The Glacsweb project at the University of Southampton, UK studies glacier behaviour using a variety of new technologies. The teams designed the first radio-linked sensor probes to place in/under glaciers. One important data stream is the movement of the ice, which has been measured over the years using differential GPS. These data help the team understand how glaciers respond to climate change.

Challenge

Traditionally data recordings have been made on the moving GPS and a fixed base. These are then processed offline to calculate precise positions. However this uses considerable battery power to receive for long periods and also to transmit the data off-site. Using real-time kinematic mode, with L1 and L2 signals promises good results while avoiding large data volumes or long recording periods. The team had to design a system that could do ths on minimal battery power for a whole year, placed anywhere in the world.

Solution

A low power microcontroller (running micropython) was used to control and schedule the Piksi Multi units and fetch a GPS fix once it was available. The units must power-up at the same time so an accurat real-time clock was used. The gathered data are then sent once per day back to the project’s server using Iridium short text messaging. The Piksi’s centimeter level accuracy was achievable in under ninety seconds, which saved power, while the precision was found to be sufficient to track the moving glacier.