High-frequency internal waves in the littoral zone of a large lake

Observations in the littoral zone of a large lake (Lake Constance) revealed strong and periodic fluctuations of temperature and current velocity on timescales between 10 and 15 min associated with high-frequency internal waves. The peak in the spectral energy of the current velocity fluctuations associated with high-frequency internal waves follows the seasonal dynamics of stratification, described by the stability frequency N. A comparison between the nearshore current velocity measurements in the littoral zone with temperature measurements in the pelagic of Lake Constance revealed a strong coupling between the occurrence of high-frequency internal waves in the littoral zone and the passage of the basin-scale internal Kelvin wave with a period of about 4 d. The littoral zone of a lake, which is usually defined as the shallow area near the shore where light penetration is sufficient for the growth of plants, is an extremely dynamic habitat where the physical conditions are highly variable on diurnal and subdiurnal timescales. Surface waves and the daily meteorological cycle are the most obvious physical processes causing a high temporal variability in the current field and in the thermal regime, particularly in the shallow littoral zone with depths of about 1 m. Here, we present data from Lake Constance demonstrating that current velocities and temperature in the deeper littoral zone, down to a depth of about 20 m, also have a high temporal variability on timescales of minutes to hours, and that the amplitude of, e.g., temperature fluctuations can be significantly larger than those in the shallow littoral zone. Measurements—Lake Constance is one of the largest lakes in Central Europe (Fig. 1a). The lake consists of two basins connected by the shallow Lake Rhine: the deep main basin Upper Lake Constance (63 km long, 14 km wide) is located upstream of the shallower basin, called Lower Lake Constance. The central part of Upper Lake Constance has a maximum depth of 252 m and a mean depth of 101 m. Its subbasin Lake Überlingen (maximum and mean depths are 147 m and 84 m, respectively) is separated from the central part of Upper Lake Constance by the Mainau Sill reaching up to 100 m water depth (Wessels 1998). For general information on the hydrodynamics of Upper Lake Constance, see Bäuerle et al. (1998). This study concentrates on data from two different moorings deployed in Upper Lake Constance: Mooring M1, located at 147 m depth in the deepest part of Lake Überlingen (Fig. 1a), measures meteorological parameters at the surface and water temperature at 34 depth levels with a fast-response thermistor chain (Precision Measurement Engineering) at a sampling interval of 60 s. The mooring has been operated continuously since December 2003. Mooring M2 was deployed nearshore (about 75 m offshore) from June to December 2004 at a depth of 19 m. Its location was near the intersection between Lake Überlingen and the central basin of Upper Lake Constance (Fig. 1a). M2 consisted of a 500-kHz acoustic Doppler current profiler (NDP, Nortek AS) in combination with five individual temperature loggers (TR-1050, RBR Ltd.) at depths of 19, 18, 15, 11, and 7 m. The sampling interval of the thermistors was set to 10 s (the nominal response time of the TR-1050 thermistors is less than 3 s). Two additional TR-1050 temperature loggers were deployed further upslope from mooring M2 at 7 and 3 m depth (at a local depth of 7 m) from 08 July to 20 August 2004 (Fig. 1b). The upward-looking NDP at mooring M2 was deployed in a gimbal mount at a depth of 18.5 m (Fig. 1b). It was operated in pulse-coherent mode and was connected to an onshore computer and power supply. The sampling interval was 1 s, and 10 individual profiles were averaged internally before the data were recorded on the computer, resulting in a profile interval of 10 s. The vertical bin size was 13 cm, and data were recorded for 64 bins, resulting in a measured depth range from 1.33 to 9.52 m above the sediment (17.7to 9.5-m depth). The NDP was mounted vertically with typical pitch and roll angles of less than 63u. The vertical velocity refers to a Cartesian (earth) coordinate system. The bottom slope at the deployment site, however, was very steep (approximately 35u) so that the vertical coordinate is locally not perpendicular to the sediment surface. The horizontal velocities were rotated in a horizontal plane to represent cross-slope and along-slope current velocities. The NDP data were contaminated by singular spikes, the frequency of occurrence of which increased with increasing distance from the instrument. These spikes were removed, and the remaining data were averaged into 1-min samples for the subsequent analysis in order to avoid data gaps or interpolations. Continuous time series of the 1-min samples, however, could only be obtained from the depth cell closest to the instrument (at a depth of 17.7). Hence, the following time series analysis will be restricted to data from