influence of spatial variability in snowmelt and active layer thaw on hillslope drainage for an alpine tundra hillslope

• Published in: Hydrological processes Vol. 23; no. 18; pp. 2628 - 2639
• Main Authors: Quinton, William L, Bemrose, Robert K, Zhang, Yinsuo, Carey, Sean K
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley '' Sons, Ltd 30.08.2009; John Wiley & Sons, Ltd
• Subjects: active layer thaw; modelling; permafrost; snowmelt; sub-surface runoff
• ISSN: 0885-6087; 1099-1085
• DOI: 10.1002/hyp.7327

In alpine tundra, hillslope drainage occurs predominantly below the ground surface, between the relatively impermeable frost table and the water table above it. The saturated hydraulic conductivity decreases by two to three orders of magnitude between the ground surface and the top of the mineral substrate at 0·2-0·3 m depth. Consequently, the rate of sub-surface conveyance from hillslopes strongly depends on the degree of active layer thaw. In addition, although rarely examined, the volume and timing of hillslope drainage and streamflow in alpine tundra are strongly influenced by the spatial pattern of active layer thaw in the runoff-contributing hillslopes. The spatial and temporal pattern of soil thaw was modelled on a 25 535 m² area of interest (AOI) on a north-facing alpine tundra hillslope in southern Yukon, Canada. Daily oblique photographs were used to pixilate the AOI and measure the snow cover depletion for the AOI. Transects of thaw depth through selected snow-free patches within or adjacent to the AOI were used to provide spatial representation of thaw depth and saturated hydraulic conductivity. To compute thaw depth for each pixel within the AOI as it became snow-free, a finite element geothermal model was driven from meteorological forcing and soil hydro-thermal data. Daily maps of snow-free area and thaw depth were generated for the AOI, which were then combined with information on flowpath tortuosity and depth-integrated saturated hydraulic conductivity for snow-free areas to provide a drainage rate for the AOI. This drainage rate varied with the spatial arrangement of the snowpack and soil thaw depth. At the beginning of melt, the lack of hydrological connectivity for the AOI inhibited drainage. The drainage rate was maximized during the mid-melt period since sub-surface connections among snow-free patches were widespread, and the hydraulic conductivity of the sub-surface flow zone was still relatively high, owing to shallow soil thaw depths. Drainage was low at the end of the melt period because the frost table was at a depth where the hydraulic conductivity was very low. The results of this study indicate that to accurately simulate runoff at the hillslope scale during snowmelt, both the spatial pattern of snowmelt and ground thaw should be considered. Copyright © 2009 John Wiley '' Sons, Ltd.