Ice pressure ridges
The thickness of the ice pack is not constant, because pressure ridges form when ice blocks are piled up above and under the ice. These ice blocks form by fracturing and cracking when ice floes or thin ice on leads are pushed against each other or against the coast by winds and currents. Although the interior of a ridge might freeze thicker than the adjacent level ice, the ice blocks of the deep parts of the keel remain only loosely connected and unconsolidated. Typical ridges in the Borneo region will be 1-2 m high, i.e. have a thickness of 10-20 m.
 EM bird accuracy Because of the “footprint”, or zone covered by the EM Bird's instruments (30-40 m, or 2-3 times the normal flying altitude of 15 m) and due to the three-dimensional structure and porosity of the ice, EM measurements underestimate the maximum thickness of pressure ridges. This is due to the greater induction in the water adjacent to the ridge keel and within the porous structure of the blocky keel. These factors increase the measured secondary field strength, which is interpreted as thinner ice by the processing algorithm.
The accuracy of EM measurements over level ice is very high (better than +/- 0.1 m), so a comparison of EM thickness with “true” thickness shows a good correlation, and the regression line has a slope of 1. However, where the ice is thicker than the typical (modal) ice thickness, the correlation becomes worse, and the regression line has a slope of <<1. The objective of the validation campaign to obtain enough data to determine the slope of the regression line for deformed ice, i.e. the amount by which EM measurements underestimate the “true” thickness.
Sea ice 3D topography compared to EM bird measurements
The extent to which EM underestimates ridge thickness can be determined by comparing EM thickness values and “true” thickness over the same ridges. To do this, we will perform extensive underwater profiling using upward-looking sonar (ULS) mounted on a ROV, and by diving. The measurements will be taken in a rectangular zone 200 m long and at least 60 m wide, perpendicularly crossing a ridge via its longer side. The ROV will profile a grid pattern with a line spacing of 2-5 m. At the surface, the surface elevation and snow thickness of the validation zone will be surveyed and measured by means of photo-stereoscopy. Afterwards, the same area and the ridge will be over flown with the EM bird. If time allows, we will extend the size of the area perpendicularly to the long side, i.e. in the ridge direction. Then, we will fly several profiles across the ridge to increase the number of measurements.
Results: EM bird accuracy, IceSat and CryoSat calibration.
Depending on the quality and resolution of the underwater data, the measurements will be either gridded or used in their raw form. We will then extract coincident profiles of surface elevation, freeboard, draft, and EM thickness using all data. And then the data will be plotted in scatter plots showing EM thickness versus “true” thickness. This will reveal the desired underestimation coefficient. We will also retrieve the footprint size of the EM Bird from the measurements, e.g. by examining the distance from the ridge, where the ridge is still “seen” by the bird.
The data obtained will also allow us to compare surface elevation with draught, which is very important for satellite measurement of ice thickness (using ICE Sat or Cryo Sat) because these satellites measure only surface elevation as a proxy for ice thickness. Therefore, surface elevation needs to be transformed into true ice thickness. Our validation campaign can yield the required transformation equation. Thanks to Archimedes' law, comparison of freeboard and draft also allows us to compute the density of the ridge. From this, the porosity can be calculated, because the ridge is a mixture of sea ice and water, whose densities are well known. The porosity is a major factor in calculating the underestimation of ridge thickness by EM device. We will be running a finite-element EM model to simulate EM induction in the ridge, and the porosity needs to be included in the model.
|
