Knowledge Bank

Monitoring and understanding motions of large ice streams in Antarctica requires vast, continuous, and accurate observations of ice flow. However, existing measuring technologies are inefficient or inaccurate in determining ice motion. An automatic, accurate, and economic method was developed to determine the parameters of ice motion (velocity, strain rates, and rotation rate) by matching multi-temporal digital images. A two-step matching process (cross-correlation followed by least-squares matching) was devised to achieve automation. Modifying traditional least-squares matching to determine the parameters of ice motion directly is one of the major achievements of this study. In order to control data quality, the tests of signal-to-noise-ratio, similarity, uniqueness, and consistency were designed to reject data resulting from mismatches. This image-matching approach has been applied to two SPOT images and two digitized aerial photographs successfully. The results show a nice consistency with manually measured data, and their accuracy (about 1 meter/year on velocity and 0.01 year-1 on strain rates and rotation rate) is very satisfactory. These results are considered to be very useful for glaciological studies, e.g., mass balance, force budget, and ice-flow mechanics. [Some mathematical expressions are not fully represented in the metadata. Full text of abstract available in document.]