Thawing permafrost poses environmental threat to thousands of sites with legacy industrial contamination

The Arctic permafrost region is experiencing rapid warming, at rates at least twice as fast as the global average, with some areas showing up to four times faster warming. This accelerated warming significantly alters ground stability and hydrological conditions, raising concerns about the mobilization of hazardous substances from various sources. Past industrial practices in the Arctic often assumed that permafrost would serve as a permanent and stable platform, and a long-term containment for industrial waste. This led to the accumulation of various toxic substances, including drilling and mining wastes, drilling muds and fluids, mine waste heaps, heavy metals, spilled fuels, and radioactive waste, often disposed of using methods that relied on the permanence of the permafrost as a hydrological barrier. These practices included creating covered waste dumps in permafrost, using hydrologically closed lakes and basins as natural dumps, and spreading substances for dilution. The 2020 Norilsk diesel fuel spill, attributed in part to loss of soil stability due to permafrost thaw, serves as a stark example of the potential environmental disasters that can result from this legacy contamination. This study addresses the urgent need for a pan-Arctic assessment of the environmental impact of these activities, focusing on the escalating risk of toxic release and mobilization due to permafrost thaw. The study aims to quantify the number of industrial sites and associated contaminated sites in permafrost regions and project how many will be affected by future permafrost thaw under different climate scenarios. Understanding this risk is crucial for developing effective long-term planning and management strategies.

Existing literature highlights several key aspects relevant to this study. Firstly, the rapid rate of permafrost thaw and its impact on infrastructure stability has been widely documented. Studies demonstrate that degrading permafrost is putting Arctic infrastructure at risk, leading to significant economic costs for maintenance, replacement, and relocation. Secondly, the potential biogeochemical risks associated with permafrost thaw, including the mobilization of hazardous substances from various sources, have been increasingly recognized. However, a comprehensive pan-Arctic assessment of industrial legacy sites and their associated contamination risks within the context of permafrost thaw has been lacking. While some regional studies exist, for example, the Contaminated Sites Program (CSP) in Alaska and the Federal Contaminated Sites Inventory (FCSI) in Canada, there's a significant data gap, particularly concerning the vast Russian Arctic. Previous research has focused on individual case studies or specific contaminant types, but a holistic pan-Arctic perspective integrating industrial sites and contaminated sites with permafrost thaw projections was necessary to fully grasp the magnitude and complexity of the problem.

This study utilized a multi-faceted methodology combining geospatial data analysis, statistical modeling, and permafrost simulations. First, a geospatial dataset of industrial sites was synthesized using data from OpenStreetMap (OSM) and the Atlas of Population, Society and Economy in the Arctic (APSEA). The Northern Hemisphere Permafrost Map (NHPM) was used to delineate the permafrost model domain, considering only areas with a permafrost occurrence probability exceeding 50%. To understand the relationship between industrial sites and contamination, regional data from Alaska (CSP) and Canada (FCSI) were synthesized. This allowed for the quantification of contamination extent and nature, and its spatial relationship to industrial sites. The spatial relationship observed in North America was then extrapolated to the Arctic as a whole, validated using a compilation of Russian sites from public sources. Two point process models were fitted to the CSP/FCSI data to provide upper and lower bounds for the estimated number of contaminated sites across the Arctic permafrost region. A numerical permafrost model (CryoGridLite) driven by past and future climate projections (using RCP 2.6 and RCP 8.5 scenarios from CCSM4 and HADGEM2-ES CMIP5 projections) was used to predict the extent of permafrost thaw and its impact on industrial and contaminated sites. The model considered the formation of persistent taliks (unfrozen soil layers) as an indicator of permafrost degradation. The model’s spatial resolution was 1 degree, and thawing within grid cells was considered likely if simulations indicated persistent talik formation. The study also analyzed the types of toxic substances associated with industrial activities, primarily using Alaskan CSP data. The frequency of different substances and their aquatic toxicity (LC50-fish) were assessed to characterize the nature of the contamination. Data gaps were addressed through manual cross-referencing and literature review. Finally, the spatial analyses utilized Python libraries (scipy, pandas, numpy, matplotlib, basemap) and R libraries (sf, maptools, raster, rgeos, geosphere, rgdal, spatstat, ncdf4).

The study revealed a significant number of industrial sites and associated contaminated sites in Arctic permafrost regions. Approximately 4500 industrial sites were identified within the permafrost model domain, with an estimated 13,000 to 20,000 contaminated sites linked to these industrial activities. The distribution of these sites is strongly skewed towards Russia, which hosts the largest percentage of both industrial sites and estimated contaminated sites. Analysis of Alaskan data indicated that fuels (diesel, kerosene, gasoline, and associated chemicals) constitute the largest fraction of substances found at contaminated sites, followed by substances with high aquatic toxicity (mercury, lead, arsenic). The permafrost thaw projections suggest that a substantial portion of these sites will be affected by permafrost degradation in the coming decades. Under current conditions, about 22% of industrial sites and 20% of contaminated sites are located in areas where permafrost degradation is possible. By 2100, under a low-emission scenario (RCP 2.6), the number of affected industrial sites could increase by 1100, and contaminated sites by 3400–5200. Under a high-emission scenario (RCP 8.5), almost all industrial and contaminated sites would be located in regions affected by permafrost degradation. The study also highlights the large uncertainties associated with the estimates, particularly for Russia, due to limited data availability. Furthermore, the analysis reveals that permafrost degradation can significantly increase the cost of mitigation and adaptation measures, including infrastructure maintenance, replacement, remediation of contaminated sites, and changes to hydrological pathways.

The findings of this study emphasize the significant environmental risk posed by legacy industrial contamination in Arctic permafrost regions. The accelerating rate of permafrost thaw significantly increases the probability of toxic substance release due to infrastructure failure and the opening of new hydrological pathways. The study’s projections underscore the need for extended time horizons in risk assessments, considering the long-term consequences of permafrost degradation that extend beyond 2050. Staying below the 2°C global warming target is crucial to mitigate the rapid increase in affected sites in the second half of the 21st century. The study's estimates are likely conservative as they don't account for rapid thaw processes, infrastructure effects on permafrost, or the influence of contamination on permafrost stability. The results highlight the urgent need for improved data collection and transparent documentation of contamination status in the Arctic. There is a substantial lack of public information on industrial sites and activities, especially in Russia, hindering a more comprehensive risk assessment. Existing databases such as CSP and FCSI provide valuable information but are limited in spatial coverage. The financial implications of managing this risk, including securing contaminants, post-operational renaturation, and remediation, are likely to be substantial and should be incorporated into cost-benefit calculations for future industrial activities in the Arctic.

This study provides the first pan-Arctic assessment of the environmental risk posed by legacy industrial contamination in thawing permafrost. Thousands of industrial sites and associated contaminated sites are identified, with a substantial portion projected to be affected by permafrost thaw by the end of the century. The results highlight the urgent need for enhanced data collection, transparent documentation, and the development of robust long-term management strategies. Future research should focus on improving data availability, particularly for the Russian Arctic, and refining models to account for rapid thaw processes and the complex interactions between infrastructure, permafrost, and contaminants. The findings underscore the need for international cooperation and the implementation of effective environmental regulations to mitigate the growing environmental risks associated with industrial legacy in the Arctic.

This study acknowledges several limitations. The extrapolation of contaminated sites from North American data to the pan-Arctic scale introduces uncertainties. Data availability for Russia, in particular, is limited, resulting in large uncertainties in the estimated number of contaminated sites. The permafrost model used has a coarse spatial resolution, neglecting rapid thaw processes and site-specific variations in ground conditions. The study doesn't consider all forms of industrial infrastructure, such as extensive pipeline networks, potentially underestimating the total number of at-risk sites. Finally, the study focuses on industrial contamination, excluding other potential sources of pollution and their combined effects on the environment.