Glacial history modifies permafrost controls over the distribution of lakes and ponds
Published:
Figure 1: Maps of ALPOD lakes. Panels (a-d) show ALPOD lake statistics calculated for HUC12 watersheds from the National Hydrography Dataset: (a) lake fraction (percentage of watershed covered by open water at least 25% of the study period), (b) average lake size within each watershed, (c) lake density calculated as the number of lakes per 100 km2, and (d) total shoreline distance (i.e., lake perimeter). Panel (e) shows the open water occurrence image for individual lakes ranging from 0.005 km2 to 66.24 km2, which shows relationship between open water occurrence and lake polygons, and (f) displays a detailed example of the vector products derived from the occurrence image within the Little Black River floodplain near the Yukon Flats.
Figure 2: Landscape controls on lake distribution. (a) Permafrost extent across Alaska and glacial coverage during the Last Glacial Maximum. (b) A comparison of lake fractions in watersheds binned by glacial history and substrate texture. The remaining landscapes (unglaciated, coarse substrate, and postglacial fine substrate watersheds) are displayed in Figure S.5. Lake fraction increases with permafrost extent in unglaciated watersheds with fine geologic substrate (p < 0.0001), whereas lake fraction decreases with permafrost extent in postglacial watersheds with coarse substrate (p < 0.0001). Panel (c) compares lake density (number of lakes per 100 km2) in watersheds binned by substrate texture and permafrost extent. Lake densities are higher in fine substrate and increase with permafrost class (p-value < 0.0001). Watersheds underlain by coarse substrate exhibit lower overall lake density and a negative correlation between permafrost extent and lake density (p-value < 0.0001). (d) Size distribution of ALPOD lakes visualized through a plot of abundance vs. lake size.
Figure 3: Schematic comparing permafrost controls on lake fraction across physiographic settings. Each panel displays the oblique view of topography and ALPOD lakes from a watershed in our study that is representative of our distinct physiographic (unglaciated vs. glaciated) and permafrost (continuous vs. discontinuous) classes. The inset panel shows the number of study sites from Webb and Liljedahl (2023)33 that exhibited decreasing, increasing, or insignificant decadal-scale lake area trends within the corresponding physiographic settings. The subsurface distribution of permafrost and seasonally freezing active layer are generated to illustrate permafrost hydrology mechanisms.
Figure 4: Estimated lake area loss based on our space-for-time substitution and projected permafrost thaw. (A) displays uniform predictions extrapolating negative correlations between lake area and permafrost extent in unglaciated-fine substrate watersheds to the entire study site (b) shows predictions constrained based on physiographic setting. Note that although the North Slope’s physiographic characteristics indicate that its lake area is sensitive to permafrost thaw, the permafrost extent is not expected to significantly decrease within this timeline56.
