High Interannual Variability of Thaw Pond Area in the Indigirka River Lowlands, a Siberian Arctic Tundra Landscape

Degradation of permafrost in the rapidly warming Arctic has the potential to release large amounts of previously locked-up carbon to the atmosphere, providing a positive feedback to global warming. Thaw ponds are characteristic features of vast areas with ice-rich permafrost in the Arctic. They influence methane fluxes, as well as albedo at the landscape-level, and act as steppingstones in the long-term cycles of larger thaw lake formation and degradation. Monitoring and understanding the dynamics in pond area and location is therefore important for predicting climate feedbacks and habitat changes in tundra landscapes, such as those found in the remote Indigirka lowlands in Siberia. Yet, previous attempts to quantify thaw pond dynamics in this region have been limited by the low temporal frequency of available remote sensing data and their grain size, given the small extent of thaw ponds that range from a few m2 to 1 ha. Using a time-series of high-resolution drone images between 2014 and 2021, we report high inter-annual variability of pond area with little to no trend at the landscape-level. However, not all ponds remained stable. We document progressive movement of several pond perimeters. Degradation of ice-wedges at the edge of some thaw ponds caused inundation and subsequent loss of dry dwarf-shrub dominated communities, while wet sedge and sphagnum-dominated communities invaded the open water surface on the other end of several ponds. The observed succession resulted in a stabilisation of the open water extent, but will affect ecosystem functions related to these vegetation communities, such as methane emissions and bird habitats. Our findings highlight the value of multitemporal drone imagery for monitoring dynamic and changing landscapes and illustrate that our understanding of thaw-pond dynamics within decadal timeframes is perhaps more limited than previously thought. Interannual thaw pond area dynamics and vegetation succession might create internal climate variability that is currently not represented in Earth system models and could result in impacts on Arctic biodiversity that are difficult to detect from satellites.