Sea Ice Topography Profiling using Laser Altimetry from Small Unmanned Aircraft Systems

摘要

Arctic sea ice is undergoing a dramatic transition from a perennial ice pack with a high prevalence of old multiyear ice, to a predominantly seasonal ice pack comprised primarily of young first-year and second-year ice. This transition has brought about changes in the sea ice thickness and topography characteristics, which will further affect the evolution and survivability of the ice pack. The varying ice conditions have substantial implications for commercial operations, international affairs, regional and global climate, our ability to model climate dynamics, and the livelihood of Arctic inhabitants. A number of satellite and airborne missions are dedicated to monitoring sea ice, but they are limited by their spatial and temporal resolution and coverage. Given the fast rate of sea ice change and its pervasive implications, enhanced observational capabilities are needed to augment the current strategies.The CU Laser Profilometer and Imaging System (CULPIS) is designed specifically for collecting fine-resolution elevation data and imagery from small unmanned aircraft systems (UAS), and has a great potential to compliment ongoing missions. This altimeter system has been integrated into four different UAS, and has been deployed during Arctic and Antarctic science campaigns. The CULPIS elevation measurement accuracy is shown to be 95 +/- 25 cm, and is limited primarily by GPS positioning error (u3c25 u3ecm), aircraft attitude determination error (u3c20 u3ecm), and sensor misalignment error (u3c20 u3ecm). The relative error is considerably smaller over short flight distances, and the measurement precision is shown to beprecision, the CULPIS is well suited for measuring sea ice topography, and observed ridge height and ridge separation distributions are found to agree with theoretical distributions to within 5%. Simulations demonstrate the inability of course-resolution measurements to accurately represent the theoretical distributions, with differences up to 30%. Future efforts should focus on reducing the total measurement error tochange.