Frost cracking dictated landslide distribution in response to temperature change since Last Glacial Maximum across the Eastern Qilian Mountains

Abstract

Temperature-dependent frost cracking is a major weathering process that dominates sediment supply and hillslope erosion by reducing shear strength and conditioning slopes to be more failure-prone in cold mountain regions. Landslides are a common erosional process in such landscapes. The linkage between these two processes has been proposed but many of the mechanisms remain to be investigated. In this study we compile a landslide inventory in the Eastern Qilian Mountains including 2482 rockfalls and 5373 elongate transport zone footprints. We use a widely accepted frost cracking model to explore this weathering process as a mechanism of paleoclimate control on the modern landslide distribution. The results show that landslides concentrate in a narrow altitudinal range and that this elevation dependence varies with slope aspect. Based on the paleoclimate record, we calculate the time-averaged frost cracking intensity (Fci) since Last Glacial Maximum (LGM) and adopt different parameters for two slope aspects. We find that time averaged Fci is a good predictor of the landslide elevation distribution for both aspects. We then introduce two slope-based landslide initiation and runout models (i.e., SHALRUN and FCIRUN) to

Elevation‐Dependent Periglacial and Paraglacial Processes Modulate Tectonically‐Controlled Erosion of the Western Southern Alps, New Zealand

“…Periglacial processes (predominantly frost‐cracking) occur commonly in regions experiencing paraglacial landscape adjustment, and they can help pre‐condition bedrock for paraglacial landsliding (Draebing et al., 2022; Geng et al., 2022; Grämiger et al., 2017; Hartmeyer et al., 2020a). Compared to paraglacial landsliding, which represents lower‐frequency, higher‐magnitude erosional events that deliver a wide range of grain sizes, including meter‐scale boulders (Ben‐Yehoshua et al., 2022), periglacial frost‐cracking represents higher‐frequency, lower‐magnitude erosional events that deliver narrower and finer grain size distributions (P. A. Allen et al., 2015; Sklar et al., 2017).…”

“…Direct rock temperature measurements are only available in a few monitored sites (Draebing & Mayer, 2021; Draebing et al., 2022; Hartmeyer et al., 2020a), and variables such as snow/regolith cover, daily temperature cycles, bedrock strength, and fracture networks are challenging or impossible to constrain at the landscape scale and over >10 2 yr timescales. Consequently, studies comparing long‐term (>10 2 yr) erosion rates to frost‐cracking predictions generally use models that consider only MAAT and annual temperature fluctuations (Delunel et al., 2010; Geng et al., 2022; Hales & Roering, 2005, 2009; Marshall et al., 2015, 2021; Savi et al., 2015).…”