Multipath mitigation of carrier-phase GPS position estimates from the Helheim glacier: using new reduced sidereal filtering approach
Examensarbete för masterexamen
Nilsson, Johan NT
The Greenland ice sheet contains a vast volume of water frozen over millennia, which have in recent decades come under investigation, motivated by the possible effects of global warming. Recent observations have shown that outlet glaciers on the edges of Greenland’s ice sheet is melting (retracting) at a unexpected rate, not perviously seen. So it is of vital importance that we understand the dynamics of the outlet glaciers on Greenland, such as Helheim, because they play a crucial part in the predictions of sea-level rise. Because resent investigations have seen an increase in glacier speed, calving rates, and glacial earthquakes. From this it is clear that the understanding of the dynamics and the link between these events are still poorly understood. On the Helheim glacier we know that glacial earthquakes, calving and glacier speed up all happen close in time with respect to each other, but due to the multitude of noise and other signals, the specific earthquake time is hard to resolve. Sidereal filtering is an effective way to remove and reduce noise (which consists mostly of multipath) which has periodic behavior that repeats itself from day to day in the position estimates for high rate GPS. This technique has been uses primarily for earthquake deformation studies where the sites involved have been assumed to be stationary relative to each other before the earthquake. But in an environment of an outlet glacier like Helheim on Greenland this generalization does not hold. There is not only large motions in the flow of ice (∼30 m/day), there are also significant tidal and diurnal motions associated with insolation and hydrology. The purpose of this thesis is to extended the sidereal filtering approach to a glacial environment with the goal to remove high levels of noise. At the same time the goal is to preserve the dynamics of interesting periodic signals inherent in the data which are usually removed by the regular sidereal filtering. I found that it is possible to extend the sidereal filtering technique to this kind of a highly variable environment. And that it’s also possible to preserve the dynamics of periodic signals using this technique. When applying this new extended technique preserving periodic behavior (reduced sidereal filtering) a reduction in noise in the range of 30−70%, a signal degradation of 18−20% and a reduction in variability of 4−20%. For the ordinary sidereal filtering we found a noise reduction range of 45−90% and signal degradation of 75−100% and a reduction in variability between −28−85 %. Further I also found that correlation is a good overall description for repeatability and filter performance of the sidereal filtering technique in this environment. I also discovered that the reduction in noise for the reduced sidereal filter also strongly depends of the stations velocity profile.
Geovetenskap och miljövetenskap , Grundläggande vetenskaper , Earth and Related Environmental Sciences , Basic Sciences