Skip to main content

Advertisement

We're creating a new version of this page. See preview

  • Full paper
  • Open Access

Time variations in the chemical and isotopic composition of fumarolic gases at Hakone volcano, Honshu Island, Japan, over the earthquake swarm and eruption in 2015, interpreted by magma sealing model

  • 1Email author,
  • 2,
  • 1,
  • 1,
  • 3,
  • 4, 5,
  • 4 and
  • 4
Earth, Planets and Space201971:48

https://doi.org/10.1186/s40623-019-1027-5

  • Received: 7 April 2018
  • Accepted: 15 April 2019
  • Published:

Abstract

Definite increases in the components ratios of CO2/H2O, CO2/H2S, CO2/CH4 and He/CH4 were observed at the fumarolic gases from Owakudani geothermal area located at the center of Hakone volcanic caldera (Honshu Island, Japan), synchronized with the earthquake swarm in 2015. Such variations were due to the dominance of a magmatic component over a hydrothermal component, suggesting the earthquake swarm was produced by the injection of magmatic gases into the hydrothermal system. The CO2/H2O ratio of magmatic gas was estimated to be 0.0045 before the earthquake swarm, which increased up to 0.013 during the earthquake swarm, likely produced by the pressurization of magma as a result of magma sealing where the pressure increment in magma was estimated to be 3% to the lithostatic pressure. The H2O and CO2 concentration in magma were estimated to be 6.3 wt% and 20 wt ppm, respectively, assuming a temperature 900 °C and a rhyolitic composition. In May 2015, a few months prior to the earthquake swarm in May 2015, a sharp increase in the Ar/CO2 and N2/He ratios and a decrease in the isotopic ratio of H2O were observed at the fumarolic gas. The invasion of air into the hydrothermal system increased the Ar/CO2 and N2/He ratios. The decrease in the isotopic ratio of H2O was induced by partial condensation of H2O vapor.
Graphical Abstract image

Keywords

  • Active volcano
  • Fumarolic gas
  • Earthquake swarm
  • Phreatic eruption
  • Mt. Hakone
  • Magma degassing
  • Magma sealing

Introduction

Mt. Hakone is an active volcano located on Honshu Island of Japan (Fig. 1a). The volcano has a caldera structure with several central cones (Fig. 1b). Three geothermal areas: Owakukdani (Ow in Fig. 1c), Kamiyuba (Ky) and Sounzan (So) developed as the foothill of Kamiyama (Ka) central cone. The history of caldera formation was first proposed by Kuno (1950) and modified by Mannen (2008). The geothermal area of Ow is a popular sightseeing spot visited by two million tourists every year. In order to prevent human injury by possible eruption, a monitor program was started in the 1970s at Mt. Hakone.
Fig. 1
Fig. 1

Location of Hakone volcano and the sampling points of fumarolic gases. a Map of the Japan and the location of Hakone volcano. b Topographic map of Hakone volcano with caldera structure. The bold dashed curve outlines the caldera rim. L, Ka and Ko represent Lake Ashinoko, Kamiyama central cone and Komagatake central cone, respectively. The ellipses marked E1, E2, E3, E4 and E5 indicate regions of hypocenters of earthquake swarms determined by Hot Springs Research Institute of Kanagawa Prefecture (HSRI). c Solid circles (N and S) indicate the fumaroles where volcanic gas was sampled in this study. Solid diamond (C) is the newly created crater at the eruption in 2015. Open circles and open square indicate location of other fumaroles (circles) and borehole, respectively. Bold lines and curves with teeth indicate the roads and cliffs, respectively

Eruptive activity at the central cones took place 50 ka ago and continued until 3 ka ago (Kobayashi et al. 1997), when a phreatic eruption occurred. With the phreatic eruption, a section of the western flank of Ka (Fig. 1c) collapsed and caused debris flow, damming a river to form Lake Ashinoko in Fig. 1b. No historical eruption has been recorded at Hakone volcano before the small phreatic eruption in 2015 at the geothermal area of Ow.

Based on the analysis of seismic wave velocities beneath Hakone caldera, Yukutake et al. (2015) recognized a region with high Vp/Vs ratio beneath Ko (Fig. 1c) likely due to a degassing magma as the depth of − 10 km. Above the degassing magma, a region with low Vp/Vs ratio was found, which was estimated to be a gas reservoir dominated by H2O and CO2. In 2001 to 2009, volcanic gases and borehole gases were analyzed by Ohba et al. (2011). The CO2/H2O and CO2/H2S ratios were high in 2001 when a seismic swarm occurred. It increased again in 2006 and 2008 when a new seismic swarm resumed. The increases in CO2/H2O and CO2/H2S ratios suggested the injection of a CO2-rich magmatic gas into the hydrothermal system, implying that a magmatic component was episodically supplied to the hydrothermal system in 2006 and 2008. The earthquake swarms resulted from the injection of the gas through the shallow crust accompanying the break of the so-called sealing zone (Fournier 1999). Daita (2013) monitored fumarole S (Fig. 1c) at Ky geothermal area in 2012 and 2013 by using a gas detector tubes system (GASTEC corp.), when a sharp increase in CO2/H2S ratio, synchronized with the start of seismic swarm, was found.

At Hakone volcano, an earthquake swarm started on April, 26, 2015. On June, 30, 2015, a small phreatic eruption occurred at the geothermal area of Ow. After the eruption, several craters formed (Mannen et al. 2018). The C crater in Fig. 1c is one of the newly born craters. In this study, we repeatedly sampled fumarolic gas at Ow and Ky geothermal areas for investigating the relationship between volcanic activity and volcanic gas chemistry.

Sampling and analysis of fumarolic gases

Fumarolic gases were sampled every month from May 2013 until February 2018, at the points N and S (Fig. 1c). In addition, fumarolic gases were sampled at point C (Fig. 1c). The outlet temperature of fumarolic gas was measured by a thermocouple with K-type sensor. Gas sampling was carried out by a titanium pipe inserted into the fumarole which was then connected to a rubber tube. The rubber tube was connected to a 120-ml pre-evacuated Pyrex glass bottle with airtight stop cock (Giggenbach 1975), where 20 ml of 5 M KOH solution was added. Steam-condensated and acidic gases (CO2, SO2, H2S, HCl etc.) were absorbed by the KOH solution, whereas the residual gases (hereafter R-gas), e.g., N2, O2, Ar, He, H2 and CH4, enriched in the headspace of glass bottle. Components in the KOH solution were analyzed according to the method by Ozawa (1968). The SO2/H2S ratio in the fumarolic gas was determined using a KI–KIO3 solution (Ozawa, 1968), the analytical accuracy of which was evaluated by Lee et al. (2016). The analysis of HCl was only carried out at the fumarole C. The total molar amount of R-gas was given by the head space volume of bottle and the inner pressure of head space as room temperature. Based on the molar amount of H2O, CO2, H2S, SO2 and R-gas, the relative concentration (µmol/mol) of those components was calculated (Tables 1, 2 and 3).
Table 1

Chemical and isotopic composition of fumarolic gas sampled at the point N in Owakudani geothermal area, Hakone volcano, Japan

Date

Temp

*CO2/H2S

H2O

CO2

H2S

SO2

He

H2

O2

N2

CH4

Ar

δDSMOW

δ18OSMOW

δDSMOW

**AET

yy/mm/dd

°C

molar ratio

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

H2O

H2O

H2

°C

2013/5/7

95.6

 

974,800

20,600

4020

92.7

0.0678

209

0.00

251

0.668

0.670

− 53

− 7.7

− 579

141

2013/6/4

97.5

 

981,400

15,160

3040

105

0.0305

158

0.00

173

0.455

1.664

− 55

− 7.2

− 608

115

2013/7/9

96.8

 

988,200

9360

2190

39.1

0.0316

123

0.00

107

0.377

0.296

− 51

− 6.8

− 619

105

2013/8/28

97.5

 

989,300

8320

2140

8.2

0.0265

110

0.00

88

0.337

0.291

− 47

− 5.7

− 618

105

2013/10/2

97.0

 

986,400

9000

3660

44.8

0.0581

46

0.00

840

0.681

14.3

− 58

− 7.0

− 626

102

2013/11/29

96.6

 

987,900

8610

3170

29.3

0.0285

101

0.31

236

0.576

2.06

− 62

− 8.7

− 619

109

2014/1/28

96.8

 

988,400

8610

2790

75.9

0.0260

81

0.41

85

0.461

0.169

− 54

− 6.6

− 625

101

2014/4/25

96.9

 

992,900

5160

1810

24.4

0.0104

61

1.20

50

0.222

0.252

− 45

− 5.3

− 620

102

2014/5/30

98.7

 

991,500

6160

2120

36.8

0.0164

101

0.00

78

0.548

0.355

− 43

− 4.7

− 622

100

2014/7/4

97.2

 

991,800

6140

1940

37.1

0.0190

43

0.31

76

0.388

0.347

− 50

− 5.9

− 626

99

2014/8/6

96.3

 

991,200

6480

2170

25.3

0.0263

45

0.00

89

0.488

0.452

− 50

− 5.7

− 615

108

2014/9/12

96.8

 

990,300

7140

2350

73.7

0.0309

60

0.47

106

0.591

0.488

− 50

− 6.1

− 627

98

2014/10/15

96.8

3.5

988,600

8910

2260

76.7

0.0320

70

0.44

116

0.641

0.600

− 55

− 7.0

− 662

72

2014/11/19

97.1

3.3

991,500

6420

1920

33.6

0.0169

31

0.50

64

0.376

0.289

− 50

− 6.0

− 654

77

2014/12/19

96.8

3.4

991,800

6110

1960

18.2

      

− 51

− 6.3

− 675

61

2015/1/20

97

3.2

992,000

5820

1950

27.9

0.0151

31

18.82

126

0.364

1.03

− 51

− 6.4

− 625

100

2015/2/16

96.5

3.1

990,800

6380

2370

40.4

0.0217

19

0.12

384

0.537

5.86

− 57

− 7.4

− 595

128

2015/3/31

97.2

3.4

989,700

5940

2260

44.0

0.0294

19

249

1750

0.522

29.9

− 60

− 8.4

− 602

123

2015/4/24

95.9

3.6

984,400

6640

3550

34.9

0.0591

36

3.76

5240

0.408

90.4

− 67

− 10.2

− 581

144

2015/5/8

95.6

4.4

988,700

8670

2400

29.0

0.0365

24

0.00

133

0.373

0.364

− 56

− 7.2

− 600

123

2015/6/2

96.1

9.8

979,300

18,250

2150

68.1

0.0718

26

0.09

230

0.276

0.477

− 52

− 6.7

− 607

116

2015/6/16

96.8

9.0

976,000

21,180

2560

30.5

0.0777

30

0.18

244

0.306

0.547

− 52

− 6.8

− 589

131

2015/7/14

96.6

6.8

971,600

24,110

3950

70.0

0.0816

49

0.19

247

0.313

0.647

− 51

− 6.7

− 611

112

2015/7/28

96.3

5.8

979,300

16,980

3480

37.8

0.0657

45

0.05

199

0.255

0.402

− 55

− 7.4

− 613

112

2015/8/20

96.1

4.6

977,200

18,160

4360

33.2

0.0633

46

0.05

204

0.312

0.441

− 55

− 7.2

− 605

118

2015/9/11

96.1

4.4

983,400

13,000

3420

32.7

0.0444

47

0.08

146

0.272

0.316

− 56

− 7.6

− 576

144

2015/10/6

97.2

4.3

984,300

11,990

3490

34.4

0.0360

35

0.05

130

0.257

0.406

− 55

− 7.3

− 608

116

2015/11/11

96.9

4.1

990,100

6590

3150

20.3

0.0307

33

0.07

106

0.252

0.348

− 54

− 7.1

− 602

120

2015/12/8

96.7

4.3

985,300

11,300

3160

29.8

0.0420

54

2.94

187

0.503

1.378

− 58

− 8.2

− 609

116

2016/1/8

96

3.3

983,000

12,580

4190

27.6

0.0339

48

1.51

198

0.517

0.997

− 52

− 6.5

− 579

140

2016/1/28

97.7

3.7

988,400

8290

3160

20.8

0.0260

32

0.36

99

0.367

0.364

− 66

− 8.9

− 612

116

2016/2/15

96.5

3.9

982,200

13,850

3810

16.7

0.0254

40

0.08

105

0.399

0.317

− 63

− 8.0

− 616

112

2016/3/2

96.5

3.8

986,800

9540

3470

36.5

0.0252

35

0.20

113

0.346

0.709

− 62

− 8.1

− 617

110

2016/4/8

96.5

3.2

989,000

7780

3050

74.1

0.0227

35

0.09

97

0.419

0.585

− 64

− 8.8

− 613

114

2016/5/13

96.5

3.3

988,900

7520

3280

29.4

0.0173

30

0.15

276

0.394

4.213

− 65

− 8.1

− 619

110

2016/6/3

96.7

3.3

990,900

6190

2710

19.9

0.0170

26

4.09

127

0.392

1.187

− 66

− 8.4

− 615

114

2016/7/5

97.7

3.1

993,300

3640

2980

15.9

0.0135

25

0.06

78

0.353

0.625

− 62

− 8.3

− 618

109

2016/8/5

96.9

3.7

994,000

3480

2320

29.3

0.0160

25

0.11

108

0.380

0.958

− 63

− 8.1

− 613

114

2016/9/5

97

3.2

991,900

5390

2610

26.4

0.0155

53

0.07

62

0.485

0.283

− 61

− 8.5

− 609

117

2016/10/7

96.9

2.7

990,900

5590

2670

9.5

0.0153

29

123

685

0.553

9.71

− 68

− 9.7

− 611

117

2016/10/19

97.2

2.7

992,300

5080

2450

23.9

0.0128

57

0.15

54

0.495

0.236

− 57

− 7.9

− 634

95

2016/11/4

96.9

2.9

993,300

4120

2390

25.7

0.0123

58

0.25

55

0.507

0.229

− 56

− 7.6

− 615

110

2016/12/7

96.4

3.1

987,500

8080

3870

90.2

0.0218

87

3.71

399

0.856

6.15

− 64

− 9.1

− 616

112

2017/1/11

96.3

3.2

992,600

5030

2210

56.9

0.0097

53

0.04

49

0.537

0.221

− 59

− 8.2

− 613

112

2017/2/6

95.9

2.9

991,700

5430

2670

92.9

0.0134

60

0.18

59

0.688

0.259

− 58

− 7.8

− 634

95

2017/3/9

95.7

2.8

993,300

4210

2340

81.7

0.0126

53

0.67

52

0.579

0.256

− 55

− 7.4

− 626

101

2017/4/10

96.2

2.5

993,900

3400

2540

45.8

0.0100

47

0.31

51

0.591

0.268

− 59

− 8.0

− 624

103

2017/5/10

96.3

3.1

          

− 84

− 12.4

− 617

118

2017/6/10

96.1

3.1

982,400

11,400

5640

267

0.0359

101

0.10

144

1.396

0.749

− 63

− 8.8

− 606

120

2017/7/4

96.4

3.2

991,800

5550

2480

22.3

0.0173

45

0.01

65

0.612

0.240

− 56

− 7.4

− 566

153

2017/8/1

96.5

3.5

991,500

6210

2160

17.6

0.0156

37

0.37

72

0.493

0.248

− 54

− 7.0

− 609

115

2017/9/2

96.9

3.7

990,200

7250

2360

22.6

0.0247

45

0.09

105

0.558

0.368

− 57

− 7.7

− 588

133

2017/10/4

95.7

4.0

          

− 72

− 11.0

− 535

192

2017/11/7

96.8

3.9

988,900

8190

2660

19.4

0.0258

48

0.05

154

0.715

1.04

− 67

− 9.5

− 581

144

2017/12/5

96.0

3.9

991,000

6430

2380

25.3

0.0199

41

0.29

87

0.656

0.283

− 61

− 8.5

− 609

117

2018/1/9

95.5

3.7

          

− 77

− 12.2

− 604

127

2018/2/5

97.0

3.9

989,400

7610

2860

3.00

0.0220

70

0.17

89

0.862

0.360

− 62

− 8.8

− 603

122

*Ratio determined by “Gas detector tubes system”

**Apparent equilibrium temperature defined between δD of H2O and H2

Table 2

Chemical and isotopic composition of fumarolic gas sampled at the point S in Kamiyuba geothermal area, Hakone volcano, Japan

Date

Temp

*CO2/H2S

H2O

CO2

H2S

SO2

He

H2

O2

N2

CH4

Ar

δDSMOW

δ18OSMOW

δDSMOW

**AET

yy/mm/dd

°C

Molar ratio

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

H2O

H2O

H2

°C

2013/5/7

98.1

 

978,600

20,200

500

12.2

0.0609

415

0.000

287

1.47

0.735

− 52

− 6.8

− 636

91

2013/6/4

96.5

 

986,500

12,540

396

7.00

0.0536

343

0.000

235

1.32

0.551

− 54

− 7.1

− 638

90

2013/7/9

97.5

 

985,400

13,770

381

9.50

0.0392

302

0.000

187

1.23

0.519

− 52

− 7.4

− 637

91

2013/8/28

97.3

 

983,000

15,940

499

5.90

0.0409

341

0.000

203

1.33

0.701

− 47

− 5.9

− 630

95

2013/10/2

96.5

 

989,800

9410

365

9.00

0.0276

289

0.000

144

1.06

0.472

− 51

− 6.2

− 636

91

2013/11/29

96.5

 

991,900

7210

465

0.50

0.0279

252

0.000

142

1.21

0.292

− 49

− 6.0

− 634

93

2014/1/28

96.9

 

990,600

8300

607

19.3

0.0349

277

0.000

155

1.39

0.363

− 50

− 5.9

− 631

95

2014/4/25

98.2

 

988,900

10,250

429

19.4

0.0267

286

0.000

153

1.31

0.543

− 50

− 6.0

− 629

96

2014/5/30

97.1

 

988,700

10,340

473

10.5

0.0279

273

0.000

155

1.21

0.637

− 48

− 5.8

− 631

94

2014/7/4

97.2

 

989,100

9890

529

6.20

0.0395

273

0.147

150

1.23

0.567

− 51

− 6.1

− 633

93

2014/8/6

97

 

989,900

9200

430

9.50

0.0263

277

0.122

141

1.34

0.426

− 49

− 5.9

− 627

98

2014/9/12

97.4

 

988,500

10,430

586

18.4

0.0367

314

0.000

174

1.70

0.534

− 50

− 6.0

− 655

76

2014/10/15

97.1

23

986,900

12,160

484

18.0

0.0466

233

0.141

171

1.66

0.446

− 53

− 6.7

− 666

68

2014/11/19

97

23

988,600

10,210

868

17.6

0.0405

205

0.307

127

1.24

0.403

− 49

− 5.9

− 630

95

2014/12/19

97.2

21

989,300

9850

517

20.9

0.0283

177

0.347

129

1.42

1.05

− 49

− 6.0

− 628

97

2015/1/20

98.6

20

990,100

8960

674

23.8

0.0231

159

0.314

117

1.15

0.338

− 48

− 5.8

− 631

94

2015/2/16

97.3

19

988,600

10,550

517

8.50

0.0413

198

0.000

143

1.56

0.423

− 49

− 5.8

− 624

100

2015/3/31

98.7

19

991,100

8170

398

20.2

0.0315

167

0.000

113

1.28

0.320

− 49

− 5.7

− 611

111

2015/4/24

97.6

20

990,200

8890

577

11.7

0.0381

184

0.146

128

1.32

0.310

− 50

− 6.2

− 609

113

2015/5/8

97.3

31

984,000

14,920

668

15.0

0.0645

183

0.000

238

1.24

0.426

− 52

− 6.3

− 616

108

2015/5/11

97.6

38

981,400

17,600

552

7.80

0.0831

190

0.000

280

1.18

0.466

− 54

− 6.8

− 564

154

2015/5/19

96.6

47

973,300

24,540

1550

40.9

0.106

236

0.625

363

1.21

0.603

− 53

− 6.5

− 589

131

2015/6/2

96.2

54

975,100

23,830

481

14.6

0.0969

227

0.261

323

1.11

0.722

− 48

− 5.9

− 607

114

2015/6/16

97.1

53

975,200

23,820

477

9.20

0.0878

234

0.394

299

1.10

0.493

− 49

− 6.0

− 616

106

2015/6/30

97.8

59

975,300

23,520

605

19.1

0.0835

216

0.000

291

1.13

0.586

− 47

− 5.5

− 606

114

2015/7/14

98.1

48

977,400

21,520

511

3.00

0.0930

280

0.256

295

1.44

0.676

− 55

− 7.2

− 606

117

2015/7/28

97

41

978,600

20,320

611

9.60

0.0663

245

0.033

236

1.35

0.509

− 54

− 7.1

− 595

126

2015/8/20

97.8

37

978,500

20,270

641

22.4

0.0540

300

0.000

255

1.47

0.556

− 50

− 6.3

− 598

123

2015/9/11

96.6

35

981,600

17,400

507

6.70

0.0493

300

0.021

226

1.37

0.542

− 51

− 6.7

− 618

106

2015/10/6

97.6

33

966,200

31,130

1820

38.0

0.0855

422

0.000

406

2.31

1.12

− 60

− 8.0

− 607

118

2015/11/11

97.7

31

983,300

14,460

614

5.60

0.0562

315

173

1160

1.54

10.9

− 53

− 6.7

− 603

119

2015/12/8

96.6

28

981,700

17,360

474

10.4

0.0424

272

1.63

225

1.47

1.83

− 49

− 6.1

− 608

114

2016/1/8

96.7

29

982,300

16,700

583

6.90

0.0478

231

1.10

188

1.37

0.381

− 63

− 8.7

− 595

129

2016/1/28

97.9

29

979,900

18,900

656

11.3

0.0627

268

0.710

241

1.86

0.699

− 53

− 6.9

− 609

114

2016/2/15

95.5

29

990,700

8510

454

6.20

0.0434

209

0.106

154

1.20

0.347

− 50

− 6.4

− 607

114

2016/3/2

95.5

28

982,600

16,320

647

15.7

0.0482

211

0.035

193

1.44

0.673

− 55

− 7.4

− 611

113

2016/4/8

97.6

28

985,000

14,090

561

13.7

0.0459

207

0.000

174

1.41

0.620

− 50

− 6.8

− 609

113

2016/5/13

97.3

26

986,300

12,830

559

8.10

0.0388

169

0.319

160

1.23

0.708

− 51

− 6.9

− 605

116

2016/6/3

97.6

23

986,300

12,820

578

1.60

0.0376

178

0.343

161

1.32

0.503

− 51

− 6.6

− 606

116

2016/7/5

97.4

23

983,700

15,110

744

3.50

0.0484

204

0.089

203

1.91

0.640

− 57

− 7.7

− 610

114

2016/8/5

96.9

26

989,600

9520

582

 

0.0415

173

0.019

154

1.50

0.451

− 51

− 6.7

− 607

115

2016/9/5

97.4

21

989,200

10,070

503

8.00

0.0305

120

0.018

129

1.30

0.443

− 49

− 6.3

− 613

109

2016/10/7

97.2

22

989,000

10,190

422

10.6

0.0347

214

0.584

135

1.35

0.451

− 51

− 7.4

− 615

109

2016/10/19

96.6

21

990,300

8960

445

13.5

0.0293

207

0.568

115

1.19

0.387

− 47

− 6.3

− 610

111

2016/11/4

97.5

22

988,800

10,300

490

7.70

0.0369

231

1.56

146

1.44

0.579

− 49

− 5.8

− 605

116

2016/12/7

97.2

21

989,800

9370

396

9.00

0.0369

258

0.064

150

1.58

1.15

− 48

− 5.7

− 612

109

2017/1/11

96.8

22

982,400

16,360

766

16.7

0.0529

257

0.033

207

2.49

0.663

− 51

− 6.6

− 611

112

2017/2/6

96.7

20

976,400

21,970

1010

13.0

0.0579

361

0.026

287

3.36

1.06

− 59

− 7.5

− 618

108

2017/3/9

96.7

21

980,700

17,670

1100

14.2

0.0552

239

0.236

230

2.67

1.07

− 56

− 7.3

− 626

101

2017/4/10

92.4

20

983,000

15,160

1380

44.9

0.0472

154

0.434

222

2.76

1.07

− 56

− 7.5

− 625

101

2017/5/10

97.2

23

984,000

14,730

905

5.40

0.0505

196

0.046

203

2.55

0.674

− 57

− 7.4

− 622

104

2017/6/10

96.5

21

975,300

22,640

1470

63.3

0.0772

227

0.000

307

3.81

1.14

− 58

− 7.4

− 604

120

2017/7/4

97.6

23

983,100

15,820

744

4.60

0.0461

107

0.168

203

2.17

0.696

− 56

− 7.0

− 589

132

2017/8/1

96.8

26

978,500

20,250

844

18.3

0.0653

139

0.121

278

2.47

0.820

− 56

− 7.4

− 584

137

2017/9/2

96.4

26

970,400

28,060

1000

17.2

0.0858

155

0.166

364

2.96

1.12

− 58

− 7.7

− 606

118

2017/10/4

96.8

25

973,000

25,220

1010

25.0

0.0716

73.5

26.7

612

2.48

6.49

− 55

− 7.1

− 599

123

2017/11/7

97.2

28

983,800

15,140

548

7.50

0.0473

255

0.000

215

1.64

0.599

− 54

− 6.9

− 618

107

2017/12/5

97.0

27

986,500

12,520

548

30.7

0.0390

258

0.000

186

1.51

0.547

− 49

− 6.0

− 605

116

2018/1/9

96.4

29

988,300

10,910

435

3.50

0.0361

165

0.107

148

1.32

0.403

− 46

− 5.1

− 585

132

2018/2/5

96.7

24

979,200

19,670

775

7.20

0.0604

126

0.040

262

2.40

0.954

− 58

− 7.6

− 589

133

*Ratio determined by “Gas detector tubes system”

**Apparent equilibrium temperature defined between δD of H2O and H2

Table 3

Chemical and isotopic composition of fumarolic gas sampled at the point C in Owakudani geothermal area, Hakone volcano, Japan

Date

Temp

H2O

CO2

H2S

SO2

HCl

He

H2

O2

N2

CH4

Ar

δDSMOW

δ18OSMOW

δDSMOW

*AET

yy/mm/dd

°C

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

μmol/mol

H2O

H2O

H2

°C

2015/12/25

131.9

988,200

5640

2050

1730

2360

0.0163

4.68

0.763

64.3

0.195

0.763

− 30

1.4

  

2016/6/7

124.4

991,100

4630

2240

1120

850

0.0148

1.88

0.210

52.2

0.161

0.151

− 31

0.6

− 307

482

2017/8/7

123

979,300

17,050

1340

2070

0

0.0488

1.32

1.55

188

0.0806

1.19

− 29

1.0

− 254

597

*Apparent equilibrium temperature defined between δD of H2O and H2

The R-gas was analyzed by using two different gas chromatographs with Ar and He as gas carriers (hereafter GC-Ar and CG-He, respectively). Concentration of He, H2, O2, N2 and CH4 was determined by GC-Ar equipped with a 6 m-long MS5A packed column and a TCD detector. The temperature of the column and detector was kept at 50 and 100 °C during analysis, respectively. The flow rate of Ar was 30 ml/min. In general, He analysis by GC is difficult due to the overlap where H2 has high concentrations. Under the above condition, the He of 64 ppm in R-gas could be separated from the H2 of 67% in R-gas without the overlap between the chromatographic peaks of He and H2. All of R-gas sample in this study had a good separation between He and H2 peaks.

The relative concentration of N2 and Ar was determined by GC-He. In the GC-He, equipped with a 6 m-long Gaskuropack-54 column (GC Sciences Inc.) and a TCD detector, the temperature of the column and detector was kept at − 70 and 50 °C, respectively. Under the above conditions, N2, O2 and Ar in R-gas were well separated on the chromatographic chart. The concentration of Ar in R-gas was calculated by,
$$X_{\text{Ar}} = X_{\text{N2}} \frac{{R_{\text{Ar}} }}{{R_{\text{N2}} }},$$
(1)
where X and R indicate the concentration obtained by GC-Ar and GC-He, respectively.

The H2O isotopic ratio was determined by the cooling of the fumarolic gas with a double tube condenser made of Pyrex glass. The fumarolic gas flowing out of the condenser was collected in a 5-l plastic gas sampling bag. The gas collected in the bag was used for on-site determination of CO2/H2S molar ratios by gas detector tube system. We used “2HH” and “4HH” detector tubes for the measurement of CO2 and H2S, respectively. The applicable range of concentration for 2HH and 4HH was 5–40% and 0.1–2.0%, respectively.

The isotopic ratio of condensed water was determined by an IR-laser cavity ring-down analyzer (Picarro Inc., L2120-i). The analytical precision of the analyzer was ± 0.12 and ± 0.05‰ for δD and δ18O, respectively. The isotopic ratio of H2 in R-gas was determined by a continuous flow system combined with a mass spectrometer (Thermo Fisher Scientific Delta V) (Tsunogai et al. 2011). The analytical precision of the analyzer was ± 0.8‰ for δD.

Results

The analytical results of fumarolic gases at N, S and C are listed in Tables 1, 2 and 3, respectively. Time variations of CO2/H2O, H2S/H2O, CO2/H2S, CO2/CH4 and He/CH4 ratios are shown in Fig. 2 with the number of earthquakes per half month at Hakone volcano observed by Japan Meteorological Agency (JMA). On April, 26, 2015, the earthquake swarm started (Mannen et al. 2018). The hypocenters of earthquakes were distributed in the regions E1 to E4 of Fig. 1b (Mannen et al. 2018). The magnitude of earthquakes during the unrest was < 2.9. The largest earthquake was observed on the July, 21, 2015 (Mannen et al. 2018). On the May, 15, 2015, the largest daily number of earthquake (955, M ≥ 0) was observed (Mannen et al. 2018). In the E5 region of Fig. 1b, a small earthquakes swarm took place in 2017 in which the largest earthquake (M = 1.5) occurred on the April 11 (https://www.onken.odawara.kanagawa.jp/earthquake/).
Fig. 2
Fig. 2

Time variation of CO2/H2O, H2S/H2O, CO2/H2S, CO2/CH4 and He/CH4 in fumarolic gases. Solid and open circles indicate the values for the fumarole N and S, respectively. Red curve indicates the number of earthquakes per half month observed at Hakone volcano

In May 2015, the CO2/H2O ratio increased up along with the frequency of earthquakes (Fig. 2a), and then, the earthquake events decreased after May 2015, as well as the CO2/H2O ratio. Similar increases and decreases were observed in the ratios CO2/H2S, CO2/CH4 and He/CH4 along the occurrence of earthquakes in 2015 (Fig. 2c, d, 2, respectively). For the indication in Fig. 2c, the CO2/H2S ratio determined by gas detector tubes system was used. The corresponding CO2/H2S curve determined by chemical analysis was not smooth compared with the curve in Fig. 2c, likely due to the contamination of native sulfur particles during the sampling. There was abundant native sulfur sublimate around the fumarole outlet, because the insertion of a titanium tube during sampling likely disturbed the fumarolic gas discharge, releasing particles of native sulfur. The native sulfur contamination increases the amount of sulfur dissolved in KOH solution.

The gradual decrease in the CO2/H2S, CO2/CH4 and He/CH4 ratios after the maximum in 2015 continued until around March 2017, followed by a gradual increase which peaked in September 2017. The increase in 2017 was not clear in CO2/H2O ratio due to the fluctuation (Fig. 2a). No increase was observed in the H2S/H2O ratio in May 2015 when the earthquake swarm observed (Fig. 2b). Beside small fluctuations, the H2S/H2O ratio was stable over the whole duration of this study.

Time variations of the SO2/H2S, H2/H2O, Ar/CO2 and N2/He ratios are shown in Fig. 3. For the calculation of Ar/CO2 and N2/He ratios, the Ar and N2 brought by the direct contamination of air during sampling were removed by the following equations,
$$C_{\text{Ar}}^{*} = C_{\text{Ar}} - \left( {\frac{0.934}{20.95}} \right) C_{{{\text{O}}_{2} }} ,$$
(2)
$$C_{{{\text{N}}_{2} }}^{*} = C_{{{\text{N}}_{2} }} - \left( {\frac{78.08}{20.95}} \right) C_{{{\text{O}}_{2} }} ,$$
(3)
where Ci is the concentration of i in fumarolic gas and \(C_{i}^{*}\) is the corrected concentration eliminating the effect of direct air contamination. In the equations, it was assumed that the fumarolic gas does not include O2 originally, and the O2 contaminated in fumarolic gas accompanied atmospheric N2 and Ar with the equivalent air abundance. It should be noted that the above correction does not account for O2 lost in the oxidation of reduced S in solution.
Fig. 3
Fig. 3

Time variation of SO2/H2S, H2/H2O, Ar/CO2, N2/He in fumarolic gases. Solid and open circles indicate the values for the fumarole N and S, respectively. Red curve indicates the number of earthquakes per half month observed at Hakone volcano

In general, the SO2/H2S ratio is a useful parameter for evaluating volcanic activity, as it tends to increase when volcanic activity was elevated (e.g., Ossaka et al. 1980). With some fluctuation, SO2/H2S ratio of N and S was < 0.06 over the whole duration of this study (Fig. 3a), being a contrast to the SO2/H2S ratio at C as high as 0.5 to 1.5 (Table 3). The H2/H2O ratio shows a weak correlation with the number of earthquakes; it started to increase slightly after May 2015; then, it maximized in the late of 2015 (Fig. 3b).

The Ar/CO2 and N2/He ratios show a sharp increase a few months prior to the earthquake swarm in May 2015. Actually, Ar/CO2 and N2/He ratios maximized on April, 24, 2015, which was 2 days before the start of earthquake swarm. On May, 8, 2015, those ratios returned to the normal levels. It should be noticed that those precursory increases only happened at fumarole N. No increase was detected at S fumarole.

The time variation in δD of H2O, δ18O of H2O, δD of H2 and the apparent equilibrium temperature (AET) are shown in Fig. 4. The AET was calculated between δD of H2O and δD of H2 based on the following equation (Richet et al. 1977),
$${\text{AET}} = \left( {4.474 \times 10^{ - 12} x^{2} + 3.482^{ - 9} x + 9.007 \times 10^{ - 8} } \right)^{ - 0.5} - 273.15$$
(4)
where
$$x = 1000 \ln \left\{ {\frac{{\delta {\text{D}}\left( {{\text{H}}_{2} {\text{O}}} \right) + 1000}}{{\delta {\text{D}}\left( {{\text{H}}_{2} } \right) + 1000}}} \right\}$$
(5)
Fig. 4
Fig. 4

Time variation of isotopic ratio of H2O and H2 in fumarolic gases with AET. Solid and open circles indicate the values for the fumarole N and S, respectively. Red curve indicates the number of earthquakes per half month observed at Hakone volcano

The averaged H2 concentration in the gas of N, S and C is 2.6, 55 and 233 µmol/mol, respectively. The depth of the borehole, from which the gas of C is discharging, is 500 m below the surface. Therefore, the most of H2 in the gas of N and S is thought to be formed in the crust shallower than − 500 m to surface. The δD of H2 in the gas of C is much higher than the ratio of the gas N and S (see Tables 1, 2 and 3), consistent with the above estimation on H2 formation. The AET calculated between the δD of H2O and H2 means the temperature of crust shallower than − 500 m to surface. The δD and δ18O of H2O decreased significantly a few months prior to the start of earthquake swarm in May 2015 (Fig. 4a, b) at N fumarole. A similar decrease was observed in 2017 at fumarole N, whereas no decrease was observed at fumarole S.

The δD of H2 and AET were stable over the whole duration of this study with the exception of two periods. In September to December 2014, those values significantly decreased relative to normal level (Fig. 4c, d). A slight decrease was also observed in March and April in 2017. It should be noticed that the both decreases happened almost simultaneously at N and S fumaroles.

Discussion

We try to group the components in fumarolic gas by use of triangular diagrams (Fig. 5). The He–N2–Ar ternary diagram shows a linear distribution of data points, suggesting two common-end members in fumarolic gases N, S and C. The end member with high He/Ar and N2/Ar ratios is estimated to be the magmatic component which is ubiquitous at the volcanoes on subduction zone (Kita et al. 1993; Giggenbach 1997). Another end member with low He/Ar and N2/Ar ratios could be air or ASW. The ASW is the atmospheric components saturated in water at 15 °C. A similar feature also found in the diagram for He–CO2–Ar system (Fig. 5), suggesting the fumarolic gases N, S and C, contains a common magmatic component.
Fig. 5
Fig. 5

Triangular diagrams for He–N2–Ar (a) and He–CO2–Ar (b) ternary systems

In Fig. 6a–c, respectively, for He–H2S–Ar, He–H2–Ar and He–CH4–Ar, the distributions of data points are different from the distribution in Fig. 5. No definite value can be found in H2S/He, H2/He and CH4/He ratios for non-atmospheric end member, suggesting H2S, H2 and CH4 belongs to a group different from the magmatic component. As explained by Giggenbach (1987), the reduced molecules group, H2S, H2 and CH4, are originating in hydrothermal system. Those molecules are generated by the reaction between magmatic gas and Fe2+ contained in crustal rock. For example, the following reaction can be written for the formation of CH4.
Fig. 6
Fig. 6

Triangular diagrams for He–H2S–Ar (a), He–H2–Ar (b) and He–CH4–Ar (c) ternary systems

$${\text{CO}}_{2} + 8{\text{Fe}}^{2 + } + 8{\text{H}}^{ + } + 2{\text{H}}_{2} {\text{O}} \to {\text{CH}}_{4} + 8{\text{Fe}}^{3 + } + 4{\text{H}}_{2} {\text{O}}$$
(6)
The behavior of magmatic gas and hydrothermal fluid is illustrated in Fig. 7. The magmatic gases such as CO2 and SO2 are converted to CH4 and H2S in hydrothermal system due to the reduction by Fe2+ in crustal rock. The reduced components are stored in hydrothermal system and entrained by the flow of magmatic gas. The fumarolic gas discharged at the surface contains both components. When the flux of magmatic gas increases, if the flux of hydrothermal entrainment is constant, the magmatic/hydrothermal component ratios such as CO2/CH4 and CO2/H2S are expected to be elevated. This situation is consistent with the observed changes of CO2/H2O, CO2/H2S, CO2/CH4 and He/CH4 ratios in May 2015 (Fig. 2a, c–e). Actually, the flux of hydrothermal component entrained in the flow of magmatic component would not be constant. However, the variation of the flux by entrainment may be much smaller than the variation in the magmatic component flow.
Fig. 7
Fig. 7

Flow of magmatic gas and the entrainment of hydrothermal components

Substituting He by CH4, the triangular diagrams of Fig. 6 can be converted to Fig. 8, where definite non-atmospheric end members can be found for H2S/CH4 ratio (Fig. 8a). It should be noticed that the H2S/CH4 ratio is similar for N and C. The H2S/CH4 ratio for S is much lower than the ratio for N and C. The geothermal area Ky, where fumarole S is located, was developed newly after the earthquake swarm in 2001. Before 2001, the place of Ky was densely forested. After 2001, the trees in Ky area were completely devastated due to geotherm. The averaged H2S concentration in the gas of N, S and C is 2800, 660 and 1900 µmol/mol, respectively. The H2S concentration of the gas S is much lower than that of N and C. The reason for the low H2S/CH4 ratio of S may be the removal of H2S along the reaction with Fe in crustal rock or soil as,
Fig. 8
Fig. 8

Triangular diagrams for CH4–H2S–Ar (a) and CH4–H2–Ar (b) ternary systems

$$2{\text{H}}_{2} {\text{S}} + {\text{FeO}} \to {\text{FeS}}_{2} + {\text{H}}_{2} + {\text{H}}_{2} {\text{O}}$$
(7)

Since Ky area has been developed recently, the crustal rock or soil contacting volcanic gas flow may contain plenty of Fe. The reaction (7) at Ky area may be effective relative to Ow area. The averaged CH4 concentration in the gas of N, S and C is 0.49, 1.6 and 0.15 µmol/mol, respectively. The concentration in the gas S is higher than that of N and C. Because the Ky area is the newly developed geothermal area, the shallow crust is expected to contain much organic matter. A part of CH4 in the gas of S originates in the thermogenic component (Taran and Giggenbach, 2003). The high CH4/H2S ratio of the gas S relative to N and C seems to be brought by the two effects, the removal of H2S and the addition of thermogenic CH4.

Figure 8b shows the existence of non-atmospheric end member for H2/CH4 ratio although the range of ratio is much wider than H2S/CH4 ratio in Fig. 8a. It should be noticed that the H2/CH4 ratio for fumarole C is significantly lower than that for fumarole N and S. Methane is a component generated in the deep part of hydrothermal system (Giggenbach 1987). The fumarole C is the newly created crater in 2015. The fumarolic gas of C represents the fluid which is transported to surface bypassing the shallow part of hydrothermal system. Therefore, the abundant H2 contained in N and S fumarolic gases would be generated in shallow part of hydrothermal system.

The stable isotopic ratio of H2O in fumarolic gas enables us to estimate the fluid evolution. As shown in Fig. 9, the data points are distributed linearly on δD versus δ18O plane. In order to explain the distribution, a model calculation was carried out. The method of calculation has been demonstrated by Ohba et al. (2011). The parameters used in the calculation are summarized in Table 4. For the isotopic ratio of H2O of magmatic gas, the typical values for andesitic volcano on subduction zone (Giggenbach 1992) are assigned. In the model calculation, a high enthalpy magmatic gas (M in Fig. 9) mixed with a low enthalpy meteoric water (L in Fig. 9). As the result of mixing, a vapor phase (V in Fig. 9) and thermal water phase were generated. Their temperature was set to be 265 °C, which is the average apparent equilibrium temperature assuming the equilibrium of,
$${\text{SO}}_{2} + 3{\text{H}}_{2} = {\text{H}}_{2} {\text{S}} + 2{\text{H}}_{2} {\text{O}}$$
(8)
for the chemical composition of gas from borehole of Ow geothermal area (Ohba et al. 2011). During the mixing of M and L, the conservation of enthalpy, isotopic ratio and CO2/H2O ratio were assumed. After the mixing, vapor–liquid separation occurs. The conservation of enthalpy, isotopic ratio and CO2/H2O ratio was assumed also during the separation step, where the isotopic fractionation and CO2/H2O distribution took place between vapor phase (V) and liquid phase. The vapor phase (V) could suffer a cooling and a removal of condensed water. During the condensation, the isotopic composition of vapor moves along the line A. The temperature was set to be 100 °C during the condensation. The observed decrease in the isotopic ratio of H2O can be attributed to a condensation of H2O vapor. The correctness of the above model is supported by the data points for fumarole C, which were plotted near the cross-point between the line V and the line M–L. In the model calculation, at the cross-point, the amount of thermal water is zero. The fumarolic gas of C contains HCl (Table 3), suggesting the amount of liquid phase coexisting with fumarolic gas C was limited. The fumarolic gas C is interpreted to be the gas phase of direct mixing between magmatic gas (M) and meteoric water (L).
Fig. 9
Fig. 9

Isotopic ratio of H2O in fumarolic gases compared with model calculation. M: isotopic ratio of magmatic H2O (900 °C), L: isotopic ratio of local meteoric water (15 °C), V: isotopic ratio of H2O vapor (265 °C), A: condensation of H2O vapor at 100 °C

Table 4

Parameters used in the model calculation of vapor phase in terms of isotopic ratio of H2O and CO2/H2O ratio

Parameter

Value

Unit

Temperature of magmatic gas

900

°C

Temperature of local meteoric water

15

°C

Temperature of vapor and thermal water

265

°C

Temperature of vapor condensation

100

°C

δD of H2O in magmatic gas

− 10

‰ to SMOW

δ18O of H2O in magmatic gas

8.0

‰ to SMOW

δD of H2O in local meteoric water

− 51

‰ to SMOW

δ18O of H2O in local meteoric water

− 8.5

‰ to SMOW

CO2/H2O ratio of magmatic gas (high)

0.013

Molar ratio

CO2/H2O ratio of magmatic gas (low)

0.0045

Molar ratio

In the model calculation, CO2/H2O ratio of fumarolic gas can be correlated with isotopic ratio of H2O. In Fig. 10a, b, the result of model calculation is compared with observed fumarolic gas composition. In the diagrams, CO2/H2O ratio is converted to the scaled CO2 fraction (ScF) as
$${\text{ScF}}\left( {{\text{CO}}_{2} } \right) = \frac{{100C_{\text{CO2}} }}{{C_{\text{H2O}} + 100C_{\text{CO2}} }},$$
(9)
where CCO2 and CH2O are the concentration of CO2 and H2O in gases, respectively. The difference between Fig. 10a and b is the CO2/H2O ratio of magmatic gas (M) and the period of fumarolic gas. In Fig. 10a, CO2/H2O ratio of M is 0.0045 and the data of fumarolic gases in March and April 2015 are plotted when the CO2/H2O ratio was low (Fig. 2a). In Fig. 10b, CO2/H2O ratio of M is 0.013. The fumarolic gas N in June and July 2015 and S in May and June 2015 are plotted. In the above period, the CO2/H2O ratios of N and S were maximized (Fig. 2a). The ScF(CO2) of fumarole N in Fig. 10a, b is separated from the line A. The separation could be explained by the addition of vapor phase originating in local meteoric water (L). The composition of the vapor phase is indicated on the bold line on the y-axis. Figure 10a, b suggests that the CO2/H2O ratio of magmatic gas was high as 0.013 when the number of earthquake increased in May 2015, and low as 0.0045 before the start of earthquake swam (Fig. 2a).
Fig. 10
Fig. 10

Correlation between δ18O of H2O and CO2/H2O ratio. The bold line on y-axis indicates the vapor phase generated from local meteoric water (L) at 100 °C. M: magmatic H2O with CO2 (900 °C), L: local meteoric water (15 °C), V: H2O vapor with CO2 (265 °C), A: condensation of H2O vapor with CO2 at 100 °C. In the figures (a) and (b), the CO2/H2O ratio of M was set to be 0.0045 and 0.013, respectively

The above variation in CO2/H2O ratio is consistent with the magma sealing model (Fournier 1999). Accepting the estimation by Yukutake et al. (2015), the depth of degassing magma and the pressure were assumed to be − 10 km and 2256 bar, respectively. A degassing magma should be saturated in term of H2O and CO2. Therefore, the H2O and CO2 concentration in magma is located on the isobaric line G0 of Fig. 11. The isobaric line was drawn by use of the PC software VolatileCalc (Newman and Lowenstern 2002). For the calculation of isobaric line, the temperature and composition of magma were assumed to be 900 °C and rhyolitic (Tsujihara et al. 2017), respectively. On the line G0, the CO2/H2O ratio of gas phase can be calculated by VolatileCalc at any point, assuming equilibrium distribution of H2O and CO2 between magma and gas phase. VolatileCalc gave 0.0045 for the CO2/H2O ratio of gas phase at the point C0. The value 0.0045 is the CO2/H2O ratio of magmatic gas (M) when the number of earthquake was low. The H2O and CO2 concentration at the point C0 was 6.235 wt% (a0) and 6.75 wt ppm (b0), respectively. The above concentration represents the portion of magma equilibrated with gas phase, namely the inner surface of bubbles generated in degassing magma. We assign a1 and b1, respectively, to the bulk concentration of H2O and CO2 in magma. The point C1 defined by a1 and b1 should be located on the line with 0.0045 of slope starting from the point C0 (Fig. 11). Fournier (1999) proposed the magma sealing model explaining the episodic volcanic activity without magma intrusion. If the sealing zone restricts the emission of degassed gas, magma is compressed by the gas itself. When the sealing zone is broken, the gas stored within the sealing zone is supplied to hydrothermal system producing earthquakes and activation of fumarolic activity or phreatic eruption. Therefore, the CO2/H2O ratio of magmatic gas, 0.013, during the earthquake swarm can be attributed to the gas phase when magma was compressed by gas itself. Assuming that CO2/H2O ratio is 0.013 in the gas equilibrium, the magma at the point C1, a1 and b1 was calculated to be 6.33 wt% and 20.04 wt ppm, respectively. The isobaric line for 2325 bar is drawn on the point C1.
Fig. 11
Fig. 11

H2O and CO2 concentration dissolved in rhyolitic magma at 900 °C with isobaric vapor lines. G0 and G1: isobaric lines at 2256 and 2325 bar, respectively

The sequence of events over the earthquake swarm in May 2015 is summarized schematically in Fig. 12. In 2014, a steady degassing of magma and the transport of magmatic gas to hydrothermal system were kept (Fig. 12a). In this period, the magmatic gas pressure PF equals to the lithostatic pressure PL. The PF equals to PC0 which is the gas pressure equilibrated with the inner surface of bubbles in magma. In February 2015 until April 2015, the magma sealing took place (Fig. 12b), restricting the emission of magmatic gas resulting in the pressurization of magma by gas itself. At the end of the period, PF was elevated up to PC1. During the magma sealing, the supply of magmatic gas to hydrothermal system was restricted. The fluid pressure in hydrothermal system may be decreased, and enthalpy of hydrothermal system may be also decreased. The pressure decrease could cause an invasion of air, which elevated Ar/CO2 and N2/He ratios prior to the earthquake swarm (Figs. 3c, d, 12b). The enthalpy decrease could prompt the condensation of H2O vapor, which reduced the isotopic ratio of H2O (Fig. 4a, b). In May 2015, the sealing zone was broken (Fig. 12c). The magmatic gas with high CO2/H2O ratio was supplied to hydrothermal system as a surge. The supply of magmatic gas lifted the fluid pressure and produced large number of earthquake and a small phreatic eruption in June 2015.
Fig. 12
Fig. 12

Sequence at Hakone volcano before and after the earthquake swarm in May 2015. a Normal situation represented by the period in 2014. M: degassing magma, R: magmatic gas reservoir, S: sealing zone, B: hydrothermal system made of brittle crust, G: fumarolic gas, PL: lithostatic pressure, PF: magmatic gas pressure, PC0 and PC1: gas pressure equilibrated with magma the composition of which is CO and C1 in Fig. 11. b Sealing zone is restricting the magmatic gas transport to hydrothermal system, corresponding to the period in few months prior to May 2015. Air invasion was allowed into hydrothermal system. c Break of sealing zone in May 2015. A large number of earthquakes (cross-marks) happened in hydrothermal system

Based on the structure estimated by Yukutake et al. (2015) and the discussions in this study, the magmatic hydrothermal system of Hakone volcano is illustrated in Fig. 13a. The magmatic gas is supplied by the degassing magma (M). The gas contains H2O, CO2, SO2, HCl, N2, H2 and He. A magmatic gas reservoir (R) is developed above the degassing magma (Yukutake et al. 2015), and it was surrounded by a sealing zone (S). According to Itadera et al. (2016), a pressure source (star mark in Fig. 13a) is located at − 8.4 km below the sea level and beneath Ka, which produced the inflation of volcanic body in May 2015. The pressure source is located within the magmatic gas reservoir (R). Through the sealing zone, magmatic gas is transported to the hydrothermal system made of brittle crust (B). Most of SO2 and a part of CO2 in the magmatic gas are converted to H2S and CH4 in the hydrothermal system with the reaction of Fe2+ contained in crustal rock within the crust shallower than − 500 m (Fig. 13b). The magmatic gas was discharged as fumarolic gas N and S with the entrainment of hydrothermal components such as H2S and CH4.
Fig. 13
Fig. 13

South to North cross section of Hokone volano. a Star mark: pressure source appeared in 2015 (Itadera et al. 2016). M: degassing magma, R: magmatic gas reservoir, S: sealing zone, B: hydrothermal system made of brittle crust, L: local meteoric water, G: fumarolic gas, Ko: Komagatake central cone, Ka: Kamiyama central cone, Ki: Mt. Kintoki, Cross-marks: occurrence of earthquakes. The rectangle of broken line indicates the region drawn in b. b Hydrothermal system beneath the fumarole of N, S, C and borehole (Bh). The shallow part of the hydrothermal system is reduced (Red), whereas the relatively deep part is oxidized (Oxi). The boundary between the region of Red and Oxi has not been certainly determined

Conclusions

Sharp increases in the components ratios such as CO2/H2O, CO2/H2S, CO2/CH4 and He/CH4 were observed at the fumarolic gas on Hakone volcano synchronized with the earthquake swarm in 2015. The increases could be explained by the dominance of magmatic component relative to hydrothermal component. The CO2/H2O ratio of magmatic gas was 0.0045 before the earthquake swarm. It increased to 0.013 during the earthquake swarm. The increase in CO2/H2O ratio can be explained by the pressurization of magma caused by magma sealing. The magma pressure increased from 2256 to 2325 bar, consistent with the increase in the CO2/H2O ratio. The bulk H2O and CO2 concentration in magma were estimated to be 6.33 wt% and 20.0 wt ppm, respectively assuming the temperature of magma is 900 °C and the composition of magma is rhyolitic. Prior to the earthquake swarm in May 2015, sharp increase in Ar/CO2 and N2/He ratios was observed at the fumarolic gas in Ow geothermal area. The increase could be explained by the invasion of air in hydrothermal system. A definite decrease in the isotopic ratio of H2O in the fumarolic gas in Ow geothermal area was observed prior to the earthquake swarm in May 2015. The decrease could be explained by the partial condensation of H2O vapor probably due to the enthalpy loss in hydrothermal system beneath Ow geothermal area. It should be noticed that the above changes prior to the earthquake swarm were not observed in the fumarolic gas S in Ky geothermal area.

Abbreviations

AET: 

apparent equilibrium temperature

ASW: 

atmospheric components saturated in water

GC: 

gas chromatograph

HSRI: 

Hot Springs Research Institute of Kanagawa Prefecture

JMA: 

Japan Meteorological Agency

Ka: 

Kamiyama

Ko: 

Komagatake

Ky: 

Kamiyuba

Ow: 

Owakudani

R-gas: 

residual gas

SMOW: 

Standard Mean Ocean Water

So: 

Sounzan

wt: 

weight

Declarations

Authors’ contributions

TO drafted the manuscript. MY, KN, NN and YD sampled fumarolic gases and analyzed them. CS, MI and UT analyzed fumarolic gas samples. All authors read and approved the final manuscript.

Acknowledgements

This work was supported by the Japanese Ministry of Education, Culture, Sorts, Science and Technology, under grant of the Integrated Program for Next Generation Volcano Research and Human Resource Development in 2016 to 2017, also under its Earthquake and Volcano Hazards Observation and Research Program in 2015 to 2017. This work was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Numbers 15K12485 in 2015 to 2017. This work was supported by the Earthquake Research Institute, The University of Tokyo Joint Usage/Research Program in 2015 to 2017. This work was supported by General Research Institute of Tokai University in 2016 to 2017. We sincerely thank the above fund providers for the support in this work. We also thank Nagoya group: Kosuke Ikeya and Fumiko Nakagawa for δD(H2) measurement.

Competing interests

The authors declare that they have no competing interests.

Availability of data and materials

The csv files for Tables 1, 2 and 3 will be stored in repositories.

Consent for publication

Not applicable.

Endnotes

None.

Ethics approval and consent to participate

Not applicable.

Funding

This work was supported by the Japanese Ministry of Education, Culture, Sorts, Science and Technology under grant of the Integrated Program for Next Generation Volcano Research and Human Resource Development in 2016 to 2017. This work was supported by Japan Society for the Promotion of Science (JSPS) KAKENHI Grant Numbers 15K12485 in 2015 to 2017. This study was supported by the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, under its Earthquake and Volcano Hazards Observation and Research Program in 2015 to 2017. This study was supported by the Earthquake Research Institute, The University of Tokyo Joint Usage/Research Program in 2015 to 2017. This work was supported by General Research Institute of Tokai University in 2016 to 2017.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Authors’ Affiliations

(1)
Tokai University, 4-1-1 Kitakaneme, Hiratsuka Kanagawa, 259-1291, Japan
(2)
Meteorological Research Institute, 1-1 Nagamine, Tsukuba Ibaraki, 305-0052, Japan
(3)
Kanagawa Environmental Research Center, Shinomiya 1-3-39, Hiratsuka Kanagawa, 254-0014, Japan
(4)
Graduate School of Environmental Studies, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8601, Japan
(5)
Present address: Center for Marine Research and Operation, Tokyo University of Marine Science and Technology, 4-5-7 Konan, Minato-ku, Tokyo 108-8477, Japan

References

  1. Daita Y (2013) Temporal variation of the fumarolic gas composition during the swarm earthquake activity in Hakone volcano in 2013. Bull Hot Spring Res Inst Kanagawa Prefect 45:29–34 (in Japanese) Google Scholar
  2. Fournier RO (1999) Hydrothermal processes related to movement of fluid from plastic into brittle rock in the magmatic-epithermal environment. Econ Geol 94:1193–1212View ArticleGoogle Scholar
  3. Giggenbach WF (1975) A simple method for the collection and analysis of volcanic gas samples. Bull Volcanol 39:132–145View ArticleGoogle Scholar
  4. Giggenbach WF (1987) Redox processes governing the chemistry of fumarolic gas discharges from White Island, New Zealand. Appl Geochem 2:143–161View ArticleGoogle Scholar
  5. Giggenbach WF (1992) Isotopic shifts in waters from geothermal and volcanic systems along convergent plate boundaries and their origin. Earth Planet Sci Lett 113:495–510View ArticleGoogle Scholar
  6. Giggenbach WF (1997) The origin and evolution of fluids in magmatic-hydrothermal systems. In: Barnes HL (ed) Geochemistry of hydrothermal ore deposits, 3rd edn. Wiley, HobokenGoogle Scholar
  7. Itadera K, Harada M, Yoshida A (2016) Relationship between ground tilts and earthquake swarms during the 2015 Hakone volcanic activity. Bull Hot Spring Res Inst Kanagawa Prefect 48:1–10 (in Japanese) Google Scholar
  8. Kita I, Nitta K, Nagao K, Taguchi S, Koga A (1993) Difference in N/Ar ratio of magmatic gases from northeast and southwest Japan: new evidence for different states of plate subduction. Geology 21:391–394View ArticleGoogle Scholar
  9. Kobayashi M, Okuno M, Nakamura T (1997) 14C ages of pyroclastic-flow deposits from central cones on the western slope of old Somma of Hakone volcano, central Japan. Bull Volcanol Soc Japan 42:355–358 (in Japanese) Google Scholar
  10. Kuno H (1950) Geology of Hakone volcano and adjacent areas. Part-I. J Fac Sci Univ Tokyo Sect II 7:257–279Google Scholar
  11. Lee S, Ohba T, Yun SH, Yang K, Jeong HY (2016) Evaluation of sampling methods for sulfur speciation in volcanic gases. Chem Geol 438:123–133View ArticleGoogle Scholar
  12. Mannen K (2008) Hakone Caldera: mode of formation, and role in present-day volcanism. Res Rep Kanagawa Prefect Mus Nat Hist 13:61–76 (in Japanese with English abstract) Google Scholar
  13. Mannen K, Yukutake Y, Kikugawa G, Harada M, Itadera K, Takenaka J (2018) Chronology of the 2015 eruption of Hakone volcano, Japan: geological background, mechanism of volcanic unrest and disaster mitigation measures during the crisis. Earth Planets Space 70:68. https://doi.org/10.1186/s40623-018-0844-2 View ArticleGoogle Scholar
  14. Newman S, Lowenstern JB (2002) VolatileCalc: a silicate melt–H2O–CO2 solution model written in Visual Basic for excel. Comput Geosci 28:597–604View ArticleGoogle Scholar
  15. Ohba T, Daita Y, Sawa T, Taira N, Kakuage Y (2011) Coseismic changes in the chemical composition of volcanic gases from the Owakudani geothermal area on Hakone volcano, Japan. Bull Volcanol 73:457–469. https://doi.org/10.1007/s00445-010-0445-9 View ArticleGoogle Scholar
  16. Ossaka J, Ozawa T, Nomura T, Ossaka T, Hirabayashi J, Takaesu A, Hayashi T (1980) Variation of chemical compositions in volcanic gases and waters at Kusatsu-Shirane volcano and its activity in 1976. Bull Volcanol 43:207–216View ArticleGoogle Scholar
  17. Ozawa T (1968) Chemical analysis of volcanic gases: I. Chemical analysis of volcanic gases containing water vapor, hydrogen chloride, sulfur dioxide, hydrogen sulfide, carbon dioxide, etc. Geochem Int 5:939–947Google Scholar
  18. Richet P, Bottinga Y, Javoy M (1977) A review of hydrogen, carbon, nitrogen, oxygen, sulphur, and chlorine stable isotope fractionation among gaseous molecules. Ann Rev Earth Planet Sci 5:65–110View ArticleGoogle Scholar
  19. Taran YA, Giggenbach WF (2003) Geochemistry of light hydrocarbons in subduction-related volcanic and hydrothermal fluids. Soc Econ Geol Special Publ 10(2003):61–74Google Scholar
  20. Tsujihara R, Ishibashi H, Hokanishi N, Yasuda A (2017) Pre-eruptive process of the 60Ka caldera-forming eruption at Hakone volcano. In: Program and abstracts the volcanological Society of Japan, P005. https://doi.org/10.18940/vsj.2017.0_125 (in Japanese)
  21. Tsunogai U, Kamimura K, Anzai S, Nakagawa F, Komatsu D (2011) Hydrogen isotopes in volcanic plumes: tracers for remote temperature sensing of fumaroles. Geochim Cosmochim Acta 75:4531–4546View ArticleGoogle Scholar
  22. Yukutake Y, Honda R, Harada M, Arai R, Matsubara M (2015) A magma-hydrothermal system beneath Hakone volcano, central Japan, revealed by highly resolved velocity structures. J Geophys Res Solid Earth 120:3293–3308. https://doi.org/10.1002/2014JB011856 View ArticleGoogle Scholar

Copyright

© The Author(s) 2019

Advertisement