For instance, different seasons cause snow and ice properties to change, which could mask or influence the signal reflected back to the satellite’s SIRAL instrument. Another example of uncertainty could be due to snow lying on top of sea ice. The weight of the snow pushes the ice floes lower into the water, so that CryoSat, which translates the height of the ice surface above the waterline into ice-thickness data, will under-estimate the true thickness of the ice.

ESA therefore goes to great lengths to ensure that CryoSat’s data are as accurate as possible. To understand and correct for any errors that could creep in, large-scale expeditions to the polar regions have to be undertaken to collect information on snow and ice properties.

These ventures were needed before launch as an essential input to the software that transforms CryoSat’s measurements into ice-thickness maps, and after launch to make direct comparisons between ground and satellite measurements.

On the ground, sophisticated equipment such as ground-radar and neutron probes, as well as more traditional techniques, such as digging snow pits and drilling through the ice, are used to measure snow and ice properties.

Data are also taken from an aircraft carrying a radar instrument very similar to that on CryoSat. These airborne data are then compared to the measurements taken on the ground and to the actual satellite measurements to fully understand and correct for sources of error.