Abstract
Small headwater catchments are highly dynamic systems, but we often lack data and understanding of the hydrological processes taking place there. This study focused on the Krycklan catchment (67.8 km2) in Northern Sweden and the dynamics of the shallow groundwater tables, stream emergence and soil moisture over time and in space. It is generally assumed that, in humid climates, the groundwater table is a subdued copy of the surface topography but there is currently no guidance on what resolution DEM to use in hydrological analyses. Nevertheless, detailed microtopography, as can be retrieved from high-resolution Digital Elevation Models (DEMs), is unlikely to affect groundwater topography. A first step, therefore, was to look at the effects of DEM- smoothing and -aggregation on the calculated flow directions and derived catchment boundaries. For more than 40 % of the Krycklan catchment area the calculated flow directions depend strongly on the degree of smoothing or aggregation of the DEM. These are areas with local slopes in the opposite direction of the general slope, flat areas, ridges, and incised streams. We calculated the drainage area for 40 locations, outlets of catchments of different sizes, and found that the processing of the DEMs affected small catchments (i.e., first-order streams) the most. This highlights the need to carefully consider the effects of DEM smoothing or -aggregation on the calculated flow directions and drainage areas as the shifts in catchment boundaries and drainage areas can have a significant effect on the calculated water balance. To compare the results from the theoretical DEM analyses with real observations, a network of groundwater wells was installed in two areas in the Krycklan catchment. One is a small headwater catchment (3.5 ha; 54 wells) and the other a hillslope (1 ha; 21 wells). The wells were 274 cm deep on average (range of 70–581 cm). The positions of the wells were determined using a Light Detection and Ranging (LiDAR) scanner. The groundwater-level variations were recorded between 18 July 2018 and 1 November 2020 using capacitance water level loggers. During the summers of 2018 and 2019, manual water-level measurements were done to validate and re-calibrate the automatic water-level measurements. The groundwater-level data were carefully processed to determine differences in absolute groundwater levels and to calculate groundwater gradients. Additionally, all wells with sufficient water were sampled once during the summer of 2019 and analyzed for electrical conductivity, pH, absorbance (at 254 nm, 365 nm, 420 nm, and 436 nm), and anion and cation concentrations, and for the stable isotopes of hydrogen and oxygen. Groundwater gradients were calculated from the groundwater levels for the period May 2019 to October 2020. As expected, the calculated gradients changed with catchment wetness. Gradient directions calculated over short distances (5 m) changed by up to 360° and gradients calculated over larger distances (20 m) varied by up to 270°. As expected from the DEM analysis, the variation in groundwater gradient directions was largest for flatter locations and locations where the local surface slope differed from the surrounding topography. Though smoothed DEMs represented the groundwater surface better than high-resolution DEMs, the optimal degree of smoothing varied over the year. It was lowest for very wet periods, such as the snowmelt period, when groundwater tables were high. To complement the groundwater data, campaigns to map the soil moisture and stream state were carried out during the summer seasons of 2018 and 2019. These campaigns were based on qualitative citizen science approaches and showed how soil moisture decreased and increased and how stream networks contracted and were reactivated in the small subcatchment throughout the two field seasons. These qualitative measurements highlight the need for detailed spatially distributed measurements to understand the spatial variation in water storage (soil moisture, groundwater, or surface water) across the catchment and the usefulness of qualitative observations.