Abstract-Sensor networks for the natural environment require an understanding of earth science, combined with sensor, communications and computer technology. We discuss the evolution from data logging to sensor networks, describe our research from a glacial environment and highlight future challenges in this field.

Abstract-A sensor network is described which obtains data from nodes on and inside glaciers. Power management through scheduling and selective control is used to allow a lifetime of at least one year on batteries. Radio links in the glacier and across 2.5km distances are used for data and commands. The prototype system was installed in Norway in 2003 and this paper describes details of the full design for 2004 through discussion of the lessons learnt.

This paper introduces a new way to investigate in situ processes, the wireless multi-sensor probe, as part of an environmental sensor network. Instruments are housed within a ‘probe’ which can move freely and so behave like a clast. These were deployed in the ice and till at Briksdalsbreen, Norway. The sensors measure temperature, resistivity, case stress, tilt angle and water pressure and send their data to a base station on the glacier surface via radio links. These data are then forwarded by radio to a reference station with mains power 2.5 km away, from where they are sent to a web server in the UK. The system deployed during 2004/05 was very successful and a total of 859 probe days worth of data from the ice and till were collected, along with GPS, weather and diagnostic data about the system.

a b s t r a c tThis study has investigated the three-dimensional movement of clasts within deformation till, using embedded wireless probes. These probes were part of an environmental sensor network, which measured subglacial properties (temperature, water pressure, resistivity, case strain and tilt) six times a day, and relayed that data via radio to the glacier surface, where they were forwarded and broadcast on-line. The system was installed at Briksdalsbreen, Norway and operated from August 2004 until August 2006. Approximately 2000 probe days worth of data were collected, with an increase in performance (41% more readings) during the second year. The probes showed similar patterns of water pressure rises throughout the two years, but with slightly different magnitudes and timings. These changes in water pressure could be related to clast behaviour. The probes decreased their dip over the year, and the rate of change was related to an increase in glacier velocity. After initial changes in dip, the probes experienced changes in orientation, followed by rotation about the a-axis. This continuous rotation was similar to the motion suggested by Jeffery [1922. The motion of ellipsoidal particles immersed in a viscous fluid. Proceedings of the Royal Society of London, Series A 102, 161-179] for the behaviour of clasts within a viscous material. In addition, some probes also showed short, frequent dip oscillations in spring and autumn, which were interpreted to reflect stick-slip events, similar to lodging; and demonstrated how local conditions can interrupt the predicted rotation pattern. Overall, it is demonstrated that when water pressures were high, decoupling occurred associated with basal sliding and dip oscillations; and when water pressures fell, the ice and sediment were coupled and till deformation occurred. These events happened during summer and autumn. It is this combination of ''lodgement'' and deformation that builds up both a complex (but predictable) fabric and a resultant composite till sedimentology.

Abstract.A wireless sensor network deployment on a glacier in Iceland is described. The system uses power management as well as power harvesting to provide long-term environment sensing. Advances in base station and sensor node design as well as initial results are described. The Glacsweb projectThe Glacsweb project [1] aimed to study glacier dynamics through the use of wireless sensor networks. It replaced wired instruments which had previously been used with radio-linked subglacial probes which contained many sensors. The base of a glacier has a large controlling effect on a glacier's response to climate change and there is a growing need to study it in order to build better models of their behaviour. Several generations of systems were deployed in Briksdalsbreen, an outlet of the Jostedal icecap in Norway. As a multi-disciplinary research project it involved people from many domains: electronics, computer science, glaciology, electrical engineering, mechanical engineering and GIS.Initial deployments had to solve the mechanical design of the probe cases and the unknown radio communication issues. The solutions involved craft as much as science and engineering but the key success has been to create data which had not existed before [2,3,4,5] while advancing our knowledge of sensor network deployments. Hot water drills are used in order to produce holes which reach the glacier bed. Most probes are placed 10-30cm under the ice while some are placed within the ice. Due to the relatively slowly changing environment the probe sense rate is normally set to once every four hours, although an adaptive sampling algorithm has been developed in the lab [6] which would optimise this sampling rate. Iceland deploymentSkalafellsjökull is a part of the large Vatnajökull icecap in Iceland and our site was chosen at (64°15'27.09"N, 15°50'37.68"W) around 800m altitude near an access road. Although there was no local internet connection there was a mobile phone signal which we used for the main internet link. The glacier is deep enough to test beyond 100m depth in the future (we used 60-80m).Due to the nature of the team and time available a few key topics were chosen for the developments for the Iceland deployment. One main area was to improve the basestation design through the use of a Gumstix processor. This would provide a better development environment and easier package management. The probes would maintain the PIC18 microcontroller but would gain an improved power supply and simplified code. In terms of sensors the temperature needed to have smaller quanta in order to sense small changes. A light reflectance sensor was also on the list of items to be tested in order to provide more information on the nature of the material surrounding the probes. A simple star network was used rather than our more complex TDMA-based protocol [7] in order to simplify debugging.The disintegration of the ice front in Briksdalsbreen [8] meant that we lost our previous basestation infrastructure. This coupled with the higher altitude of the Icel...