Glaciofluvial Processes and Sediment Yield
Glacier coverage is a well-known control of sediment yield (Hallet et
al., 1996; Meade et al., 1990), and heavily glaciated catchments in
Alaska have been found to produce an order of magnitude more sediment
than other Alaskan catchments (Hallet et al., 1996). Against this trend,
we found that specific sediment yields (SSYs) for Chamberlin Creek (23%
glacier coverage) were similar to Carnivore Creek (10% glacier
coverage) in 2015, and less in 2016 (Table 5). Our results are
comparable with Gurnell et al. (1996), who also found an inverse
relation between SSY and glacier coverage for Alaskan catchments. If the
rates of glacier recession and glacier thermal regimes are consistent
between sub-catchments, this signifies the importance of non-glacial
processes. Non-glacial sediment sources (i.e. proglacial extra-channel
and hillslope sources) eroded during rainfall events are likely
significant, as reported for Matanuska Glacier in southern Alaska
(O’Farrell et al., 2009). At Lake Peters, more exceptional sediment
delivery from Carnivore Creek during high discharge events dominates the
yield to Lake Peters, despite lower glacier coverage than Chamberlin
Creek (Table 4; Figure 4). Conversely, at low flow, Chamberlin Creek’s
steeper slopes result in persistently turbid water on the alluvial fan
(63 NTU average for Q < 0.325
m3s-1), compared with relatively
clear water in Carnivore Creek (30 NTU average for Q < 15
m3s-1), although low flow delivery
is a meager proportion of the sediment yield.
Glacier processes may enhance sediment delivery in Carnivore Creek under
the current hydrological regime. Ellerbrook (2018) report that old water
(glacier melt and subsurface flow) contributed a higher proportion of
the hydrograph than rainfall in the Carnivore sub-catchment, whereas
rainfall dominated over old water in the Chamberlin sub-catchment. If
the glaciers have surface-bed connections, it is possible that more
intense rainfall over the Carnivore glaciers contributes to erosive
glacier processes, compared with Chamberlin sub-catchment, which is a
more isolated massif. Subglacial conduits may have melted by the time
the most peaked rainfall events occur (mid-July to early-August),
supporting subglacial erosion and enhancing hydrological response (e.g.
Bogen & Bønses, 2003; Gurnell et al., 1996; Hodson & Ferguson, 1999;
Hodson et al., 1997). The nearby polythermal McCall Glacier
(~50 km west of Lake Peters) was found to have a zone of
basal sliding, and moulins—likely transferring surface meltwater to
the glacier’s base, but a complex subglacial drainage network was
probably not active (Pattyn et al., 2009). Although the contemporary
subglacial network at Lake Peters has not been studied, Benson et al.
(2019) relates millennial-scale changes in sediment accumulation and
other sediment properties in Lake Peters to large-scale glacial
fluctuations and other hydro-climatic Holocene trends, and note that
increased accumulation rates during the last century may reflect
contemporary glacier retreat.