Northern Sweden is mainly covered by till in which small, rather shallow aquiferes are hosted. During an annual cycle, the groundwater level in these aquiferes has one well defined maximum and one minimum. Maximum groundwater level occurs at the end of the snowmelt season in May, and this is the period when the major part of groundwater recharge occurs. After maximum level is reached, the groundwater level steadily decreases until the next snowmelt season starts. The minimum groundwater level is usually found in late April, just before snowmelt. These changes of groundwater levels can on a local scale change the groundwater flow path, which in turn may result in variations of the groundwater chemistry.
In a small cathcment situated within the Kalix River watershed, northern Sweden, two 2" polyethylene wells were installed in a till slope facing a small stream. The till, supporting mainly pine and spruce forest, is unsorted and consists of granitic material. One well was placed at the top of the slope, and the other one close to the stream. Groundwater was sampled once a week from April to November 1994, and from March to July 1995, with bailer samplers. Stream water was sampled from April to November1994, and from May to July 1995. All waters were immediately filtered through 0.45 µm filters and then analysed for major elements (ICP-AES) and trace elements (ICP-MS). The sampling area is remote (N66°47' E21°49'), and little affected by anthropogenic activities.
During winter, the groundwater table undulation was goverened by a "large"-scale topography, resulting in a groundwater flow path away from the stream towards the groundwater wells. When snowmelt commenced, there was a rapid response in groundwater level. The elevation was however much greater at the upper well so that the flow path changed direction, i.e. downslope towards the stream. As the groundwater table later on was lowered, the flow path changed direction again, back into the winter direction. This occurred in August.
The concentration of most major elements was almost constant throughout the year at the upper well while there were considerable variations at the lower well. Before the commencement of the snowmelt, there were approximately equal concentrations of Ca, Mg, Na, K, and Si in both wells. At early snowmelt these elements were slightly diluted at the lower well, but by the time when the flow path changed direction the concentrations started to increase. From mid-May to early August, the concentrations increased from 1.5 to 4.8 mg/l for Ca, from 0.6 to 1.5 mg/l for Mg, from 0.6 to 2.8 mg/l for Na, and from 3.3 to 7.8 mg/l for Si. During August when the flow path returned to the pre-snowmelt direction, the concentrations decreased to the same levels as before the snowmelt. The alkalinity followed this pattern. The enhancement of the concentrations may be explained by upwelling of deeper and older groundwater discharging
into the stream. During autumn, there was probably a considerable amount of streamwater infiltrating down to the groundwater, causing the rather rapid decrease in groundwater concentrations. Silica has been used as a natural tracer for partitioning of different waters (e.g. Maulé & Stein, 1990). Such a calculation suggests that during autumn, up to 85 % of the groundwater close to the stream consists of infiltrating streamwater .
The increase in concentration of K was not as pronounced as for the other base cations and silica, but there was a clear decrease in the K/Mg ratio during growing season which might indicate uptake by vegetation. The ratio increased from October to November.
The concentrations of Fe and Al were enhanced during early snowelt as well as in July and August. There was a good correlation between Ce and Al concentrations. The highest concentrations of Cu, Co, and Ni occurred during early snowmelt (0.8, 0.3, and 2 µg/l, respectively), while the lowest concentrations were found in autumn (0.3, 0.1, and
1 µg/l, respectively). The concentrations of Ba and Sr followed the concentrations of K and Ca, respectively.
Maulé, C. P. & Stein, J., Water Resources Res. 26, 2959-2970 (1990).