Permafrost degradation increases risk and large future costs of infrastructure on the Third Pole

The warming and thawing of near-surface permafrost poses a significant threat to infrastructure in cold regions. This degradation reduces substrate strength, increases mass movements, and causes thermokarst activity, shortening infrastructure lifespan and increasing maintenance costs. In the Northern Hemisphere, a substantial portion of infrastructure is at risk from permafrost degradation. Previous research has highlighted the substantial economic impacts in regions like the circumpolar Arctic, estimating billions of dollars in additional costs to maintain service functions. The Qinghai-Tibet Plateau (QTP), also known as the Third Pole, is a climate hotspot with extensive permafrost, crucial for the socioeconomic development of over 10 million people and vital for regional and national development. The QTP is experiencing warming rates double the global average, making its permafrost particularly vulnerable. While adaptation measures like roadbed cooling and foundation modifications have been implemented, comprehensive economic assessments considering future permafrost degradation and adaptation strategies remain limited for the QTP. This study aims to quantify the future risk of current infrastructure on the QTP due to permafrost degradation and to evaluate the associated additional costs, considering both adaptation measures and varying climate change scenarios.

Existing research has demonstrated the severe impacts of permafrost degradation on Arctic infrastructure, including increased maintenance costs and reduced lifespan. Studies in Alaska and Russia have quantified these economic damages, reaching billions of dollars in projected costs. However, similar comprehensive assessments for the QTP, a large permafrost region in middle and low latitudes, are lacking. Previous work on the QTP primarily focuses on biophysical changes, neglecting the economic implications of permafrost degradation. This study aims to fill this gap by using a data-driven approach to assess the economic costs and benefits of adaptation strategies in response to permafrost degradation.

This study employed a three-stage approach: (1) projecting future permafrost thermal state, (2) quantifying permafrost degradation hazard, and (3) assessing economic costs. First, a well-trained ensemble statistical/machine learning model (GLM, GAM, SVR, RF, GWR) predicted mean annual ground temperature (MAGT) and active layer thickness (ALT) at 1km resolution for 2008 (reference), 2050, and 2090 under four Shared Socioeconomic Pathways (SSPs: 126, 245, 370, 585) and two Paris Agreement warming targets (1.5°C and 2°C). WorldClim data served as climate predictors, supplemented by soil and terrain data. Second, a composite hazard index was developed by combining five individual indices (thermal index, settlement index, bearing capacity index, risk zone index, and expert-based index). A majority vote procedure was used to classify areas into high, medium, and low hazard zones. Third, an equivalent lifespan replacement model, calibrated using data from the Qinghai-Tibet Highway, quantified the additional costs of maintaining infrastructure under permafrost degradation scenarios with and without adaptation. The model considered infrastructure type (roads, railways, powerlines, buildings), replacement costs, useful lifespan, and adaptation costs. A 2.85% discount rate was used, and uncertainty was assessed using the 97.5th and 2.5th percentiles of the ensemble MAGT and ALT predictions.

The study projected substantial changes in the QTP's permafrost thermal state. By 2090 under SSP245, approximately 63.3% of the permafrost area will be in high-hazard zones, exposing around 60% of current infrastructure. This would necessitate an additional ~$6.31 billion to maintain service functions without adaptation. However, strategic adaptations can save ~$1.32 billion (20.9%). The warming rate significantly impacted the proportion of infrastructure in high-hazard zones. By 2050, under SSP245, 38% of roads, 39% of railways, 39% of power lines, and 21% of buildings were predicted to be at high risk. Limiting warming to 1.5°C could reduce high-risk infrastructure by approximately half compared to the 2°C target. The additional costs to maintain infrastructure were projected to be ~$3.98 billion (2008-2050) under SSP245, with transportation infrastructure accounting for the majority. Adaptation measures could save approximately 15.36% of costs by 2050, increasing to 21% by 2090. Regional variations in additional costs were substantial, with central-eastern regions facing the highest costs and northern/south-eastern regions the lowest. Reducing global warming to below 1.5°C could save ~$1.32 billion by 2090 compared to the 2°C target.

The findings highlight the significant economic risks associated with permafrost degradation on the QTP. The study's projections underscore the vulnerability of the region's infrastructure to climate change and the importance of both mitigation and adaptation. The substantial savings achievable through strategic adaptations demonstrate the economic viability of proactive measures. The regional variations in projected costs emphasize the need for targeted interventions, accounting for local conditions. The comparison with economic damage estimates from Alaska and Russia reveals that while the amount of infrastructure on the QTP is less, the economic damage is comparatively large due to the higher degradation rate of permafrost in the QTP. The results provide crucial data for informed decision-making regarding infrastructure investment and adaptation strategies in the QTP and other permafrost regions.

This study provides a quantitative assessment of the future economic costs of permafrost degradation on QTP infrastructure, emphasizing the critical need for both climate change mitigation and adaptation strategies. Proactive adaptation measures are economically advantageous, and limiting global warming is crucial in minimizing economic damage. Further research should focus on refining the damage relationships and integrating more detailed engineering-economic data to reduce uncertainty in cost estimates. The findings are essential for sustainable regional planning and resilient infrastructure management in the QTP and other permafrost regions.

The study's uncertainty stems from several sources: the limitations of large-scale permafrost models, incomplete infrastructure databases, simplified damage relationships, and the inherent limitations of the equivalent lifespan replacement model. Local-scale processes influencing permafrost degradation (e.g., talik formation) were not fully resolved, potentially leading to underestimation. Data scarcity regarding infrastructure replacement costs and adaptation measures introduced further uncertainty. The study assumed constant useful lifespans and adaptation costs, which may vary over time. Future research should address these limitations using more detailed datasets and advanced modeling techniques.