This review article deals with bank erosion from the perspective of rivers affected by seasonal ice formation. These rivers drain half of the terrestrial land surface globally, and are mainly located in both periglacial and cold, non-periglacial environments across the Northern Hemisphere. This review is based on a literature survey of 126 publications (articles, technical reports, conference papers and book chapters) documenting case studies in temperate and polar climates. The first details the global issues of bank erosion and pinpoints concerns specific to northern environments. The second describes the dominant erosion processes (fluvial vs. terrestrial), mechanisms (mechanical vs. thermal) and typical landforms encountered in the literature. The third reviews the environmental factors (hydraulic vs. non-hydraulic) controlling bank erosion, with a focus on the different forms of river ice. The fourth deals with the spatial and temporal variability in bank-erosion processes, discussing the distribution of process dominance occurring at the reach scale and the catchment scale, and describing the temporal window in which each process dominates. The fifth reviews the expected impacts on bank erosion resulting from climate-induced disturbances on hydrological cycles and from increasing anthropogenic pressures along riverbanks in northern countries. The relationships among erosion processes, environmental factors, climate change, and human impacts are summarized in a sixth that introduces a new synthetic conceptual diagram of bank erosion. Research needs that should be investigated in the future are highlighted in the seventh while the final synthesizes all the aspects presented in this review.

To better understand the linkage between lake area change, permafrost conditions and intra-annual and inter-annual variability in climate, we explored the temporal and spatial patterns of lake area changes for a 422382-ha study area within Yukon Flats, Alaska using Landsat images of 17 dates between 1984 and 2009. Only closed basin lakes were used in this study. Among the 3529 lakes greater than 1 ha, closed basin lakes accounted for 65% by number and 50% by area. A multiple linear regression model was built to quantify the temporal change in total lake area with consideration of its intra-annual and inter-annual variability. The results showed that 80.7% of lake area variability was attributed to intra-annual and inter-annual variability in local water balance and mean temperature since snowmelt (interpreted as a proxy for seasonal thaw depth). Another 14.3% was associated with long-term change. Among 2280 lakes, 350 lakes shrank, and 103 lakes expanded. The lakes with similar change trends formed distinct clusters, so did the lakes with similar short term intra-annual and inter-annual variability. By analysing potential factors driving lake area changes including evaporation, precipitation, indicators for regional permafrost change, and flooding, we found that ice-jam flooding events were the most likely explanation for the observed temporal pattern. In addition to changes in the frequency of ice jam flooding events, the observed changes of individual lakes may be influenced by local variability in permafrost distributions and/or degradation. Copyright (c) 2012 John Wiley & Sons, Ltd.