Thermokarst lakes are abundant landforms of permafrost regions, particularly in lowland areas with high ground ice content (Grosse et al., 2013). Initiating as depressions by some surface disturbance formed in thawing ice rich permafrost (van Everdingen, 1998), thermokarst depressions fill with water causing ponds to gradually expand and coalesce to often form large lakes (Britton, 1957; Carson, 1968). The abundance and dynamic nature of thermokarst lakes makes these landforms play a dominant role in landscape processes, such as surface energy balance, permafrost dynamics, hydrologic storage and export, carbon cycling, and provision of habitat for fish and water birds (Rovansek, 1996; Bowling et al., 2003, Arp et al., 2012b, Boyd, 1959; Dmitriev and Tolstikhin, 1971; Heinke and Deans, 1973; Martin et al, 2007; Alessa et al., 2008, Sibley et al., 2008; Jones et al., 2009, Alerstam et al, 2001, Vincent and Hobbie, 2000; Berkes and Jolly, 2001).
Lakes are spatially distributed across much of the northern high latitudes, with the highest concentration of lakes on Earth occurring between 50° - 70° N (Lehner and Döll, 2004), largely due to the presence of permafrost or to a history of glaciation. Thermokarst lakes, specifically, then are found in any lowland permafrost region with high ground ice content and a thick sediment package, a terrain susceptible to their formation. This includes broad swaths of land across Siberian Russia, Northwest Canada, and Interior and Northern Alaska (Grosse et al., 2013). As a region with high ground ice contents that was not glaciated during the last glacial maximum, thermokarst lakes dominate the Arctic Coastal Plain of Alaska, covering greater than 20% of land area (Sellmann et al., 1975; Hinkel et al., 2005, Arp & Jones, 2009). Total land area change in other continuous permafrost regions has been demonstrated to be rapidly increasing in Siberia (Smith et al, 2005) to slightly increasing on the eastern Alaskan ACP (Riordan et al., 2006), while results from the Tuktoyatuk Peninsula of NW Canada (Plug et al, 2008) and other studies on the Alaskan ACP (Jones et al., 2009) show that these lake area change effects may be masked due to hydromorphic effects of lake level change. Further still, discontinuous permafrost regions show shrinking lake area due to effects of subsurface lake drainage (Smith et al., 2005) and increased evapotranspiration vs precipitation regimes (Riordan et al., 2006).
Thermokarst lakes are highly dynamic landscape features, continuously evolving and interacting with the surrounding landscape. Prior studies of thermokarst lake expansion typically focus on the thermokarst or thaw lake cycle, the process by which these lakes are formed through permafrost degradation, expand through shoreline erosion, and can eventually drain by a variety of mechanisms (Cabot, 1947; Hopkins, 1949; Britton, 1957; Carson & Hussey, 1962; Carson, 1968; Everett, 1980; Billings and Peterson, 1980; Mackay, 1988; Brewer et al., 1993; Hinkel et al., 2007; Jorgenson and Shur, 2007;Marsh et al, 2009; Jones and Arp, 2015). In this context, lake shoreline expansion is viewed as a process and part of this thaw lake cycle, and can include physical thermomechanical erosion as well as lateral expansion through coalescing lake basins. Small ponds that form through this process continue to expand and coalesce, eventually forming larger lakes with progressively deepening basins (Britton, 1957; Carson, 1968) and thawed sub-lake permafrost or a talik if/when the mean annual lake bottom temperature exceeds 0°C (Brewer, 1958; Lachenbruch, 1962; West and Plug, 2008; Arp et al., 2016). The growth of Lake shoreline expansion has also been studied through examination of the salient orientation of lakes on the Alaskan ACP (Black and Barksdale, 1949; Livingstone, 1954; Carson and Hussey, 1962) and on the Tuktoyatuk Peninsula of Canada (Côté and Burn, 2002). These studies focus on the physical, wind driven circulation pattern leading to oriented expansion of thermokarst lakes. In all concepts of thermokarst lake evolution, shoreline erosion and lateral expansion are seen as key processes determining local and regional variation in the thaw lake cycle and how thermokarst lakes will respond to changes in climate.
The processes by which thermokarst lakes expand vary depending upon the dominant climate driven forces and environmental controls at play. These forces express themselves across various spatial and temporal scales, and it is this network of interactions that contribute to the heterogeneity of shoreline expansion in thermokarst lakes (Grosse et al., 2013). At the regional scale, the local geology and climate conditions likely serve as the strongest controls on shoreline expansion rates. Previously reported erosion rates in thermokarst lakes from regions in the circumpolar north show variability between regions as a whole (Jones et al., 2011), from a low of 0.10 m/yr on average for the Fish Creek area on the ACP (Jorgenson and Shur, 2007) and locations within Interior Alaska (Jorgenson and Osterkamp, 2005) to a high average rate of 0.7 m/yr in areas such as the Barrow Peninsula (Lewellen, 1970), the Northern Teshekpuk Lake Special Area (Arp et al., 2011) and Mayo Territory, Canada (Burn and Smith, 1990). This suggests that shoreline material composition and condition are strongly controlling factors in the expansion of thermokarst lakes (Jones et al., 2011).. Variation in expansion rates among lakes in a region has generally been attributed to differences in lake size, lake depth",, bluff height (Roy-Léveillée and Burn, 2010; Jones et al., 2011), lake-marginal permafrost temperature, and terrain units (Jorgenson and Shur, 2007) Because many of these factors can change within a single lake, between lakes within a region, and between regions simultaneously, understanding the role these factors can play at all scales is necessary for achieving a landscape scale perspective on the process of thermokarst lake expansion.
Most thermokarst lakes on the ACP are shallow due to their formation in permafrost with high ice content only near the surface, particularly on the outer coastal plain (Jorgenson and Shur, 2007). Because lake depths are often close to the maximum thickness of seasonal ice, many freeze solid with bedfast ice and other slightly deeper lakes retain liquid water below floating ice (Brewer, 1958; Lachenbruch et al., 1962, Mackay, 1992; West and Plug, 2008, Arp et al., 2011). The lake depth separating these ice regimes generally exists somewhere between 1.5 m and 2 m, where maximum winter ice growth is reached by the end of winter (Brewer, 1958; Sellmann et. al. 1975), though evidence suggest a trend toward thinner maximum ice thickness and a shift toward more floating ice conditions (Arp et al. 2012a; Surdu et al. 2014; Engram et al. 2018). The most critical effect created by floating ice conditions is the distinctly warmer thermal regime of a lake bed that leads to development of a thaw-bulb (talik) (Lachenbruch, 1962; West & Plug, 2008). The thermal impact generated by floating ice conditions may also occur laterally into surrounding lake shorelines and bluffs (Arp et al. 2011), as the lateral heat flux from lakes into these surrounding bluffs has been shown dissipate up to 70% of vertical heat flux into these water bodies (Langer et al. 2016). In lakes with bedfast ice regimes, warming of the permafrost can still occur (Langer et al., 2016) and may be becoming more rapid as trends in thinning lake ice increases floating lake ice conditions (Arp et al. 2016). The extent to which warmer sublake permafrost is also apparent along shorelines is uncertain, but evidence from other continuous permafrost regions suggest talik development beneath water as shallow as 0.2m depth along expanding shorelines (Roy-Léveillée and Burn, 2017), similar to lake centers, where perennially thawed ground will not exist due to the presence of liquid water year round and mean annual temperatures greater than 0°C (Brewer, 1958; Lachenbruch, 1962; West & Plug, 2008). Because permafrost temperatures are colder beneath and adjacent to bedfast ice lakes relative to floating ice lakes, they may be less susceptible to thermal erosion, which would be manifest in lower overall thermokarst lake expansion rates. This pattern was observed by Arp et al. (2011), with bedfast ice lakes within the YOCP showing lower expansion rates than floating ice lakes over the time period 1979-2002.
This study addresses the potential impact of regional geology, including ground ice content and soil characteristics, through the analysis of 40 lakes in three distinct regions on the ACP. The impacts of localized factors, including lake ice regime and terrain unit, on the processes that contribute to thermokarst lake expansion rates are addressed through comparisons of 20 lakes within an individual region and within an individual lake. We hypothesize that lakes will predominantly be controlled by their regional setting and ice conditions, with wintertime processes playing a larger role than previously considered. Through a multi-scale approach using remote sensing techniques in addition to field based data collection, this analysis aims to better inform those involved in land and water resource management on the ACP as to the dynamics of lake expansion processes.
