Thomas S. James, John J. Clague, Kelin Wang, Ian Hutchinson

GSC, Contribution Series 1998125

A postglacial rebound model is developed to explain tilted shorelines of proglacial lakes and relative sea level change at the northern Cascadia subduction zone. The modelling shows that mantle viscosity in this tectonically active region, which features a relatively youthful oceanic plate (6 Ma at the trench), is less than 10²⁰ Pa s. As a consequence of the low mantle viscosity and resulting rapid isostatic recovery, present-day postglacial rebound uplift rates are predicted to be less than 0.1 mm/yr, in contrast to earlier models predicting rates of a few millimeters per year. The relative sea-level observations require rapid collapse of remaining marine-based portions of the ice sheet around 12.5 to 12.25 ka.

The Cascadia subduction zone is a region of enhanced seismic hazard. Consequently, it is the focus of a variety of geodetic investigations, including continuous GPS monitoring. Before crustal motion observations can be interpreted in terms of plate-boundary-related crustal deformation, however, it is necessary to estimate and remove the signal from other processes that may produce detectable crustal motion. Surface loading from ice sheets (glacial isostasy) can potentially produce vertical crustal motion at the millimeter per year level.

A glacial rebound model is developed to explain the crustal tilting and uplift that occurred at the northern Cascadia subduction zone during retreat of the Cordilleran ice sheet at the end of the Pleistocene. A good fit to the observations is obtained with mantle viscosity values ranging from 5 x 10¹⁸ Pa s to 5 x 1019 Pa s. This is consistent with studies of crustal deformation following subduction zone earthquakes, but is ½ to 2 ½ orders of magnitude smaller than the viscosity typically attributed to the upper mantle underlying tectonically less active regions (1020 to 1021 Pa s). Owing to the low viscosity values, our model predicts present-day crustal uplift rates of the order of 0.1 mm/yr or less, much smaller than the predictions of a global postglacial rebound model that were previously used to adjust geodetic observations from northern Cascadia.

Tectonic setting larger image [GIF, 49.1 kb, 600 X 408, notice]

Tectonic setting of the northern Cascadia subduction zone, where the Juan de Fuca plate subducts beneath North America. Data analyzed in this study include tilts of proglacial lake shorelines in Puget Sound (profiles RH and Br) and relative sea level observations at Courtenay and Parksville.

Ice sheet baseline model larger image [GIF, 97.5 kb, 600 X 520, notice]

The baseline ice sheet model developed in this study at the glacial maximum about 14 ka (heavy line). Ice sheet thicknesses (m) are given for each grid element. Model thicknesses are given at 13 other times between 25 and 10.5 ka. The large circular disks are for ICE-3G (Tushingham and Peltier, 1991), a postglacial rebound model previously used to correct geodetic observations in the region.

Sketches of the effect of a glacial load on the Earth:

Peak glaciation larger image [GIF, 11.2 kb, 600 X 296, notice]

(a) At peak glacial conditions the Earth's surface is depressed beneath the ice sheet and slightly elevated outside the ice sheet owing to mantle flow.

Deglaciation larger image [GIF, 11.6 kb, 600 X 308, notice]

(b) During deglaciation the depressed region rises and peripheral regions subside. Uplift of the Earth's surface is frequently observed as relative sea level fall in recently deglaciated areas. During the initial stages of recession proglacial lakes formed in the Puget Lowland. Shoreline features of these lakes are now tilted up to the north.

Lake location map larger image [GIF, 43.1 kb, 600 X 786, notice]

(a) Location of Lake Russell-Hood (triangles) and Lake Bretz (diamonds) proglacial lake shoreline sites in the Puget Lowland (Thorson, 1989). The projection of the Seattle thrust fault is shaded.

Lake Russell Hood linear regression larger image [GIF, 16.0 kb, 400 X 445, notice]

(b) Linear regression (filled triangles and solid line) showing the tilt of present-day shoreline elevations for Lake Russell-Hood. Also shown is the linear regression after northern Lake Russell-Hood sites are corrected for possible movement on the Seattle fault (open triangles and dotted line).

Lake Bretz linear regression larger image [GIF, 13.3 kb, 400 X 444, notice]

(c) Linear regression for western Lake Bretz.

Crustal displacement and tilt larger image [GIF, 35.9 kb, 500 X 647, notice]

(a and b) Crustal displacement and (c and d) crustal tilt resulting from a circular disk of ice whose size and duration of loading is chosen to approximate that of the southwestern Cordilleran ice sheet. In the left-hand panels the mantle viscosity is 10¹⁸ Pa s and the lithospheric thickness is varied, while in the right-hand panels the lithospheric thickness is 35 km and the mantle viscosity is varied. It is difficult to obtain the observed tilt of about 1 m/km if the lithosphere is too thick or the mantle viscosity too large.

Predicted shoreline tilts larger image [GIF, 36.1 kb, 600 X 400, notice]

Predicted shoreline tilts (in m/km) for (a) Lake Russell-Hood and (b) western Lake Bretz for a range of mantle viscosity values and lithospheric thicknesses for the detailed ice sheet model. Observed tilts, shown as heavy lines, are 0.85 m/km for Lake Russell-Hood and 1.15 m/km for western Lake Bretz (see figures b and c on proglacial shorelines). Point A indicates the baseline solution satisfying both lake tilts and Point B shows the solution when the effect of slip on the Seattle Fault is included (see figure b on proglacial shorelines).

Mantle viscosity larger image [GIF, 40.1 kb, 500 X 636, notice]

Mantle viscosity and lithospheric thickness results of varying a number of model assumptions. The shaded region indicates the range of likely solutions, taking into account combinations of various effects. Seismicity depth ranges are given for western Puget Sound. An upper bound on mantle viscosity in this region is about 10²⁰ Pa s.

Relative sea-level larger image [GIF, 21.5 kb, 500 X 534, notice]

Relative sea-level observations at Courtenay and Parksville compared to model predictions for the baseline ice sheet and three mantle viscosity values. The sea level observations also require a mantle viscosity less than 10²⁰ Pa s.

Ice Sheet Volume History larger image [GIF, 19.8 kb, 400 X 599, notice]

(a) The ice sheet volume history, expressed as meters of sea-level equivalent, for the baseline ice sheet (solid line) and a variant in which deglaciation occurs relatively uniformly (dotted line).

Relaitve sea-level predictions larger image [GIF, 21.9 kb, 400 X 674, notice]

(b) and (c) Relative sea-level predictions at Courtenay and Parksville for the two ice sheet histories and a mantle viscosity of 5 x 10¹⁸ Pa s. Rapid deglaciation around 12.5 to 12.25 ka is needed to explain the observed sea level change.

Baseline model larger image [GIF, 39.7 kb, 500 X 475, notice]

ICE-3G larger image [GIF, 38.8 kb, 500 X 475, notice]

Predicted present-day crustal uplift (mm/yr) for (a) the baseline ice sheet and a 1020 Pa s mantle viscosity, and (b) the ICE-3G deglaciation history and a 1021 Pa s upper mantle viscosity.

A postglacial rebound model developed to explain tilted shorelines of proglacial lakes and relative sea level change at the northern Cascadia subduction zone reveals:

1. Mantle viscosity in this tectonically active region, which features a relatively youthful oceanic plate (6 Ma at the trench), is less than 1020 Pa s.

2. As a consequence of the low mantle viscosity and resulting rapid isostatic recovery, present-day postglacial rebound uplift rates are predicted to be less than 0.1 mm/yr, in contrast to earlier models predicting millimeter per year rates.

3. The relative sea-level observations require rapid collapse of remaining marine-based portions of the ice sheet around 12.5 to 12.25 ka.

Thorson, R. M., Glacio-isostatic response of the Puget Sound area, Washington, Bulletin of the Geological Society of America 101, 1163-1174, 1989.

Tushingham, A. M., and W. R. Peltier, ICE-3G: A new global model of late Pleistocene deglaciation based upon geophysical predictions of post-glacial relative sea level change, Journal of Geophysical Research 69, 4497-4523, 1991.

Dr. Thomas S. James works in the Geodynamics Section of the Geological Survey of Canada (Sidney), British Columbia.

Dr. Kelin Wang works in the Cordilleran Tectonics Section of the Geological Survey of Canada, Sidney, British Columbia.

Dr. John J. Clague works in the Earth Sciences Department at Simon Fraser University, Burnaby, British Columbia.

Dr. Ian Hutchinson works in the Geography Department at Simon Fraser University, Burnaby, British Columbia.