Integrating uncertainty: Canyon Creek hyperconcentrated flows of November 1989 and 1990

Abstract

Canyon Creek drains a 79 km 2 watershed in northwestern Washington State. Extensive logging occurred from the mid-1960s to 1980s, which resulted in numerous slope instabilities and a several order of magnitude increase in sediment supply to the creek. On November 9, 1989, a hyperconcentrated flow with a peak discharge of 450 m 3 /s destroyed one house on the fan. A forensic investigation of the event suggests that a temporary landslide dam may have formed at two coalescing earthflows about 4 km above the fan apex. The 1989 hyperconcentrated flow caused significant aggradation on the fan. One year later to the day, a significant flood occurred, which ran over the aggraded fan surface from the 1989 event. This latter event destroyed four more homes mostly through bank erosion and rendered a section of county road impassable. FLDWAV, a flood routing model capable of simulating unsteady flow conditions, was used to model landslide dam breaches for a number of different dam heights at the earthflows. Modeling results were then combined with historic air photograph interpretation, dendrochronology, and eyewitness accounts to construct a frequency-magnitude relationship for hyperconcentrated flows at Canyon Creek. FLDWAV results were combined with a hyperconcentrated flow runout model (FLO-2D) on the fan to estimate maximum flow depth and flow velocity for the design event, a 500-year return period with a predicted peak discharge of 710 m 3 /s. A large range of mitigation measures were reviewed, but it was concluded that buy-outs would be the most effective risk reduction measure. Property acquisition commenced in 2004.

Unprecedented rates of landslide and surface erosion along a newly constructed road in Yunnan, China

“…Both stream reaches are shown in Fig. 1 less likely scenario, a major flooding event will provide ample energy to transport the majority of the accumulated sediment and debris resulting in a large debris flood or hyperconcentrated flow (e.g., Slaymaker 1988;Yumuang 2006;Jakob and Weatherly 2008). Quantitative investigations into sediment and debris transport in this channel together with probabilistic assessment of flood frequency are necessary to address this environmentally important question.…”

A Multi-Faceted Debris-Flood Hazard Assessment for Cougar Creek, Alberta, Canada

“…This realization necessitates some broad assumptions, or resorting to approaches that do not rely solely on sediment transport calculation. This degree of uncertainty has been described numerous times in the scientific literature (i.e., Jakob [4], Rickenmann [10], Jakob and Weatherly [66]). Similarly, the frequency of events that are neither gauged, nor reliably documented temporally requires a combination of methods, each of which is associated with their own benefits and challenges and none of which would provide reliable results in isolation.…”

“…For example, a recently aggraded channel section when followed by a high flood event will lead to tree damage even if that flood event was of lesser magnitude than the preceding one. An example of this cycle in aggradation and degradation and its hazard potential has been reported in Jakob and Weatherly [66] at Canyon Creek, Washington County, US.…”

Numerical simulation of the inundation area for landslide-induced debris flow: a case study of the Sha-Xinkai gully in southern Taiwan

“…The FLO-2D [4] routing model is software designed for two-dimensional mathematical modeling of water movement and fast flowing slope processes including debris flows. The FLO-2D model has been used successfully for debris flow simulations by many researchers in a variety of countries (Lin et al [5]; Tecca et al [6]; Sosio et al [7]; Stolz et al [8]; Jakob and Weatherly [9]; Hsu et al [10]; Sodnik and Mikos [11]). Data required for model simulations include a digital terrain model, an inflow hydrograph, rheological properties of the sediment water mixture, and the Manning roughness coefficient.…”