| Summary: | Trapridge Glacier, a surging glacier located in the St. Elias Mountains, Yukon Territory, Canada, went through a complete surge cycle between 1951 and 2005. Air photos (1951-1981) and groundbased optical surveys (1969-2005) are used to describe the modifications in flow and geometry that occurred over this period. The acceleration of flow during the surge is detected by repeated measurement of poles drilled into the ice. Between 1974 and 1980, the median velocity in the lower basin went from 15.6ma⁻¹ to 38.6ma⁻¹. Downstream from this zone of active flow, coldbased ice accumulated during the previous surge impeded the flow, and a steep front formed at the boundary between the two ice masses. Over the following ten years, this bulge propagated downglacier, advancing faster than the ice and integrating stagnant ice by continuous deformation. After it peaked in 1984, the flow in the lower basin remained above 25ma⁻¹ for 12 years, but was on a slowing trend. The slowdown followed strangely regular 4-year pulses: 1-2 years of timid acceleration (2-3ma⁻¹) , followed by 2-3 years of rapid deceleration (4-8ma⁻¹) . The 1997-1999 acceleration was particularly vigorous, as the median velocity went from 20.3ma⁻¹ to 28.5ma⁻¹. After this last pulse, the glacier gradually slowed down to pre-surge velocities. In 2005, the lower basin was flowing at less than 8.5ma⁻¹. Based on this flow history, I divide the 1969-2005 period into three parts: the activation phase (1969-1974), the surge (1980-1999), and the waning phase (1999-2005). To quantify the geometrical changes that occurred during each phase, digital elevation models are constructed from air photos and optical survey measurements. Optical and radar surveys are joined to photogrammetric measurements of the proglacial field to obtain a bed topography map. DEMs for 1951, 1970, 1972, 1977, and 1981 are generated by stereographic analysis of air photos. These background models are then updated year after year by ground-based survey data, using a Bayesian kriging algorithm. By assuming the ice surface to be similar in shape from one year to the next, only shifted vertically, this algorithm allows the propagation and diffusion of information on the ice surface topography. The interpolation obeys the data, and follows the prior model where no data are available. Uncertainty is attributed to each estimation, based on the estimated covariance structure of the field and on the uncertainty of the prior model. The assumption of stationarity and the parametrization of the covariance are tested by the analysis of orthonormal residuals. Despite the significant changes in surface area that occurred between 1969 and 2005, the total volume of ice remained relatively constant at 0.14 ± 0.03 km³. The spread of the glacier during the surge was accompanied by a corresponding decrease in thickness. From the surface slope and ice thickness fields, the stress driving the flow is calculated. After filtering the surface slope to eliminate small-scale variations, the shallow ice approximation is used to estimate creep velocities from the stress and ice thickness fields. Deformation of the ice is shown to accounts for a negligible portion of the flow observed in the lower basin. Sliding or bed deformation must therefore be responsible for most of the motion in that area. In the upper basin, thicker and flowing more slowly, creep accounts for maybe as much as half of the recorded motion. Science, Faculty of Earth, Ocean and Atmospheric Sciences, Department of Graduate
|
|---|