Rocky Mountain (63rd Annual) and Cordilleran (107th Annual) Joint Meeting (18–20 May 2011)

Paper No. 5
Presentation Time: 8:00 AM-6:00 PM

DISCHARGE ESTIMATES FOR A GLACIAL OUTBURST PALEOFLOOD ON THE UPPER ARKANSAS RIVER, COLORADO, FROM AN ICE-DAM FAILURE MODEL


BRUGGER, Keith A.1, LEONARD, Eric M.2, LEE, Keenan3 and BUSH, Meredith A.2, (1)Geology Discipline, University of Minnesota, Morris, 600 E. 4th Street, Morris, MN 56267, (2)Department of Geology, Colorado College, Colorado Springs, CO 80903, (3)Dept. of Geology, Colorado School of Mines, Golden, CO 80402, bruggeka@morris.umn.edu

Glaciers advancing down eastern flank of the Sawatch Range in the Clear Creek and/or Pine Creek valleys dammed the upper Arkansas Valley several times during the Pleistocene, impounding waters of Three Glaciers Lake. Evidence for the lake consists largely of preserved shoreline gravels, some lacustrine sediments, and presumed ice-rafted boulders. During late Pleistocene phases (ca. 19.3 and 17.8 ka; Young et al., 2011), lake level was at 2855 m, about 150 m above the modern channel of the Arkansas River, corresponding to a volume of 1.72 x 109 m3. Lake volume during a possible higher, middle Pleistocene stand might have been as much as 3.77 x 109 m3. Catastrophic drainage of Three Glacier Lake is indicated by flood terraces, terrace gravels, and large flood boulders 1.5 to more than 10 m in (mean) diameter. Based on ice deployment and glacier terminus geometry, the mode of ice-dam failure is likely to have been a subaerial breach wherein outflow from the lake is controlled by ice-marginal drainage. Numerical simulations of the time-dependent evolution of breach channel geometry yield flood hydrographs and thus provide estimates for peak discharges and flood duration. Using a midrange value for channel roughness, peak discharge for the late Pleistocene flood(s) is estimated to have been on the order 46,000 m3 s-1, achieved about 10 hrs after initiation of the breach. This estimate is in reasonable agreement with those obtained using empirical relationships between lake volume and discharge during dam failure. The entire paleoflood event (lake emptied) is complete in less than a day. Modeling results are not sensitive to lake temperature. However, allowing for differences in channel roughness broadens the range of estimated peak discharges to between 21,000 and 65,000 m3 s-1.