Presentation Information
[ACC42-03]Present and Future Mountain Permafrost Distribution in Japan Derived from a Land Surface Model and Observations: Impacts of Soil Physical Properties
*Tokuta Yokohata1, Go Iwahana2, Kazuyuki Saito3, Tetsuo Sueyoshi4, Takashi Arakawa5, Tomoko Nitta6, Yukihiko Onuma9, Irina Melnikova1, Yasuto Watanabe1, Noriko N Ishizaki1, Haruto Okubo7, Yuka Sawa7, Shiona Nagane7, Kumiko TAKATA8 (1.National Institute for Environmental Studies, 2.University of Alaska Fairbanks, 3.Japan Agency for Marine and Earth Science and Technology, 4.National Institute for Polar Research, 5.CliMTech, 6.Chuo University, 7.Tsukuba University, 8.Azabu University, 9.Meteorological Research Institute)
Keywords:
Climate Change,Permafrost,Land Surface Model
Permafrost is defined as ground that remains below 0°C for more than two consecutive years and is widely distributed across the tundra and taiga regions of the high latitudes in the Northern Hemisphere. Japan lies at the southernmost boundary of the global permafrost distribution. Consequently, permafrost has only been confirmed at a very limited number of sites in high-latitude, high-elevation mountainous regions with extremely cold climatic conditions. Because permafrost formation is a subsurface process, it is inherently difficult to observe directly. This challenge is further amplified in mountainous areas, making it difficult to obtain a comprehensive understanding of permafrost distribution.
In this study, we used 1 km resolution temperature projections for Japan as input to the Integrated Land Simulator (ILS), a land surface physical model capable of estimating subsurface physical processes. ILS uses atmospheric variables—such as air temperature, precipitation, radiation, wind speed, and humidity—to simulate near-surface physical processes and calculate ground temperature, soil moisture, river discharge, and runoff etc. The 1 km climate forcing used here was developed using the latest climate projections and bias-correction techniques to support stakeholders, including local governments and private companies, in developing climate change adaptation strategies. We estimated the distribution of permafrost in Japan by applying surface physical conditions from the pre-industrial period to the present within ILS.
The ILS simulation results revealed that permafrost exists over a much narrower area compared with the distribution estimated from temperature projection data and empirical relationships used in our previous study (Yokohata et al., 2022). This discrepancy is likely due, at least in part, to the empirical relationships having been developed from environments such as Alaska and Siberia, which differ significantly from Japanese conditions. Meanwhile, previous observational studies have shown that permafrost in areas such as the Daisetsu Mountains is typically found in wind-exposed sites where strong winter winds inhibit snow accumulation. To account for this effect, we conducted numerical experiments by modifying model parameters to reduce snow depth. The results indicated that decreasing snow depth—representing wind-exposed conditions—led to a broader permafrost distribution. Using model parameter settings that successfully reproduced the observed permafrost distribution, we performed both historical climate reconstructions and future projection experiments to estimate the current permafrost distribution in Japan and its potential future changes.
Our estimates suggest that mountain permafrost is currently distributed extensively in the Hidaka Mountains. Therefore, since June 2024, we have conducted ground temperature observations near the summit of Mount Poroshiri in the Hidaka Mountains of Hokkaido. Permafrost is defined as ground that remains at or below 0°C for at least two consecutive years; however, our observations revealed no such persistently frozen layer within the upper 10 meters of subsurface near the summit of Mount Poroshiri. Although shallow layers fall below 0°C from winter through spring, temperatures below a depth of 6 meters remained above 0°C year-round.
Despite the land surface physical model indicating the presence of permafrost near the summit of Mount Poroshiri, no actual permafrost was found. One plausible explanation for this discrepancy is that the soil physical properties assumed in the model differ from those of the actual soils in the Hidaka Mountains. To investigate this, we modified soil physical parameters within the model and analyzed the resulting changes in permafrost distribution in the mountainous regions of Hokkaido. Our analysis demonstrated that soil thermal conductivity plays a critical role in determining the distribution of mountain permafrost.
In this study, we used 1 km resolution temperature projections for Japan as input to the Integrated Land Simulator (ILS), a land surface physical model capable of estimating subsurface physical processes. ILS uses atmospheric variables—such as air temperature, precipitation, radiation, wind speed, and humidity—to simulate near-surface physical processes and calculate ground temperature, soil moisture, river discharge, and runoff etc. The 1 km climate forcing used here was developed using the latest climate projections and bias-correction techniques to support stakeholders, including local governments and private companies, in developing climate change adaptation strategies. We estimated the distribution of permafrost in Japan by applying surface physical conditions from the pre-industrial period to the present within ILS.
The ILS simulation results revealed that permafrost exists over a much narrower area compared with the distribution estimated from temperature projection data and empirical relationships used in our previous study (Yokohata et al., 2022). This discrepancy is likely due, at least in part, to the empirical relationships having been developed from environments such as Alaska and Siberia, which differ significantly from Japanese conditions. Meanwhile, previous observational studies have shown that permafrost in areas such as the Daisetsu Mountains is typically found in wind-exposed sites where strong winter winds inhibit snow accumulation. To account for this effect, we conducted numerical experiments by modifying model parameters to reduce snow depth. The results indicated that decreasing snow depth—representing wind-exposed conditions—led to a broader permafrost distribution. Using model parameter settings that successfully reproduced the observed permafrost distribution, we performed both historical climate reconstructions and future projection experiments to estimate the current permafrost distribution in Japan and its potential future changes.
Our estimates suggest that mountain permafrost is currently distributed extensively in the Hidaka Mountains. Therefore, since June 2024, we have conducted ground temperature observations near the summit of Mount Poroshiri in the Hidaka Mountains of Hokkaido. Permafrost is defined as ground that remains at or below 0°C for at least two consecutive years; however, our observations revealed no such persistently frozen layer within the upper 10 meters of subsurface near the summit of Mount Poroshiri. Although shallow layers fall below 0°C from winter through spring, temperatures below a depth of 6 meters remained above 0°C year-round.
Despite the land surface physical model indicating the presence of permafrost near the summit of Mount Poroshiri, no actual permafrost was found. One plausible explanation for this discrepancy is that the soil physical properties assumed in the model differ from those of the actual soils in the Hidaka Mountains. To investigate this, we modified soil physical parameters within the model and analyzed the resulting changes in permafrost distribution in the mountainous regions of Hokkaido. Our analysis demonstrated that soil thermal conductivity plays a critical role in determining the distribution of mountain permafrost.