Joseph Walder | United States Geological Survey (original) (raw)

Papers by Joseph Walder

Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, 1993

Arctic and Alpine Research, 1986

... 309. 9 pp. Lautridou, JP and Ozouf, JC, 1982: Experimental frost shattering: 15 years of rese... more ... 309. 9 pp. Lautridou, JP and Ozouf, JC, 1982: Experimental frost shattering: 15 years of research at the Centre de Geomorphologie du CNRS. ... McGreevy, JP, 1981: Some perspectives on frost shattering. Progress in Physical Geography, 5: 56-75. ...

Water Resources Investigations Report, Jun 30, 2006

Agu Fall Meeting Abstracts, Dec 1, 2001

The theoretical foundations of glacier hydrology deal almost exclusively with idealized steady-st... more The theoretical foundations of glacier hydrology deal almost exclusively with idealized steady-state flow. Röthlisberger and Shreve, in their seminal papers on water flow through channels surrounded by ice, showed that in the steady state there exists an inverse relationship between water pressure and discharge, and concluded that this would favor development of an arborescent drainage system. Steady-state basal linked cavities, in contrast, should (as shown most clearly by Barclay Kamb) exhibit characteristics of what is known in current parlance as a "distributed" drainage system: a direct relationship between water pressure and discharge, no tendency to develop arborescence, and relatively slow water flow. The two theoretical constructs--steady-state arborescent- and distributed drainage systems--have dominated the interpretation of field data. Yet no temperate glacier is in a steady state hydrologically: meltwater runoff rate commonly varies by a factor of 3 or 4, say, on a diurnal basis, and by orders of magnitude on a seasonal basis. The "success" of a classification scheme grounded in steady-state theory thus seems paradoxical and suggests that some key piece of physics has received short shrift. In our view, the interaction of qualitatively different bits of the basal drainage system--between arborescent- and distributed-like portions--in both time and space should be more thoughtfully regarded. We already know something of the time-dependent interaction: Barclay Kamb showed that a linked-cavity system is unstable to large pressure perturbations and would transform to an arborescent system; an arborescent system in turn is likely to collapse when discharge plummets during the accumulation season. These instabilities probably have a great deal to do with the evolution of the drainage system during the course of the ablation season and with the abrupt release of water from storage. Our own work on the drainage of glacier-dammed lakes casts light on interactions of drainage-system components in space. The picture that emerges is that of a "patchy" basal drainage system, comprising at the height of the ablation season an arborescent channel system poorly linked to a distributed system that covers most of the glacier bed. The patchy drainage system can accommodate substantial variability in discharge and pressure--variability that cannot be accommodated if the drainage system consists wholly of either arborescent channels or linked cavities.

Proceedings of the Royal Society a Mathematical Physical and Engineering Sciences, 1993

We examine theoretically the creep closure of subglacial tunnels cut into basal till, generalizin... more We examine theoretically the creep closure of subglacial tunnels cut into basal till, generalizing Nye's classical analysis of tunnel closure in glacier ice to rheologies in which the creep rate depends on effective pressure (the difference between total pressure and pore-water pressure). The solutions depend critically on a dimensionless permeability parameter. For the appealingly simple Boulton-Hindmarsh rheology in which strain rate depends on powers of applied stress and effective pressure, solutions to the closure problem may not exist; this is related to the existence of a `failed' zone next to the channel, where piping occurs, and also to a non-physical degeneracy of the assumed rheology, whereby the viscosity is indeterminate at zero effective pressure. Consideration of the failed zone allows solutions to be obtained and shows that the closure characteristics of high permeability tills and low permeability tills are very different.

Journal of Glaciology, 1986

Jökulhlaups, sudden releases of water impounded by a glacier, produce large floods unrelated to w... more Jökulhlaups, sudden releases of water impounded by a glacier, produce large floods unrelated to weather events. We draw on observations from 90 years of annual jökulhlaups from Hidden Creek Lake, Kennicott Glacier, Alaska and from detailed monitoring in 1999 and 2000, to examine conditions that trigger outburst floods. For the class of jökulhlaups caused by subglacial drainage, the trigger must be related to formation of subglacial conduits, a pivotal problem in glaciology. Hidden Creek Lake water level at drainage has declined over the last century, during which time the glacier has thinned. The water level trend is mirrored by a tendency toward earlier release dates in the summer. These observations suggest that a minimum threshold lake level must be exceeded for drainage to occur, and that this threshold is related to ice thickness. The release date varies by over a month, however, and lake level varies by as much as 10 m over spans of a few years, which indicates that more is involved than simple exceedance of a threshold. Kennicott Glacier impounds several other small lakes. In two summers with fairly complete observations of their behavior, these lakes drained in sequence from nearest to furthest from the terminus. More frequent observations have been made of drainage of one of these lakes: Erie Lake, located roughly the same distance from the terminus as Hidden Creek Lake, along a major tributary to the Kennicott Glacier, usually drains within days of Hidden Creek Lake. These patterns are consistent with a trigger that is related to glacier-wide evolution of the hydrologic system, rather than each drainage reflecting purely local conditions. Perhaps there is a linkage between timing of outbursts and seasonal upglacier extension of the subglacial conduit system. While we have no direct information on annual evolution of the conduit system at Kennicott Glacier, some have suggested that conduits and the snowline move upglacier in tandem. In 1999 and 2000, the snowline had retreated far upglacier, beyond Hidden Creek Lake, before it drained. Estimated melt rates in the days preceding Hidden Creek Lake outbursts, calculated with a degree day model, show no pattern: lake drainage has occurred during times of both low and high melt production. Jökulhlaup triggering appears to be controlled by a number of conditions, among them the height of the ice barrier, organization of the subglacial hydrologic system, and specific, probably transient, conditions within that system at the time of drainage.

As a glacier-dammed lake fills, a large volume of water may be stored beneath the ice dam (the pa... more As a glacier-dammed lake fills, a large volume of water may be stored beneath the ice dam (the part of the glacier adjacent to the lake) and then released during a subsequent jökulhlaup. The surface of the ice dam rises and falls accordingly. At Kennicott Glacier, Alaska, the kinematics of such motion during the annual jökulhlaup cycle of Hidden Creek

Page 133. 7 Time-Dependent Hydraulics of the Earth's Crust AMOS NUR Stanford University JOSE... more Page 133. 7 Time-Dependent Hydraulics of the Earth's Crust AMOS NUR Stanford University JOSEPH WALDER University of Washington ABSTRACT The deceivingly simple question posed here is how deep free water extends in the Earth's crust. ...

Since the end of the 1981-1986 episode of lava-dome growth at Mount St. Helens, an unusual glacie... more Since the end of the 1981-1986 episode of lava-dome growth at Mount St. Helens, an unusual glacier has grown rapidly within the crater of the volcano. The glacier, which is fed primarily by avalanching from the crater walls, contains about 30% rock debris by volume, has a maximum thickness of about 220 m and a volume of about 120 million

Eos, Transactions American Geophysical Union, 2005

Text S1 to S3 Figures S1 to S17

A pyroclastic density current moving over snow is likely to transform to a lahar if the pyroclast... more A pyroclastic density current moving over snow is likely to transform to a lahar if the pyroclasts incorporate enough (melting) snow and meltwater to bring the bulk water content of the mixture to about 35% by volume. However, the processes by which such a mixture forms are still not well understood. Walder (Bull. Volcanol., v. 62, 2000) showed experimentally the

Following the cataclysmic 1980 eruption of Mount St. Helens, a lava dome grew in the crater until... more Following the cataclysmic 1980 eruption of Mount St. Helens, a lava dome grew in the crater until 1986, at which time the crater contained only discontinuous patches of perennial snow and ice. By Sept. 2004, when a new dome began to grow, the crater hosted a glacier with a surface area of ~1 km2 and a thickness locally as much as ~200 m. This extraordinarily rapid growth occurred because the glacier sits in a north-facing amphitheater-like basin and is fed by avalanches from the crater walls. During the first two months of the 2004-05 eruption, the glacier was cut in half as the new lava dome grew toward the south crater wall. Subsequently this lava dome grew toward the east, with the dome/glacier contact migrating at a rate of ~1 m/d. The east crater glacier was thus squeezed between the growing lava dome and the east crater wall. East/west trending crevasses formed and the previously smooth surface bowed upward. The intense squeeze caused ice thickness locally to more than double ...