OCR Text |
Show and Gilbert (- 1300 m) shorelines and these patterns have been used to estimate the viscosity of the upper mantle in the Eastern Great Basin. A very simple model, which supposes that an elastic layer overlies a uniform visco- elastic substrate, is able to fit the observations quite well. The inferred elastic layer thickness is ( 23 ±_ 2) km, and the substrate viscosity is - 10 ( 19" 5 ± 0J) Pa- s, substantially lower than the 1021 Pa- s value which has been inferred from glacial rebound studies, and is often represented as a global average. More detailed models allow some resolution of the variation in viscosity with depth. A robust feature of such models is an extremely low viscosity region ( 1017- 1018 Pa- s), extending from 40- 150 km depth. These values are anomalously low for continental settings, and are as low as those found beneath Iceland. At greater depth under Lake Bonneville, the viscosity increases to values close to the 1021 Pa- s nominal upper mantle value. If separate estimates of viscosity structure are extracted from the rebound patterns seen on the Bonneville and Provo shorelines, using any of the published lake level chronologies, discrepant estimates are obtained. Since the time scale for significant change in viscosity is likely 107 years or more, this suggests that the input chronologies are in error. It appears that adjustment of the chronologies, within their present error bounds, will yield concordant viscosity estimates from the separate shorelines. In addition to the rebound signal, there is evidence of a regional tilt signal in the Bonneville basin shoreline elevations. When the best fitting rebound model predictions are subtracted from the observed heights, the Bonneville, Provo and Gilbert shorelines all exhibit a nearly planar tilt down to the northeast. The east and north components of the tilt, in units of cm/ km, are: Bonneville (- 4.9, - 3.2), Provo (- 4.1, - 5.3), Gilbert (- 2.7, - 1.9). The cause of this tilt is still quite uncertain. Lake Lahontan was the second largest of the Great Basin pluvial lakes. At its greatest extent it had a surface area of 25,100 km2 in a drainage basin of 110,900 km2', which it shared with several smaller lakes, all of which were spilling into Lahontan. The largest of these satellite lakes were Tahoe ( 550 km2) to the west, Kumiva ( 300 km2) and Granite Springs ( 400 km2) on the big island in the middle of lake Lahontan, and Diamond ( 800 km2) situated far to the east and spilling as a tributary to the Humboldt river. For most of the late Pleistocene, the lakes which eventually coalesced to form Lake Lahontan were operating as hydrologically independent entities. This is much more so for Lahontan than Bonneville. The isostatic rebound pattern associated with Lake Lahontan has only recently been studied in any detail. The primary reason is likely that the amplitude is considerably less (- 20 m amplitude) than in the Bonneville basin. Using a recent compilation of 170 height measurements on the highest Lahontan shoreline, and previously published lake level chronologies, an initial investigation of viscosity structure in the eastern Great Basin has been attempted. As with the Bonneville data, a simple two parameter model provides a reasonably good fit to the data. However, the results are somewhat surprising. The effective elastic plate thickness is ( 28 +. 4) km, and the viscosity of the substrate is 10 ( 18* 7 ± 03) Pa- s. The lower value of viscosity in the western Great Basin is consistent with previous observations of higher crustal heat flow, more |