Presentation Information
[O12-P14]Hot Spring Water Chemistry Changes Before the May 2025 Minamidake Eruptions at Sakurajima Volcano
*Sotaro Seki1, *Mio Kawaji1 (1. Ikeda Senior Highschool)
Keywords:
hot spring volcanic gas components,magma degassing,B type earthquake counts
Objective
At Sakurajima, magma intrusion can cause volcanic gases to dissolve into hot-spring waters. Ota1) and Iguchi2) reported that variations in volcanic-gas components in Sakurajima's hot springs correspond to volcanic activity. Prior studies3)4) propose a model (Fig.1) in which CO2 and SO2 exsolve from deep magma, while water and chlorine degas in a shallow magma reservoir beneath the crater, triggering eruptions. Adopting a similar model for Sakurajima, we began monitoring hot-spring volcanic-gas components to forecast activity.
Methods
2-1 Sampling
From May 2024 we collected hot-spring water monthly (mid-month) at Arimura Onsen (south; closed April 2025, thereafter Furusato Onsen) and Shirahama Onsen (north), and seawater at nearby ports (Fig.2). Following the Guidelines5), we measured air temperature, atmospheric pressure, water temperature, EC, pH, Na+, K+, Ca2+, F-, NO3-, SO42-, Cl-, and total carbonate.
2-2 Total-carbonate method
Because aqueous carbonate species are hard to measure, we developed a simple total-carbonate method (Fig.3) and calibrated it (Fig.4).
Results and Discussion
We compared gas concentrations with monthly eruption/explosion counts6), total ashfall7) within 20 km at 33 sites (index of magma discharge) (Fig.5-9), and monthly B-type earthquake counts8) linked to bubble formation (Fig.10).
3-1 F- and CO2
F- and CO2 rose in May 2025, coinciding with increased eruptions and ashfall, and were elevated from two months earlier (Fig.6). Thus, F- and CO2 appear useful for forecasting. Total-carbonate variations align with Fig.10 and likely reflect reservoir degassing.
3-2 Cl- and [Cl-]/[Na+]
Cl- is affected by seawater and meteoric input (Fig.11); we corrected this using the molar ratio [Cl-]/[Na+] (Fig.12). The ratio rose about four months before increased eruptions and ashfall, then fell during peak activity in May, which we interpret as reduced steam and overall degassing during eruption.
3-3 SO42-
SO42- tended to increase with eruptions and ashfall; it rose in June 2025, consistent with reservoir degassing (Fig.10).
3-4 B-type earthquakes and a pre-May hydrothermal model
Time-lag correlations between monthly B-type earthquake counts and gas concentrations (Tables 6,7) and references 9-13) support the model in Fig.11: before activity intensifies, B-type earthquakes increase, magma-derived gases feed hydrothermal reservoirs, and hot spring concentrations rise. CO2 reaches the Shirahama reservoir 30-15 days before eruption. HCl and SO2 begin supplying Arimura 20-10 days before eruption, with CO2 added from 15 days prior. SO2 and HCl at Shirahama show low correlation with B-type quakes, suggesting supply from the Aira caldera reservoir or conduits. HF at both springs also shows low correlation; as activity and magma temperature rise, HF likely degasses from surrounding magma into hydrothermal reservoirs. Continuous eruptions then follow.
Conclusions and Future plan
We established hot-spring water analyses, including a simple total-carbonate method. Rises in total carbonate, [Cl-]/[Na+], and F- preceded the May intensification of activity. Based on B-type earthquake variations linked to shallow-reservoir degassing, we proposed a Sakurajima hydrothermal-system variation model. Continued monitoring is needed to validate the model and enable volcanic-activity forecasting from gas observations.
References
1)K. Ota, Report of the 5th Intensive Observation of Sakurajima Volcano, 1986, 103-114
2)M. Iguchi et al., 2014: "Study to Elucidate the Eruption Preparation Process at Sakurajima Volcano Based on Multi-parameter Observations," 93-96
3)N. Spilliaert et al., Earth and Planetary Science Letters 248, 2006, 772-786
4)M. Edmonds et al., Elements, Vol.13, No.1, 2017, 29-34
5)Ministry of the Environment, Guidelines for Analysis of Mineral Spring Water, 2014
6)Kagoshima Local Meteorological Observatory Web, Sakurajima eruption observation table
7)Kagoshima Prefecture official Web, Sakurajima ashfall observation results
8)Kyushu Regional Meteorological Observatory, Sakurajima volcanic activity briefing (May 2025), 1-14
9)K. Nogami et al., Annual Report, Disaster Prevention Research Institute, Kyoto University, No.47B, 2004
10)https://gbank.gsj.jp/volcano/Act_Vol/sakurajima/text/exp01-1.html
11)0190214_01sakurajima_web.pdf
12)https://www.tohoku.ac.jp/japanese/newimg/pressimg/tohokuunivpress0190214_01sakurajima_web.pdf
13)M.Iguchi, Journal of Geophysical Exploration, Vol.60-2, 2007, 145-154
At Sakurajima, magma intrusion can cause volcanic gases to dissolve into hot-spring waters. Ota1) and Iguchi2) reported that variations in volcanic-gas components in Sakurajima's hot springs correspond to volcanic activity. Prior studies3)4) propose a model (Fig.1) in which CO2 and SO2 exsolve from deep magma, while water and chlorine degas in a shallow magma reservoir beneath the crater, triggering eruptions. Adopting a similar model for Sakurajima, we began monitoring hot-spring volcanic-gas components to forecast activity.
Methods
2-1 Sampling
From May 2024 we collected hot-spring water monthly (mid-month) at Arimura Onsen (south; closed April 2025, thereafter Furusato Onsen) and Shirahama Onsen (north), and seawater at nearby ports (Fig.2). Following the Guidelines5), we measured air temperature, atmospheric pressure, water temperature, EC, pH, Na+, K+, Ca2+, F-, NO3-, SO42-, Cl-, and total carbonate.
2-2 Total-carbonate method
Because aqueous carbonate species are hard to measure, we developed a simple total-carbonate method (Fig.3) and calibrated it (Fig.4).
Results and Discussion
We compared gas concentrations with monthly eruption/explosion counts6), total ashfall7) within 20 km at 33 sites (index of magma discharge) (Fig.5-9), and monthly B-type earthquake counts8) linked to bubble formation (Fig.10).
3-1 F- and CO2
F- and CO2 rose in May 2025, coinciding with increased eruptions and ashfall, and were elevated from two months earlier (Fig.6). Thus, F- and CO2 appear useful for forecasting. Total-carbonate variations align with Fig.10 and likely reflect reservoir degassing.
3-2 Cl- and [Cl-]/[Na+]
Cl- is affected by seawater and meteoric input (Fig.11); we corrected this using the molar ratio [Cl-]/[Na+] (Fig.12). The ratio rose about four months before increased eruptions and ashfall, then fell during peak activity in May, which we interpret as reduced steam and overall degassing during eruption.
3-3 SO42-
SO42- tended to increase with eruptions and ashfall; it rose in June 2025, consistent with reservoir degassing (Fig.10).
3-4 B-type earthquakes and a pre-May hydrothermal model
Time-lag correlations between monthly B-type earthquake counts and gas concentrations (Tables 6,7) and references 9-13) support the model in Fig.11: before activity intensifies, B-type earthquakes increase, magma-derived gases feed hydrothermal reservoirs, and hot spring concentrations rise. CO2 reaches the Shirahama reservoir 30-15 days before eruption. HCl and SO2 begin supplying Arimura 20-10 days before eruption, with CO2 added from 15 days prior. SO2 and HCl at Shirahama show low correlation with B-type quakes, suggesting supply from the Aira caldera reservoir or conduits. HF at both springs also shows low correlation; as activity and magma temperature rise, HF likely degasses from surrounding magma into hydrothermal reservoirs. Continuous eruptions then follow.
Conclusions and Future plan
We established hot-spring water analyses, including a simple total-carbonate method. Rises in total carbonate, [Cl-]/[Na+], and F- preceded the May intensification of activity. Based on B-type earthquake variations linked to shallow-reservoir degassing, we proposed a Sakurajima hydrothermal-system variation model. Continued monitoring is needed to validate the model and enable volcanic-activity forecasting from gas observations.
References
1)K. Ota, Report of the 5th Intensive Observation of Sakurajima Volcano, 1986, 103-114
2)M. Iguchi et al., 2014: "Study to Elucidate the Eruption Preparation Process at Sakurajima Volcano Based on Multi-parameter Observations," 93-96
3)N. Spilliaert et al., Earth and Planetary Science Letters 248, 2006, 772-786
4)M. Edmonds et al., Elements, Vol.13, No.1, 2017, 29-34
5)Ministry of the Environment, Guidelines for Analysis of Mineral Spring Water, 2014
6)Kagoshima Local Meteorological Observatory Web, Sakurajima eruption observation table
7)Kagoshima Prefecture official Web, Sakurajima ashfall observation results
8)Kyushu Regional Meteorological Observatory, Sakurajima volcanic activity briefing (May 2025), 1-14
9)K. Nogami et al., Annual Report, Disaster Prevention Research Institute, Kyoto University, No.47B, 2004
10)https://gbank.gsj.jp/volcano/Act_Vol/sakurajima/text/exp01-1.html
11)0190214_01sakurajima_web.pdf
12)https://www.tohoku.ac.jp/japanese/newimg/pressimg/tohokuunivpress0190214_01sakurajima_web.pdf
13)M.Iguchi, Journal of Geophysical Exploration, Vol.60-2, 2007, 145-154