The northern Japanese Alps on the Sea of Japan side is located in the heavy snow region, with snow depths reaching 6 m on Murodo lava plateau above the timberline, where topographic effects are minimal (Iida et al., 2018). In the cirque landforms surrounded by steep rock walls, snow depths can reach up to 30 m due to pronounced topographic effects (Arie et al., 2022). In the alpine zone, located within a periglacial environment in a heavy snow region, strong northwest monsoon often generates variations in snow cover over a narrow elevation range, leading to heterogeneous temperature fluctuations and freeze-thaw processes.
Freeze-thaw processes play a significant role in triggering bedrock failures, such as collapses and rockfalls (Matsuoka, 2019). Therefore, it is crucial to elucidate the relationship between snow cover and freeze-thaw action in heavy snow region. Despite this importance, the relationship between snow cover and freeze-thaw action remains under-evaluated, and publicly available datasets for such analyses are scarce, especially when compared to the comprehensive datasets recently documented in the European Alps (e.g., Kellerer-Pirklbauer, 2017; Draebing and Mayer, 2021).
Two indicators can be used to evaluate freeze-thaw action. 1) Freezing Index (FI) quantifies the potential impact of freezing conditions over a specific period, calculated using the cumulative sum of sub-zero temperatures. 2) Time within the Frost Cracking Window (FCW) is the duration within temperature ranges, typically between −3°C and −8°C, where ice segregation processes are more active (Anderson, 1998; Hales and Roering, 2007). Prolonged exposure to such conditions facilitates the growth of ice lenses, contributing significantly to rock fracturing.
We investigated rock temperature and snow accumulation in the Hakuba Mountains (about 3,000 m a.s.l.) in the northern Japanese Alps over a three-year period from 2021 to 2024. Our analysis focused on examining changes in FI and FCW duration with variations in snow depth. Snow depth was calculated using SfM-MVS software and a multi-year 3D terrain model created from Cessna and UAV images.
Results indicated that FI was higher on windward slopes, which receive more significant solar radiation and experience reduced snow cover. In contrast, the freezing index was significantly lower on flat, leeward slopes characterized by prolonged thick snow cover during winter.
At sites with lower winter surface temperatures and higher FI, the time below the lower limit of FCW ( -8°C) tended to be longer, and the FCW recording time tended to be shorter. However, some sites recorded long FCW despite large FI. Such locations included cliffs at the top of flat slopes covered by thick snow in the latter half of the winter season or under cornice near mountain peaks, where temperatures were low in the first half of the winter season and did not rise easily in the latter half.
Furthermore, a comparison of FCW duration across different depths revealed that the depth at which the FCW peak was pronounced and the time of year when it occurred varied significantly between locations. These depth-specific results highlight the importance of monitoring seasonal changes in FCW.