60-year evolution of retrogressive thaw slumps on the central Qinghai-Tibet Plateau

Permafrost underlying the central Qinghai-Tibetan Plateau has experienced significant degradation in recent decades due to the warming climate. Retrogressive thaw slumps (RTSs), a typical form of abrupt permafrost disturbance, retreat rapidly and can impact the local environment for decades before stabilizing. Previous studies have revealed rapid growth of RTSs in the central plateau, with their areas and numbers increasing by 154% and 70% from 2016 to2022, respectively. To gain a deeper understanding of the long-term activity and distribution of RTSs, we utilized multi-source satellite imagery, including Keyhole (1965–1984), Landsat (1984–2015), WorldView (2006–2013), and PlanetScope (2016–2024), to trace their activities over six decades. We manually delineated RTSs on high-resolution (<5 m) satellite images in 1965–1984 and 2016–2024. To fill the temporal gap between 1984 and 2016, we acquired 30-m-resolution Landsat imagery and applied deep learning-based heatmap regression to estimate RTS areas.

We identified 126 RTSs affecting 108 ha areas in 1965; while after sixty years, the number and affected areas increased by 4.48 and 13.9 times. Notably, around 50 new RTSs developed between 1970 and 1973. Since then, the activity of RTSs has slowed, with numbers increasing slightly from 274 to 287 in 1973-2010. However, from 2010 to 2013, the number rose to 324, affecting 482 ha. Between 2016 and 2017, RTSs surged from 407 to 697, impacting 1,167 ha. By comparing the average air temperatures from station records for the thawing season (from June to August), we found that the episodically rapid growth of RTSs during 1970–1973, 2010–2013, and 2016–2017 was associated with high summer temperatures.

In conclusion, we compiled a comprehensive timeseries of RTS evolutions in the central Qinghai-Tibetan Plateau, based on which we found that the active initiation of RTSs may be attributed to high summer temperatures. However, a more detailed analysis incorporating other climatic and environmental factors is necessary. Based on the long-term evolution of RTSs, we will further enhance our analysis of their changing numbers and areas to gain a quantitative understanding of the decadal evolution of these hillslope thermokarst landforms on the central Qinghai-Tibetan Plateau.