Under-ice thermal stratification dynamics of a large, deep lake revealed by high-frequency data

We measured under-ice thermal stratification from before ice-on through after ice-off in Lake Sunapee, New Hampshire, a large, deep, north temperate lake, using a high-frequency monitoring buoy in the winter season of 2007–2008 to quantify how lake thermal stratification varies throughout the under-ice season. We examined potential drivers of variation in under-ice stability, identified diel-scale patterns in under-ice stratification, and used this dataset to test the hypothesis that there are two distinct under-ice phases driven by heat flux from the sediment followed by increased solar radiation as winter progresses. High-frequency measurements demonstrated that only a small fraction of the under-ice period exhibited the traditional inverse stratification previously thought to prevail, based on temporally discrete under-ice temperature profiles. Local short-term weather conditions altered under-ice conditions throughout the ice season with brief periods of snow melt, resulting in several days of disrupted thermal stratification. Our data indicate that thermal structure under the ice in Lake Sunapee is dynamic, and in contrast to smaller, shallower lakes, may be categorized in three, not two, distinct phases. As the under-ice season continues to become shorter due to climate change, under-ice thermal stratification in lakes will likely decrease further. Temperate lakes around the world are responding to the changing climate (Adrian et al. 2009). Many temperate, seasonally frozen lakes are experiencing later ice-on and earlier ice-off dates than in the past, leaving fewer days each year with ice cover (Magnuson et al. 2000; Hodgkins et al. 2002). This trend is expected to continue (Shuter et al. 2013). The timing and duration of ice cover plays a critical role in lake water column mixing, delivery of nutrients to the epilimnion, and the subsequent characteristics of the plankton communities each spring (Adrian et al. 1999; Weyhenmeyer et al. 1999; Winder et al. 2009), as well as the occurrence of summer hypoxia (Wilhelm et al. 2014). Earlier onset of the open water season each spring can threaten water quality in many lakes because summer stratification develops over a longer period of time, thereby increasing the duration of summer hypolimnetic hypoxia (Jeppesen et al. 2010). Ice-covered lakes were once thought to be “dormant” (Kirillin et al. 2012); however, that view is beginning to change (Kirillin et al. 2012; Bertilsson et al. 2013). The under-ice period has traditionally been characterized by a pattern of inverse thermal stratification, represented with the coldest water immediately beneath the ice and a warmer, stable bottom layer of more dense water (Wetzel 2001). Convective processes result in very little mixing of stratified layers (Kirillin et al. 2012). Under-ice circulation and mixing patterns are hypothesized to occur in two distinct phases: Winter 1, with hydrodynamics driven by heat flux from the sediment, and Winter 2, with mixing driven by increased heat from solar radiation under the ice (Kirillin et al. 2012; Bertilsson et al. 2013). Sediment heat flux during Winter 1 is important in small, shallow lakes (less than two meters; Kletetschka et al. 2013), but thought to be diminished in larger, deeper lakes (Farmer 1975; Kirillin et al. 2012). While previous studies have documented how weather conditions influence ice cover (Beier et al. 2012), less work has been conducted to determine how weather influences under-ice lake thermal stratification, especially the onset and length of Winter 1 and Winter 2 phases. Most of our knowledge of under-ice dynamics in large, temperate, and seasonally ice-covered lakes to date has been derived from the measurement of temporally discrete (monthly or weekly) thermal profiles (Wetzel 2001; but see Farmer 1975) or short-term collection of higher temporal resolution data (Mironov et al. 2002). In contrast to large *Correspondence: [email protected] 347 LIMNOLOGY and OCEANOGRAPHY Limnol. Oceanogr. 60, 2015, 347–359 VC 2014 Association for the Sciences of Limnology and Oceanography doi: 10.1002/lno.10014 temperate lakes, more under-ice data are available for Arctic lakes (Arp et al. 2010) and smaller seasonally ice-covered lakes (Forrest et al. 2008; Kletetschka et al. 2013). Temporally discrete measurements do not provide sufficient data to detect the occurrence of diel-scale changes under-ice. High-frequency data—that is, measurements made on the scale of minutes collected during ice formation, ice cover, and ice melt are needed to understand the local drivers (e.g., weather patterns) and small-scale temporal variation in winter lake stratification. However, due to the difficulties of sampling and equipment damage by ice, it remains a challenge to collect high-frequency data, especially during the transitions of ice-on and ice-off, as well as during ice periods (but see Arp et al. 2010; Pierson et al. 2011; Kletetschka et al. 2013). Our primary objective was to quantify variability in thermal stratification under the ice in a large, deep, temperate lake. We used a high-frequency under-ice dataset of vertical temperature profiles from Lake Sunapee, New Hampshire, which is representative of large, deep north temperate lakes. Lake Sunapee has a high-frequency monitoring buoy (Klug et al. 2012) that was successfully deployed through the winter season of 2007–2008. This rare winter dataset successfully captures the entirety of the under-ice period, from before ice-on through after ice-off. Characterizing under-ice patterns of stratification is a first step toward understanding how a changing ice season may cascade to influence lake ecosystem dynamics, especially biogeochemical cycling (Moss 2012), microbial ecology (Twiss et al. 2012; Bertilsson et al. 2013; Wilhelm et al. 2014), and fish community structure (Shuter et al. 2012). Specifically, we focused on the following three questions: (1) how do vertical temperature profiles and lake stability vary under the ice, and what are the drivers of this variability?; (2) does the temperature profile exhibit diel fluctuations under the ice, as it does during the ice free season, and if so, how deep into the water column do these diel fluctuations occur?; and finally, (3) can patterns of lake thermal stratification and stability reveal distinct phases under the ice progressing from ice-on to ice-off (Winter 1 and Winter 2; sensu Kirillin et al. 2012), or in response to local weather?