Cordilleran Section (104th Annual) and Rocky Mountain Section (60th Annual) Joint Meeting (19–21 March 2008)

Paper No. 3
Presentation Time: 9:00 AM

DEBRIS FLOWS IN SOUTHEASTERN ARIZONA: HOLOCENE FREQUENCY OF EXTREME EVENTS


WEBB, Robert, GRIFFITHS, Peter and MAGIRL, Christopher, U.S. Geological Survey, 520 N. Park Ave, Tucson, AZ 85719, rhwebb@usgs.gov

Deposits near the apices of alluvial fans emanating from some mountain ranges in southeastern Arizona indicate that debris flows have periodically occurred from at least the latest Pleistocene until the present. Before 2006, few debris flows were known to have occurred historically in this region. On July 31, 2006, following a week of relatively heavy rainfall, more than 500 hillslope failures in the Santa Catalina Mountains spawned numerous debris flows. Flows from five canyons reached or exited the mountain front to damage structures, roads, and other infrastructure north of Tucson. The July 31, 2006, debris flows caused considerable channel aggradation; upstream from the snout, poorly sorted sediment dominated by cobbles and boulders aggraded one channel by 5-7 m, and runout facies downstream caused sedimentation that limited flood conveyance. Although total storm precipitation and intensity on July 31 were not unusual for a summer storm, precipitation during the preceding 5-7 days was high, creating essentially saturated conditions on hillslopes. Preliminary estimates of storm and flood frequency indicate a return periods >1000 years over the seven day period in some locations. We used conventional 14C as well as the cosmogenic 10Be to date debris-flow deposits on alluvial fans, which provides some indication of the frequency of events that reach developed land. Because we expected that some deposits may be late Holocene, and because the 10Be production rate at 32º N latitude and 900 m elevation is relatively low, we developed a custom, high-purity 9Be carrier from phenacite for our youngest samples. Our results indicate that the largest, fan-forming debris flows, which left the most spatially extensive and highest level deposits, occurred in the latest Pleistocene <13 ka. These debris flows likely initiated as large, deep-seated landslides, whereas the 2006 events began as relatively shallow slope failures of limited spatial extent. Debris-flow deposits at lower elevation with respect to the channel were mostly Holocene, with the youngest dated at 570±100 14C years BP. Our work indicates that large particle transport into channels is a highly infrequent event that determines subsequent channel geometry over long periods in this semiarid environment.