techniques, the most standard being based on the propagation and deformation of electrical or magnetic waves. These methods allow detection of discontinuities but these are not necessarily linked to groundwater. Neither are they able to yield estimates of water volume. An IRD research team from the ‘Laboratoire d’étude des transferts en hydrologie et environnement (LTHE)’ have for several years been developing a method based on proton nuclear magnetic resonance (Proton or 1H NMR), in conjunction with the BRGM and the Geophysical Institute of Israel.
A very different method
The 1H NMR was developed for non-intrusive exploration for water in rocks, to depths of 0 to 100 m. An alternating current generated at the ground surface creates an electromagnetic field which in turn triggers resonance of water molecules held in the rock underground. The research team subsequently took measurements –still from the surface– of the alternating magnetic field these water molecules produced. This method is sensitive only for water, making it quite distinct from traditional geophysical techniques which analyse anomalies in structures or physical parameters.
Glaciers can trap water
This special NMR technique has proved highly suitable in many tropical regions for quantification of water resources. A recent application in a mountainous area demonstrated its additional potential for the management of risks associated with water which in this case can no longer be considered hidden but lost. The water cycle in a mountain environment has a strong seasonal component, with accumulation as snow in winter and as melt water in summer. In the glacial environment, this melt water is drained away by mountain streams which flow over the surface then disappear through glacial rills( 1) which close up in winter but can reach an impressive size at the height of summer. Water lost over the surface in this way collects in subglacial streams which then emerge down stream from the glaciers. However, some underground formations preclude the existence of these subglacial streams that drain melt water away. The lost water stays trapped under the glacier. This is the case at the Tête Rousse glacier, in Haute-Savoie, which although small (0.08 km²), is vitally important owing to the dangers that lurk within. An 80 000 m3 subglacial water pocket suddenly burst out in 1892, causing 175 deaths in the valley below. Ever since that time, the local authorities and people have feared that such an event could happen again, but without being able to assess what quantity of water was enclosed under the glacier or at what rate it was accumulating. The current global warming –already strongly apparent in the Alps– brings fears of a weakening of the glacial “lock” which holds the water in.
A real danger
In 2009 and 2010, the LTHE team, prompted by glaciologists from the LGGE( 2), set to work at the local authorities’ request. They offered their services to check for the presence of a water pocket and estimate its volume if there was one. This is a real feat, in both scientific and technical terms, and the first of its kind. No standard geophysical method provides a means of making such an estimation and the proton NMR has never before been applied in such a situation (water under ice, high altitude, manœuvering difficulties in high mountain terrain).
In June 2010, a newly developed 3D software package was used to assess the volume of water stored. The figure was 55 000 m3, large enough compared with the water pocket which triggered the 1892 disaster for the local authorities to engage a pumping operation to drain off the pocket. Using this artificial drainage system 48 000 m3 of water were pumped away. Now after drainage about 4000 m3 remains but this does not pose any threat.
The usefulness of this innovatory detection technique for underground water resources had already been proved. This more recent work has shown a new potential for the method adapted to glacial risks, applicable to other regions particularly concerned by the problem, like the Andes or the Himalaya, especially promising in the context of climate change.
Authors: Thierry Lebel et Gaëlle Courcoux
