Lake C2 is drained by one well defined outlet stream, which flows westward 230 m into Taconite Inlet. (IMAGE: Lake C2 and the outlet) The outlet channel is approximately 5-8 m wide leaving the lake. The stream drops 1.2 m through unconsolidated till and marine sediments to the inlet, a gradient of 0.3 degrees, or 5.2 m/km.
The outlet was occasionally visited, primarily in 1990, to obtain measurements of discharge, suspended sediment concentration, and electrical conductivity. The objectives of these measurements were to assess in a general way how the outlet influenced the lake water level, lake chemistry, and to determine the effectiveness of the lake as a sediment trap.
Discharge measurements and calculations at the outlet followed the same procedure as for the stream. All velocity measurements were done with the Swoffer meter. The verticals were 50 cm apart for all measurements, across a very stable cross section. The spacing of verticals resulted in some segments comprising 10-20 percent of total discharge. Seven measurements were made in 1990, and three in 1991.
A water-stage recorder was operated for 18 days during the 1991 field season (23 June to 10 July). The stilling well was located at the far upstream end of the outlet. This allowed a determination of the phase lag between inflow and outflow from the lake, and the relationship between lake level and outflow.
The salinity, electrical conductivity, and temperature of water in the outlet were measured 4 times in 1990. This was done with the YSI model 33 meter, in the stream channel. These measurements were made at relatively high discharges (2.1-3.0 m^3 s^-1).
Samples of outlet water were taken for determination of suspended sediment concentration. These were obtained by both grab sampling and depth-integration, following the methods discussed in the procedure section. Samples were taken four times in 1990 and twice in 1991. At all times, including 1992 observations, the water flowing through the outlet appeared to be extremely clear.
A staff gage was installed in Lake C2 at the edge of the delta to measure the height of the lake surface. Observations were made to investigate the influence of inflowing streamwater on lake level fluctuations, outlet discharge, and changes in the water volume of the lake.
Measurement of lake level was done in all three years by manually reading the staff gage. The initial site of the gage, installed in 1990 while the lake was frozen, became unreliable in early July due to downslope movement. The gage was moved to a new site on July 14. The staff gage was securely reset at this site on July 21, and heights were referenced to a benchmark on large boulder by leveling. Height differences between the staff gage and the benchmark were measured six times over the three years and were always within 1 cm, permitting interannual lake level height comparisons.
Lake level observations were made at irregular intervals, generally in the morning and evening, by several different field personnel. There were a total of 41, 106, and 88 measurements made in the years 1990-92, respectively.
Through the 1991 and 1992 field seasons, samples were taken of all waters potentially contributing to streamflow from the watershed. These samples included summer precipitation at Delta and Echo, bulk snowpack cores from several locations, glacier ice, and actual meltwater - where the source could be located and confidently identified (e.g. snowmelt, thawing soil moisture). In addition, streamwater samples were regularly collected through each field period. One sample each was taken from the Lake C2 outlet, one day after flow began, and from 3 m water depth in Lake C2 itself. The objective of this sampling was to assess the relative differences in stable isotope compositions, and given significant differences, undertake an evaluation of the proportional contributions to streamflow.
The desire to delineate relative contributions to streamflow stemmed from uncertainty about the role of rainfall at the field site, both as a source of streamwater (runoff), and as a control on sediment flux from the watershed. A minor rainfall event in July 1990, followed by a significant hydrologic response, suggested rainfall was indeed hydrologically important, even at 83°N. The subsequent collection of samples through the next two field seasons was done anticipating the recurrence of such an event.
Summer precipitation samples were obtained from both the Delta and Echo precipitation gages. The collection interval varied, although was typically daily at the Delta site, during precipitation events. Longer intervals occurred between collection of Echo samples. Precipitation inside the gage was effectively isolated from the atmosphere by the small orifice between the tube and collection funnel above. Nonetheless, the volume of air inside the gage collection tube was considerably greater than the volume of precipitation collected (e.g. 25:1 if 1 mm precipitation). As a result, the influence of evaporation, and consequent isotopic fractionation, on the isotopic concentration of samples is not known.
Sampling of the snowpack was done to provide a bulk measure of the isotopic composition. In 1991, a composite core was taken from each site with a Snowmetrics density sampler (1000 cm^3). The cutter was rotated 90° from the standard use position, so as to obtain a profile with uniform vertical volume. The snow was placed into a plastic sample bag, and sealed following evacuation of air to the extent possible. One profile was taken at each site. During the 1992 field season, snowpack cores were taken with an acrylic tube, approximately 7.5 cm in diameter. Three cores were taken at each site whenever samples were collected, and bagged separately, using the same method as in 1991. In both years, snow cores were melted in the sample bags, and following agitation to insure adequate mixing, a 30 mL subsample was collected.
Four samples of glacier ice were collected, in July of 1991. A 5 cm CRREL auger was used to drill into what appeared to be glacier ice (rather than superimposed ice), after removal of 5 to 20 cm of snow. The four sites were at elevations between 500 and 550 m, and probably within the ablation zone of the glacier most summers. Drill cuttings were placed in sample bags, as done with snow samples.
Samples were also taken of meltwater from snow and soil. Snowmelt was sampled for comparison with bulk snow samples, and was generally easy to locate, dripping or flowing from the snowpack. The melting active layer samples were taken in 1991, on the south slope of Echo Peak. Collection was made in areas free of snow cover, among talus or tussocks of vegetation. In several cases, segregation ground ice was visible at the collection site.
Streamwater samples were taken at the gaging station site generally twice per day, close to the daily discharge minima and maxima. Additional samples were collected during important periods, including the initial slushflow and high discharges. Typically, the filtrate from processed suspended sediment samples was used.
Water samples were all collected in 30 mL Nalgene Brand HDPE Environmental Sample Bottles. The bottles were filled nearly to capacity, sealed, and stored for transport in a dark box at temperatures generally between 0° and 5°C. Upon return from the field, samples were kept refrigerated at a constant 5°C.
The water samples from all sources were analyzed under the supervision of Dr. Jim White, at the Stable Isotope Laboratory, Institute of Arctic and Alpine Research, at the University of Colorado, Boulder. Oxygen isotope values are in per mille versus SMOW, with a reproducibility of ±0.06 per mille.
Bedload measurements were made during the 1991 field season to assess the importance of bedload transport relative to that of suspended sediment, and to provide a rough indication of how bedload transport varied with discharge. In 1990, bedload was not measured due to lack of equipment. Measurements were not continued in 1992 primarily because personnel time was devoted to higher-frequency sampling of suspended sediment. In addition, extreme spatial variability of bedload transport was observed in 1991 on the Lake C2 delta, raising questions about the significance of point measurements. There are numerous difficulties associated with direct measurement of bedload transport (cf. Edwards and Glysson, 1988; Yuqian, 1989).
Measurements of 1991 bedload transport were all made at points across the cross section, at the stream gaging station. The tag line installed for discharge measurement was used to establish lateral position.
A Helley-Smith bedload sampler, of the original design, was used with a wading rod for all measurements. The sampler has a 7.62 by 7.62 cm opening (3 * 3 inch), and collects sediment coarser than 0.25 mm in a mesh sample bag (Style #1 from GBC, Inc.). The efficiency of the sampler has been shown to vary between 100 and 150 percent depending upon the transport rate and particle size (Hubbell et al., 1985).
At Taconite Inlet, bedload transport was initially measured at several locations across the stream. Sampling efforts were then shifted to obtain three samples at each sampling time, all at the same vertical. The sampled vertical was selected to represent the most favorable site for bedload movement, because observations indicated that the rough gravel channel bed was concentrating bedload movement in a narrow "thread" (cf. Bathurst, 1987 p. 275; Warburton, 1992 p. 198). Samples were collected for a duration of 1 or 2 minutes. Samples were taken at intervals dependent upon the amount of material in motion, up to 12 times daily.
Bedload samples were initially drained of water in the collection bags, then air dried in plastic containers. Dry weighing was done on a Fisherbrand model S-400G electronic balance, with a manufacturers stated linearity of +/- 10 mg. In general, one of the three samples taken at each sampling time was retained for further analysis.