Fluvial and hillslope geomorphology of Hoseanna Creek Watershed, central Alaska

Abstract

Hoseanna Creek Watershed is rapidly eroding and provides excellent opportunities to describe and quantify hillslope and fluvial processes in the subarctic and in discontinuous permafrost terrain. High landslide and badland densities occur due to asymmetric geologic structure and weakly consolidated lithologies. Late Quaternary regional glaciofluvial processes and tectonism have changed local base level at least 100 m, inducing headward incision through weak lithologies and yielding high rates of sediment production. Earthflows, translational blocks, rotational blocks, lateral spreads or complex landslide types form in coal-bearing formations in response to lateral corrasion of toes by avulsing streams or to undermining of foot areas by headward incising streams. Slides undergo episodic resurgent activity when new lithostatic or hydrostatic thresholds are reached. Average horizontal displacement rates of seven slides monitored between 8/85 and 9/88 ranged from 0.2 m/yr in rotational blocks to 48 m/yr in the earthflows. Although unique sliding mechanisms are not apparent, permafrost and subarctic climate generate delays or catalysts for failure atypical of warmer climates. Freezing/thawing fronts affect soil strength and permeability; break-up/freeze-up processes affect the timing of water supply to the slide mass and affect development of aufeis-related ground-water pore pressures. Aspectual and lithologic variations combine to yield three geohydrologic subbasin types which govern discharge ranges. Regressions were performed on multiple sets of sediment-discharge (Ts-Q) data. Regression variance (r$\sp2$) was found to have a maximum natural threshold indicative of intrinsic variability. Wide ranges in Q (0.001 to 2350 cfs) and Ts (0.005 to 1600 g/l) necessitated log-log scales and power functions. Each geohydrologic subbasin has a unique Ts-Q relationship termed here the mean sediment concentration potential Cp. Systematic differences in regression parameters indicate that variations in spatial conditions define Cp, while systematic changes in Ts-Q regression residuals R (termed here the maintenance rate R$\sp*$) describe the temporal variability of Ts through time with respect to Cp. 50-95% of the annual sediment load is transferred during less than 3% of the year. Erosion rate indexes were established from peak load estimates; Nenana Gravel basins are eroding 260 times faster than schistose basins and ten times coal-bearing basins.