Geochemical Journal Vol. 19, pp. 27 to 44, 1985

Origin of thermal waters from the Hakone geothermal system, Japan

SADAO MATSUOI, MINORU KUSAKABE2, MARIKO NIWAN03, TOMIO HIRANO4 and YASUE OKI4

Department of Chemistry, Tokyo Institute of Technology, O-okayama, Meguro-ku, Tokyo 1521, Institute for Study of the Earth's Interior, Okayama University, Misasa, Tottori 682-022, Department of Chemistry, Tokyo Kyoiku University, Otsuka, Bunkyo-ku, Tokyo 1123, and Hot Springs Research Institute of Kanagawa Pref., Hakone-Yumoto 997, Kanagawa 250-034, Japan

(Received June 18, 1984 : Accepted August 1, 1984)

In the Hakone geothermal system a variety of waters including precipitation, surface water, ground water, thermal water and steam condensate were analyzed for 5 D and S 180, dissolved sulfate and carbonate for S34S, 5180 and 513C, and some sulfur-bearing gases for S34S. Fifteen samples were collected monthly to see if there is any monthly change in the isotopic composition of water. Except in precipitation and steam condensates, no significant monthly changes were observed. 8180SO4 and 534SS04 analyses indicate that surface oxidation of volcanic sulfur produces isotopically light sulfate in water occurring at relatively high elevations. Sulfate minerals in the basement rocks formed by the Miocene submarine volcanism are another source of dissolved sulfate in waters at lower elevations. C02 originally derived from decomposition of marine carbonate is suggested as a carbon source for dis solved bicarbonate at higher elevations, although contribution of organic carbon becomes significant in waters at lower elevations. In a 5D versus 8180 plot, surface waters including precipitation and ground water lie on the line, 6D = 88180 + 17. On the other hand, thermal waters lie on a regression line, SD=2.16180-33.5. Around the intersect of the two lines, S D = -51 %o and 8 180 = -8.5 %o, there is a swarm of point for groundwaters. We call the ground water with 8 D = -51 %o and 8180 = -8.5 %o "representative" groundwater (RGW). From both chemical and isotopic view points, thermal waters are interpreted to be a mixture of RGW and high temperature dense steam (HTDS), the latter being ultimately evolved from hydrothermal interaction of RGW with rocks containing appreciable amounts of hydrous silicates. Values of 5D and 8180 for HTDS have been interpreted to be a result of rock-RGW interaction with the rock/water weight ratio of about 10, on the basis of isotopic material balance in a closed system with the use of SD of hydrous silicates in rocks as well as 8D and 8180 values of RGW. When the ratio of about 10 is compared with those of other geothermal areas, the Hakone geothermal system is rock-dominated. Interaction between meteoric water and rocks including hydrous silicates under a rock-dominated con dition can account for both hydrogen and oxygen isotopic shifts found in thermal waters.

tion through water-rock interactions, and c) INTRODUCTION subsurface boiling. In most cases the above Isotopic study, especially that of hydrogen factors contribute to different extents. and oxygen isotopes of water is now an essen The Hakone geothermal system may be one tial approach to the origin and behavior of of the best fields to study the hydrological fea geothermal waters, because geothermal waters ture of various types of thermal waters, since are characterized by their isotopic and chemical the caldera can be regarded as an independent compositions. Hydrogen and oxygen isotopic hydrological unit. The Hakone volcano, situated ratios of geothermal waters are controlled by 80km south-west of Tokyo, is a strato-volcano a) mixing of meteoric water with various types with double caldera and is composed of three of water such as sea water, connate water, meta geological units, i.e., old somma, young somma morphic water, and magmatic water, b) altera and central cones (Kuno, 1950). The volcanic

27 28 S. Matsuo et al.

activity started in the late Pleistocene and still 1974) extended the scope and discussed the continues in the form of fumarolic activity in genetic relationship of thermal waters on the the Kamiyama and Komagatake (the central basis of extensive chemical analyses. cones) with occasional volcanic earthquake Since thermal waters are mostly of meteoric swarms. Many hot springs are found along the origin (Craig et al., 1956), a better understand deep valleys dissected by the rivers Hayakawa ing of the isotopic features of the local meteoric and Sukumogawa. The Hakone caldera with water helps interpret the isotopic characteristics dimensions of l 1 km (north-south) and 7 km of local thermal waters. In this respect, Matsuo (east-west) is morphologically separated from et al. (1979) estimated in the Hakone caldera the surrounding area by a distinct old comma quantitatively the hydrological water budget, (Fig. 1). using hydrogen and oxygen isotopic composi A thorough study of the geology of the tion and chloride content of rain-, river-, lake Hakone volcano and adjacent areas has been and ground-waters of the area. made by Kuno (1950, 1951). Yuhara et al. In this paper we discuss the ultimate origin (1966) and Yuhara (1968) measured the amount of thermal waters in the caldera on the basis of of steam and heat discharge from the active isotopic information referring to the previous fumaroles, which enabled estimation of the chemical and hydrological studies. energy output in the Hakone geothermal system. There are a number of chemical studies on SAMPLES AND ANALYTICAL METHODS thermal waters of this system. Among these, Sato (1962) emphasized a zonal distribution of About seventy samples were collected from the thermal waters. Oki and Hirano (1970, the Hakone caldera for isotopic analyses as

loon

n 00 ~ 00 ~ T 1000 okyo • 800 Q~ t Hakone 800 800 l Pacific Ocean 600 I 0 Zone 11 Zon`e~ • t ED6) 'r w p'' . %Hayakawa. 1 ~a ` Zone I Zone III •j KAMIYAMA ...... • >1~ X 1200 Zone Nb F 600 r KOMAGAKAKEAt ogawa ~ t' 400 Suku Zone i I t,, 1 Zone II ~` t 000 t f& 0 Zone III 800 op . A ` t 00 O Zone Na I 4 _ ^ \_ •` O Lake ^ C b Zone N `~ V A v/ A Q Steam Ashinoko w 723 Surface water O / 400 Ground water V ~• j L800 AI t Rain 600 -l

Caldera rim s km Fig. 1. Sample localities in the Hakone caldera. Heavy lines indicate the zonal distribution of thermal waters with various chemical compositions (after Oki and Hirano, 1970).

~ Origin of thermal waters 29 described in Table 1. They are precipitation, was combined with the water evolved as water river and lake water, ground water, thermal vapor. The water thus collected was analyzed water from the four zones in the caldera which for D/H ratio in the manner described for water were proposed by Oki and Hirano (1970) and samples. 180/160 ratio of water samples was steam condensates from steam wells and fuma measured by the established C02-H20 exchange roles. Fifteen samples with asterisks in Table 1 technique, followed by the mass spectrometric were collected once a month from May 1971 to measurement. Both isotopic ratios are presented May 1972 in order to see if there were any in the 6-notation with SMOW as the standard. seasonal variations. These samples were chosen The overall error was ±1.3%o (one sigma) for so that they cover the various types of waters 6D and ±0.2%o (one sigma) for 6180. occurring in the caldera. For carbon and sulfur isotopic analyses of Figure 1 shows sampling localities in the the dissolved materials, 1 to 2 liters of water Hakone caldera with the zonal distribution of samples were taken into plastic bottles to which thermal waters. In Table 1 are included isotopic a small amount of mercuric chloride had been analyses, water temperature, pH values, chloride added to suppress any biological activity after contents, sampling localities with altitude above collection. Dissolved sulfate was precipitated as sea level, and the registered numbers for ther BaSO4 for which S 18OSO4 and 634SSO4 analyses mal waters (numbered by Kanagawa Prefecture). were made on C02 and S02 gases respectively, It should be noted that the in situ pH value is obtained by using the preparation techniques certainly not retained, since most of the thermal described by Rafter (1957), Mizutani (1971) waters are discharged by an air-lift pump. Con and Robinson and Kusakabe (1975). Sulfur densates from steam wells and fumaroles were bearing gases from steam wells were recovered collected by sucking steam through a cold trap in the following manner; steam was introduced with the help of a manually operated pump. into a water-cooled alkaline solution containing Some rock samples from the Yugashima group, cadmium acetate, and the cadmium sulfide thus the basement rocks consisting of thick piles of formed and other sulfur species in higher oxida submarine pyroclastic sediments of the Miocene tion state remaining in the solution were re age, were analyzed for D/H ratio in order to covered for sulfur isotopic analyses. estimate the extent of interaction between the The total dissolved carbonate was extracted thermal water and rocks. from the samples by acidifying a solution in D/H ratio analyses of water samples were vacuum with concentrated phosphoric acid carried out by passing 5 to 10mg of water over (Deuser and Hunt, 1969), and 613C values of uranium metal heated to 700'C (Bigeleisen et the resulting C02 were determined mass spectro al., 1952) and comparing D/H ratio of the result metrically. ing hydrogen with that of the standard hydrogen on a mass spectrometer with a dual-inlet and GENERAL FEATURES AND MONTHLY VARIATION dual-collector system. Hydrogen extraction OF IsoTonc RATios from hydrous silicates in rock samples was made in the following way. Each sample was pow The results of hydrogen and oxygen isotopic dered, dried and then loaded in a platinum analyses of the water samples are presented in crucible, and hydrogen in the form of water Fig. 2. In Fig. 2 we draw a straight line re was extracted by means of induction heating in presenting the relationship between 6D and 6180 vacuum at 1,200'C after absorbed and interlayer of meteoric water in Japan proposed by Sakai water had been removed at 200°C (Savin and and Matsubaya (1974), i.e., SD = 86180 + 17. Epstein, 1970). A small amount of H2 gas It is seen in Fig. 2 that the average rain and most formed during heating was converted to H20 of ground water samples in the Hakone caldera through a CuO furnace and the H20 recovered fit quite well the straight line. Likewise, a 30 S. Matsuo et al.

b a O U .0 0 H U •t:• d . 1 w I w I w I 1 A I I I A A I I I I I I w A as PQPQ PQ U

U

00 0) .-1 3 a 3 O0 O.t 0) 0 O 0 0 z I 0, C) g4 4) (D A 3 0 0 N 0 z zN~ 3 N 3 10 aE O G7 v) w O y ~b33 N 0 ,~ z o tn CD 3 0 0 0 ~bzz a 0 t~l O tN a0ia~i ~ 0 ~ '~ z ..a) 0 >0 C4 b 4-13 3 O ai 0, 0, 0 z (D Z x 0 ~x0~, ~ ~ .co 2 y .: o .)

w 4 °' o O 0 N O 0 0 0 1010 0 c'yd IF :2P 0 to Q0 a0) a 18 Cn to 00xXvW~0 x h E400>+~4 .v

00 o t- -v 00 CD q* qt e v .-, O ,~ .-4 0 •--4 0 • -1 -( .-.4 M --~ -~ O a qt tn * tnv)C"4NNNt- v) It) 0 * W) tn * N * Its .d a ON-i 0 0 0 0 0 •-~ * O O O N N 0 I-q 1-4 r. 10 CA N N NNNNNNNNN'.O N N r N y

l~ 0 00 00 I I 09 O\ 00 .--~ N 00 N M x I 00 N I I I I 00 N I N M N N N 00

0 t7 N

f1 /1 C~ llO M .-~ t~ N ~t M ti N I-D ~--~ O O t-d' I d' t t` -4 I I I MM'.O tn•-+ ON 0) q*r-o,Ntn to N N I .--1 N N .--IN It to N V) I!) %D m 0) 0) O,

CDo O to I I I I I I I I voio° I X000 I 1 00 C) 00 N I I o 4%c -I 3b~ N to N d' N W)

O O O 0 M O 0 O 0tn000I)tn000 W) 00 W) O W) O O O O W) O N v) to v) O N v) to M M 00 v) M 0 M qtt v) 00 h 0 W) 0) M N 0) 0 N WhCD N N0000 N [~ N N N N lp 00 N tN 00 0) ) 00 00 N 00 W) to to '.o

O O N M ~0 Nd'v)M.•iMN O M M o00 A 0) I 00 00 00 '.0 00 I'D to 00 0) %0 00 0O 00 00 00 00 00 t 00 ~..~ t; 00 00 t-~ N h I I I I I I I I I I I i I I I I I I I '--' I I I I I 0) C a 1 1 1 1 . O C'r O N N O N C O ~+ F ~+ F U O In 10 •--1O N 00 00 00 (71 .--~ O, t q* O O 0v)N0,0,w) 4 OM.-4 o A O&a 01 N '.0 .--i y) '.O D\ --i d 0; O O --~ 0) 00 N 4' C 06 -r 00 N oo n O 1 ~O "D M to d' to to 't tt•) M M It v) v ~+to 'ct' et to "t ~t V) ELw I I I I I I I l ~) It) W)V v v) I I I I I I 1 0)) too I I I I Uw a () 3 3 i td 10 00 end C •w dF 6. O 0 U * dF a N M c"M; .~ N M 'T to v~ t~ op T * Cy .g a + ~F.-+ N M to . --I N M 10 0~ a > X3333 aa~ x rr~ Nr'.i x ~--t~-r r~ rr ti~-r ~-.~-r .-arr

b WOy4 Origin of thermal waters 31

04a U a a N ,•-i N d U 00 N dVUUQ pgAq I I I M to WWWWLitWW t,7 I I I I I U 0 U U 0 O N N Ux N x

qrn ~ 3 to O M M to o ~

00 ,00 00 M xo-4 O 0 00 >+ 00 -4 , o O 0 a% c Z 0 N ~Z W ?E10 0 Z C; N r z O >. : •N C O^v\ 0 :3 N 0M WC) C)M Z ZC7C14 is V] >1:9 0 C) H W0 •-' 0 I0 10a.4sa, 504 Ad RI O O Q vq >1 O 0 r.' b0 0 o (D 0 O0 0 end y

N cdO 954 > +~-' t O , 0 s. o q. 1'0 0 .c 0 y ao a' En ~t 0014' ~,xW rAWIO d xd0>-+cn W O O D ci 1C/nx r~ o t3

v ,-~ ,•-t r-t o N N N N N, * kn tn h t/) II) 01 * tn h tn *I) try N N N N N n 0 0 0 0 0 0 I * 0 0 0 0 0

C O t3-`

00 lD ti N ID 00 N 0\ 00 . M O 00 "i I I I I I ts NO E 00 N 00 00 00 N N I I I i I I I ,y y o) ~ e~'ts at et to M 'n M to q O O~ M N Oo O~ t,p N O I N I I M %Dw-q It t to 1.0 le t0 to 'T 1 -4 '1) 1-4 M M o I O N I o 0 01 01 00 40 to 'I) 00 Q\ Q et t1 in InNIt%tn4D , ON 0, O

96.

0 -400 0 000 M Go 0 o to o %0 (D w I I 01% o I It to o 0 O is to can I v°) v to tN to v) N O I I I I I

0

0 00 to 0 0 0 0 0 0 coo Co0o I M I 1 01-4 tI)N t` 000\ 00N tn MtNN'1MO~ kn t I I I I I v v0 0) 0) p to 00 h tD 1I) try d' It N -N -NN.-I t, O\ O\ lb V ti ti I) ~ h y M 0\ 0\ to td tom.-4 In 4D0~0\0 - tr) M to M t.o O .-' 3 t` N tD o0 N "t Ln W) tn ~D Ci [i li ~D%0 %0 010 n l~ l~ ~ 00 t~ l~ l~ 1 1 1 1 1 1 + O ~oto v) 00 I) h 0 N

O N o0 O O~ tI) N r . -i N N -•i tD M to d O 00 0 00 in t to N M't 4140406 0,N0000000 4-t00 00 0~ M -4 ~6 6 0) O\ 0) M to N 't It. Id,~r 'd 'T ~tV M N N t ~t t a ? ? ' tj)d 't~tn I i I I 1 I I ~ b b I' to tI) 'd• to b 0) 0) 0) 8 G 00 d * b ~ ~ O 3rn .-•1N M tI) 00 0l~ rn .1 N M er•to ~ , %9 t'7 ooo, -0 -0 0 C40UUU000U 4

a0N O 32 S. Matsuo et al

straight line, 5D = 2.15180 -33.5, can be drawn various types exists. The results of hydrogen by the least squares best fit method for all the and oxygen isotopic analyses of monthly col thermal waters except steam condensates. The lected samples are given in Tables 2 and 3, two lines intersect at 5 D = -51%o and 5180 = respectively. The isotopic variation is the gratest -8 .5%, There is a swarm of points of ground for precipitation. The variation with month in water centered at this point which we call SD and 5180 of the other waters is almost equal "representative ground water" (RGW) hereafter . to or slightly greater than the accuracy of mea It is seen in Fig. 2 and Table 1 that 6D and 6'80 surement except for the two steam condensates. values of the surface water having open surface This means that the monthly fluctuation in such as lake water (SW-1) and pond waters (SW precipitation has been smoothed out under the 3, -4) are higher than those of most of the ground (Matsuo et al., 1979). The same conclu ground and thermal waters, and the isotopic sion has been obtained in the Nasu volcanic area, enrichment relative to RGW is most likely to Japan (Kusakabe et al., 1970). be due to kinetic evaporation. The discharge rate of I-1 hot spring, spouting If the line for thermal waters, 5 D = 2.15180 from the slope of a central cone, is known to be 33.5, represents a mixing line of RGW with positively correlated with precipitation inten another component, the latter, presumably a sity, and the water temperature and salt con high temperature water, should lie somewhere centration are inversely correlated (Hirano and along the thermal water line in Fig. 2. This Tajima, 1969; Hirota and Odaka, 1975). The idea will be pursued in the later section. isotopic ratios, however, show practically no One of the purposes of this study is to see monthly fluctuation, and the annual averages of whether or not any systematic change in iso 6D and 5180 values are only slightly higher than topic ratios of monthly collected water of those of the precipitation by about 10%o and

6180 -10 -8 -6 -4 -2 0 +2 +4 1 i i i i i -20 i i i

Zone I A Zone 11 i *- 0 Zone III i 0 Zone 240+C IVa n -30 c Zone IVb liquid 0 Surface water reservoir • Ground water

e Steam ° i ° 0 Average rain i i A -40 0 ° ° 0 0 6D =2.1 6180 -33.5 o ~ o 6 p (°/oo) '. o 0 oi° 0 RGW .~®]~c-o -50 -~ O

°

6D=86180 +17 -60

Fig. 2. 6D versus S 180 plot of the Hakone water samples. The isotopic compositions of thermal waters excluding steam condensates are best approximated by a least square regression equation, 6D = 2.15 180--33.5. The isotopic compositions of the representative ground water (RGW) are defined by the intersection of the thermal water line and the meteoric water line, SD = 85 180+ 17. A point at the far-right end of the thermal water line indicates the isotopic composition of the hot water reservoir occurring below the active fumarolic areas of the Hakone volcano. Origin of thermal waters 33

Table 2. Variation with time in hydrogen isotope ratio of monthly collected samples and precipitation depth

Sampling Amount P-1 SW-1 I-1 II-1 III-1 IVa-1 IVb-1 CW-1 date (mm) RW-1 CW-2 1971 May -48 .9 -42 .7 -50.7 -44.6 -44.2 -45.3 -48 .9 -47 .2 -24 .1 June -68 .0 220 -49 .3 -42 .4 -51.7 -46.5 -47 .9 -45 .7 -49 .1 -45 .8 -24 .6 July -71 .1 184 -49 .1 -41.7 -51 .1 -52 .3 -47 .1 -44 .6 -49 .1 -42 .1 -22.6 August -38.2 141 -49.4 -42 .4 -49 .0 -49 .7 -46.9 -46.2 -47 .7 -41 .7 -25 .0 Sept. -60 .4 645 -50 .2 -42 .8 -50 .8 -49.4 -46.5 -46 .6 -48 .8 -39 .0 -26 .7 Oct. -52 .9 309 -47 .2 -42 .3 -50 .0 -46 .7 -46 .2 -47 .7 -48 .6 -32 .5 -22.5 Nov. -79.6 289 -50 .4 -43 .9 -50 .0 -48.3 -46.5 -46.6 -47.6 -30 .5 Dec. -49 .0 29 -47 .9 -41 .3 -50.0 -47.1 -45 .3 -47 .3 -48 .4 -23 .6 -19 .3 1972 Jan. -82 .0 238 -51.5 -42.5 -51 .1 -50 .3 -46 .4 -46.8 -46.9 -45.3 -26.3 Feb. -62.0 152 -48 .5 -42 .4 -48 .6 -47.8 -47.7 -47 .9 -49 .3 -43 .7 -24 .0 March -67 .9 74 -48 .2 -44 .4 -47.9 -51 .6 -44 .0 -46 .4 -47 .5 -37 .1 -23.2 April -46 .1 262 -47.4 -42.4 -49 .4 -47 .6 -46 .8 -47 .1 -47 .6 -40.0 -25.0 May -46.5 >115 average -61.7* -49.0 -42.6 -50.0 -48 .5 -46.3 -46.5 -48.3 -39.8 -24.5 standard _ _ 1.1 0.8 1.1 2.1 1.2 0.9 0.8 6.5 2.6 deviation

* weighted for precipitation amount .

Table 3. Variation with time in oxygen isotope ratio of monthly collected samples and chloride content of meteoric precipitation

Sampling C1 SW-1 I-1 II-1 III-1 IVa-1 date P-i (ppm) RW-1 IVb-1 CW-1 CW-2 1971 May -7 .9 -7 .4 -8 .3 -7 .7 -5 .5 -7 .0 -7 .6 -6 .0 +1.2 June -10 .5 -8 .4 -7 .3 -8 .0 -8 .1 -6 .0 -6 .8 -8 .0 -5 .0 +1.4 July 9.9 -7.7 -7 .4 -8 .4 -7 .8 -6 .2 -6.8 -7 .6 -4.1 +1.6 August 5.4 3.0 -7 .5 -6 .4 -8 .2 -7.9 -6.2 -6.7 -7.8 -4.2 +0.8 Sept. 9.4 -8 .2 -7 .4 -7 .9 -6 .9 -6.0 -6.4 -6.3 -4.0 +1.9 Oct. 8.7 2.3 -8 .2 -7.1 -7.5 -7 .6 -6 .2 -7 .1 -7 .5 -3 .9 +1.1 Nov. -12.2 1.6 -8.0 -6.6 -7.8 -7 .2 -5 .7 -6 .6 -7 .6 +1.1 Dec. 9.6 5.8 -8.3 -6 .7 -7 .9 -7 .1 -6 .2 -6 .4 -7 .6 -2 .4 +1.9 1972 Jan. -13 .4 1.4 -7 .5 -6 .2 -7 .8 -7.5 -6.1 -6.5 -6.8 -5.1 +1.5 Feb. -11 .5 1.3 -8 .7 -6 .9 -7.8 -7.7 -6.1 -6.7 -7.8 -5 .5 +1.4 March -11 .5 2.4 -7.7 -6.4 -7.1 -7.8 -5 .8 -6 .3 -7 .7 +2.3 April 8.2 2.5 -8 .0 -7 .0 -8 .4 -7 .7 -6.5 -6.8 -8.1 -5.4 +1.0 May 7.0

average 9.9* 2.1* -8.0 -6.9 -7.9 -7.6 -6 .0 -6 .7 -7 .5 -4 .6 +1.4 standard deviation 0.4 0.4 0.4 0.3 0.3 0.2 0.5 1.0 0.4

* weighted for precipitation amount .

1 %o, respectively. Considering a relatively large perature ranged from 93 to 123°C. The sample annual isotopic variation in precipitation (P-1 in consisted of steam and mist. Steam discharge Tables 2 and 3), we can conclude that I-1 is an was made each time only for sampling with essentially well-mixed ground water of shallow different extent of valve opening, which may depth discharged by the water-head built up by determine the outlet steam temperature. In precipitation. addition, there is no relationship between iso A large monthly variation has been found topic values of steam and outlet temperature for a steam condensate, CW-1. The steam tem of steam. This implies that the mist to steam 34 S. Matsuo et al.

ratio in each. sample is not related to the mea Table 4. Altitude effect on liD of precipitation ob sured outlet temperature. The observed month served at south-western slope of the Koma ly isotopic variation may be a reflection of gatake secondary factors during sampling and not be Altitude SD Precipitation related to the real variation in the original ther (m) (%0) amount (min) 730 -65 .9 405 mal water. 920 -68 .5 397 We see in this section that monthly change 1050 -72 .3 438 -72.2 in isotopic ratios of water in the caldera is 1190 496 1350 -70 .8 492 negligible except for precipitation and steam 780* -66 .3 478 condensates. It can be safely said that thermal * collected at Shimoyu Earthquake Observatory , Gora, and ground waters collected at an arbitrary Hakone. time of year have isotopic compositions close to the annual averages. In the following discus water are definitely higher than those of the sion, we will use one-shot samples together with average precipitation. This indicates that the the averages of monthly collected samples with annual average of isotopic composition of an equal weight. precipitation fluctuates with year and/or the average precipitation has undergone enrichment of heavy isotopes through evaporation and BEHAVIOR OF SURFACE AND GROUND WATER evapotranspiration during the formation of the IN THE CALDERA homogenized ground water (Matsuo et al., In order to reveal more detailed isotopic 1979). features of surface and ground water in the The annual averages of isotopic ratios of the caldera, we collected rain water samples from lake water are S D = -43 %o and 8'80 = -7 %o, September 9 to October 20, 1971 at five dif which are even much higher than those of RGW. ferent altitudes on the south-western slope of The possibility that RGW is a mixture of the Komagatake (one of the central cones) from lake water and average precipitation is unlikely 730 to 1,350m elevation at about 200m inter because ground waters with SD values almost vals. The result of D/H measurement is given in equal to that of RGW can be found even at Table 4. It is seen in Table 4 that SD decreases topographically higher levels than the lake. The about 1 %o per 100m at this particular site. permeable pumice formation extending towards This rate is about one fourth to one half of the the lake bottom is found at the uppermost rates in the Alps (Moser and Stichler, 1970) region of the Sukumogawa River. The water and the Sierra Nevada (Friedman and Smith, sample (RW-3) collected in this area has isotopic 1970, 1972), probably reflecting the smaller values not close to those of the lake water, size of the field investigated. So we do not take but to RGW. This indicates that no appreciable into consideration the change in isotopic ratios subsurface leakage of lake water occurs. There of precipitation with altitude and we will use is a possibility, however, that the bottom water the annual average values of 5D = -61.7%o and of the lake is isotopically different from the sur 5180 = _9.9%0 (Table 1) for isotopic ratios of face water which was actually sampled, and the the precipitation supplied to the caldera. bottom water is the major source of the ground The values of S D and 5"0 of ground water water in the caldera. In order to check this (almost the same as those of RGW) are distrib possibility, a SD profile at the deepest point uted in a narrow range and are practically of the lake was measured in June 1971 when the independent of their altitudal and lateral posi thermal stratification started. The water sam tion in the caldera. In other words, ground ples from 0, 10, 20, 30 and 38m (almost the water in this caldera is isotopically homogene bottom) showed no appreciable vertical change ous. However, S D and 8"0 values of ground in SD values which were -41.5, -41.5, -42.0 Origin of thermal waters 35

-44 .1 and -42.8%o respectively. It should be Fig. 2, Zone I and II waters can be regarded as concluded that no sizable contribution of the ground water in origin. Zone I water extracted lake water to ground water is expected, though S04 and Zone II water extracted HC03 from the lake water is the largest surface water re central cone materials without significant iso servoir in the caldera. topic exchange. The origin and behavior of S04 and HCO3 will be discussed in the next section RELATIONSHIP BETWEEN CHEMICAL ZONING Zone III waters are represented by high tem AND ISOTOPIC COMPOSITION OF HOT SPRING WATERS peratures (near boiling point) and intermediate pH values, major constituent being sodium As mentioned already, Oki and Hirano chloride. Three tongue-like branches of the zone (1970) divided the occurrence of thermal waters III waters are recognized in the eastern flank of into the four zones according to anion com the Kamiyama (Fig. 1). It is seen in Fig. 2 that position of the waters (see Figs. 1 and 3). SD values of the Zone III waters are slightly Zone I waters have low pH values (2-3) higher than those of waters from the other and high sulfate ion concentrations. They are zones, and S 180 values are definitely higher than located near the top level of the Kamiyama, one those of the others. of the central cones (see Fig. 1) and have SD and Oki and Hirano (1970) postulated the ex 5180 values as low as those of ground waters istence of "high temperature dense steam (Table 1). (HTDS)" (White, 1957; White et al., 1971) Zone II waters are characterized by high at depths of the Hakone geothermal system and pH values (^-8), and high bicarbonate and sulfate estimated its contribution to the zone III waters ion concentrations. Zone II waters are pumped to be 2530%, assuming that HTDS contains from depths of several hundred meters around only sodium chloride, and that sulfate ions of the central cones. According to Oki and Hirano the zone III waters are supplied exclusively by (1970), the waters may come from a major the zone II waters. They also estimated the aquifer beneath the central cones. As seen in contribution of HTDS to the zone III waters on the basis of the enthalpy budget to be 17 23 %. Cl Oki and Hirano (1974) later revised their value 66 1ONE III ob to 30-50% on the basis of chemistry. The Go validity of these figures will be discussed in a later section in view of isotopic data.

Q Zone IV waters are sodium chloride-bicar bonate-sulfate type or mixed type (Fig.3). Oki and Hirano (1970) further divided zone IV ZONE IVq waters occurring on the eastern side of the & IVb 0 caldera into two subgroups, IVa and IVb ; the former occurs at higher levels in the Quaternary volcanic rocks around the central cones and the latter in the Yugashima group, the Miocene ZONE II basement rocks at lower levels. 10 ZONE I ®® The chloride content is plotted against FC02 S04 5"0 for thermal water as shown in Fig. 4. It Fig. 3. Anion composition of the thermal waters studied is clear that zone III waters with a few excep with respect to the zonal distribution (Oki and Hirano, tions have the highest S 180 and chloride content, 1970). The symbols are the same as those in Figs. 1 and and that zone IV waters result from mixing of 2. zone II and III waters. The results of isotopic 36 S. Matsuo et alL

analyses agree quite well with the chemical 0 C C zoning of thermal waters in the Hakone geo thermal system. 0 0 2000 0 0 0

E ISOTOPIC EVIDENCE FOR THE a a O DISSOLVED MATERIALS U 0 C 0 The results of sulfur and oxygen isotopic 1000 00 analyses of sulfate ions dissolved in the various 0 types of thermal waters are given in Table 5, 0 0 together with some additional sulfur isotopic C C results for sulfur-bearing gases from the steam 0 0 0 wells. Figure 5 shows the isotopic relationship -9 -8 -7 -6 -5 -4 between oxygen and sulfur of dissolved sulfate. 6180 (°/ao) There is a positive correlation between S 18Oso4 Relationshi p Fig.4, betweenchloridecontent and and S34Sso4 values, with a trend that the sulfate 6180 value o the thermal waters. Symbols are f the in zone I and II waters is isotopically lighter same as those in Figs. 1 and 2.

Table 5. Sulfur, oxygen and carbon isotopic composition of dissolved sulfate and bicarbonate of thermal waters, and sulfur-bearing gases from steam wells in the Hakone geothermal system

S04 HCO1 Sample conc. conc. 634 s 8180 S13C (ppm) (CDT) (SMOW) (ppm) (PDB) Hot spring water from zone I I-1 405 -3 .2 +0.3 1-2 502 -3 .4 +0.1

Hot spring water from zone II II-1 551 -2.2 +3.2 121 +0.4 11-2 370 +14.1 +8.5 443 +2.0 11-5 202 -2 .5 +3.0 113 +2.1

Hot spring water from zone III 111-2 93 +5.8 +2.0 196 +1.0 111-3 107 +4.9 +1.8 61 -2 .3 Hot spring water from zone Na IVa-1 81 +4.2 +2.5 73 +0.6 Na-2 174 +6.7 +4.1 254 -0 .2 IVa-8 140 +5.9 +4.3 178 -1.5 Na-9 54 +4.1 +1.7 77 -6.7 Hot spring water from zone Nb Nb-1 34 +12.1 +5.9 51 -12.1 IVb-2 409 +17.5 +8.7 15 -11.2 IVb-4 127 +10.0 +6.6 25 -14.1 Groundwater GW-10 193 +7.2 +5.4 43 -6.4 Gases from steam well CW-1 H2S -5.3 S02* -3.2 total sulfur -3 .7 CO2 CW-2 H2S -4 .0 S02* +2.9 total sulfur 0.0 * sulfur compounds other than H 2S. ** taken from Craig et al. (1978). Origin of thermal waters 37

(with an exception for the sample 11-2), the (Oki and Hirano, 1970). The fact that the sulfate in zone IVb waters heavier and the 6'8OSO4 values of zone II waters are slightly sulfate in zone III and IVa waters intermediate. higher than those of zone I waters is due probab Similar results have been given by Matsubaya ly to partial oxygen isotopic exchange with et al. (1973) and Sakai and Matsubaya (1977). water in which the sulfate is dissolved. The The range of 634S values of sulfur-bearing dissolved sulfate should have 6180 value of species from steam wells is indicated on a. 514S about +9 %o, if complete equilibrium was axis of Fig. 5. Zone I waters are typical acid attained in water with S 18OH2O = -8%o at sulfate water. Such sulfate is considered to have 100°C (Mizutani and Rafter, 1969). Recent resulted from the surface oxidation of vol studies on reaction kinetics show that only canic sulfur at shallow depths where atmo partial equilibration is possible under low tem spheric oxygen is available. Bacterial oxidation perature (100° C) and neutral to weakly acid of reduced sulfur to form sulfate in geothermal (pH down to 2) conditions encountered in the environments may also be possible (Mosser et Hakone geothermal system (Chiba, pers. com.). al., 1973). Agreement of V'S values between As mentioned previously zone III waters are the zone I sulfate and sulfur gases in volcanic considered to be formed by mixing of sulfate steam from the Owakudani fumarolic area rich zone II waters with deeply generated dense supports the view of supergene origin of the steam carrying predominantly sodium chloride. sulfate, since it has been known that the isotopic Zone III and IVa waters are isotopically similar ratio of sulfur of reduced form is inherited by to each other as far as dissolved sulfate is con oxidized sulfur through oxidation processes cerned, suggesting a common origin. Both (Sakai, 1957; Ivanov et al., 1968; Schoen and 534SS04 and S18Oso4 values of sulfate in these Rye, 1970). waters, however, are higher than those of When atmospheric oxygen is utilized in zone I waters, and are close to those of a ground oxidative sulfate formation in an aqueous water sample GW-10. According to a hydrogeo medium, the oxygen isotopic ratio of the result chemical study of ground-waters in the Hakone ing sulfate is determined by the isotopic ratios caldera (Hirano and Oki, 1971), a shallow of both atmospheric and water oxygen with groundwater flow from which GW-10 was relative contribution of about 1/3 and 2/3, sampled is estimated to come down from zone respectively (Lloyd, 1967). If the S 18OH20 I through the central cone debris. GW-10 values are taken to be -8.5 %o for zone I waters contains sulfate as well as a measurable amount (Table 1) and 23.5 %o for atmospheric oxygen of ferrous ion with a Fe"/Fe 3' ratio as high as (Kroopnick and Craig, 1972) with the kinetic 25. Eh-pH calculations indicate that sulfate fractionation factor of -8.7 %o for the in ion, rather than hydrogen sulfide, is the stable corporation of atmospheric oxygen in the sul form of aqueous sulfur species under the redox fate (Lloyd, 1967), then the resulting sulfate and pH conditions estimated for GW-10 assum should have the 8110 value of about -0.7%o, ing attainment of equilibrium between Fe" and which is in good agreement with the observed ferric hydroxide (Hirano and Oki, 1971). How values of O%o for sulfate in zone I waters. ever, the solution can become more reducing Similarity in sulfur isotopic ratios in zone I during down-permeation through rocks con and II waters (excluding 11-2) suggests that taining ferrous minerals. Under such reducing sulfate in zone II waters may also have a similar conditions, dissolved sulfate can be reduced origin. In other words, acid zone I waters are probably by anaerobic bacteria. Bacterial neutralized during permeation through the reduction of sulfate is mostly accompanied by central cone materials to form zone II waters the enrichment of heavier isotopes in the re which occur about several hundred meters maining sulfate (Mizutani and Rafter, 1973), below the zone I level, forming a major aquifer to which isotopic enrichment observed for 38 S. Matsuo et alL

sulfate in zone III and IVa waters relative to implying mixing of down-permeating sulfate that in zone II water may be ascribed. with lower isotopic ratios as discussed previous If HCO3 in the Hakone hot spring waters is ly. derived from decomposition of fossil organic materials intercalated in the volcanic rocks as ORIGINOF THE THERMALWATER was proposed by Oki and Hirano (1970), its 813C values are expecred to be close to those of Estimation of isotopic composition of HTDS terrestrial plants which are typically in the range The hghest fumarolic activity in the of -20 to -25%o (PDB). Carbon isotopic caldera is seen in the Owakudani area, about analyses given in Table 5, however, show ex 1,000m above sea level at the northern slope istence of isotopically much heavier HCO3. of the Kamiyama, one of the central cones. The average 513C value of -0.5 ± 2.7%o for There are several steam wells of which highest HCO3 dissolved in zone II, III and IVa waters temperature is 1800C (CW-2) and natural indicates only small contribution of CO2 of fumaroles with boiling temperature of water. organic origin. This average value is consistent The thermal output due to steam and hot water with a 613C value of -0.7 %o for C02 in the discharge of this area, 1.1 X 10' cal s-1 (Yuhara, zone II steam well (CW-1) reported by Craig 1968) is about one third to one half of the et al. (1978). The above isotopic evidence sug entire thermal output from the Hakone geother gests that the ultimate source of carbon in the mal system (Oki and Hirano, 1970). There is zone II, III and IVa waters is sought in marine also a fumarolic area in the So-unzan area at the carbonates which occur as secondary minerals north-eastern flank of the Kamiyama, where in the Yugashima formation, the Miocene base several shallow steam wells are installed. The ment of submarine basalts and andesites under temperature of steam from this area is slightly lying the Hakone caldera. The CO2 may have less than 100° C. There are two steam wells in been carried up from the depths upon decom the Yunohanazawa area at about 1,000m level position of carbonate and dissolved in ground of the eastern slope of the Komagatake, one of water at the uppermost part of the Hakone geo the central cones. thermal system. Such carbonate-rich water may be mixed with descending sulfate-bearing ground waters to give isotopically heavy HCO3 10 rich waters in zones II and III. c Zone IVb waters form a group characterized 8 by sulfate with high oxygen and sulfur isotopic 0 0 ratios as shown in Fig. 5. Concentration of 0 6 c 0 HCO3 in these waters is low and its 513C X 4 value is more negative than the others (Table 5), 0 a a indicating higher contribution of organically 0 2 derived carbon. These waters occur at the coo lowest elevation of the Hakone geothermal 0 system, where the Yugashima formation is Z Range of 6345 for volcanicsulfur exposed. It is most likely that the zone IVb -2 -5 0 5 10 15 20 sulfate was leached out of the Yugashima forma 634s (%o) tion which contains sulfate as anhydrite (or gypsum) with 534S values of +15 ~+16%o Fig. S. S 180 versus S 34S plot of dissolved sulfate in the (Mizutani et al., 1975, Matsubaya et al., 1973).. thermal waters. The range of 6 34SHakone values f or The points for zone IVb waters extend towards sulfur from the steam wells is given on a horizontal axis lower 6180so ,4 and 634Sso4 values in Fig. 5, for comparison. Symbols are the same as those in Fi g s. 1 Fig. 2. Origin of thermal waters 39

Since the steam discharge from the Owaku the chloride content of HTDS is estimated to dani area is the greatest in the caldera as far as be 10,000 ppm. thermal output is concerned, the following dis According to our latest experimental results cussion will mainly be based on the isotopic (Shinohara et al., 1984) chlorine bearing species data of steam condensate from the Owakudani in the steam phase coexisting with acidic silicate area (CW-2), which exhibits the highest SD and melt saturated with water are mostly NaCl with 6180 values among all the Hakone samples much less HCI, in the range from the total studied. pressure of 1 to 6 kb at the temperature of The high 8 D and 8180 values of the Owaku 800 900°C. So that the steam derived from a dani steam (CW-2) suggest a close relationship deep-seated magma can contain a significant between CW-2 steam and HTDS. The average amount of NaC1. According to Sourirajan and values of CW-2 (a triangle at the far right end Kennedy (1962), the vapor with the tempera of Fig. 2), however, do not lie on the extension ture higher than 390'C with the pressure higher of the thermal water line, 5D = 2.18180 33.5, than 150b can contain 10,000ppm of chloride but slightly above the line. Yuhara (1968) ions in the form of NaCI. A large proportion estimated the temperature of subsurface reser of HTDS with the minimum temperature of voir of hot water beneath the central cones to 390°C encounters the major aquifer of RGW to be as high as 240"C on the basis of the enthalpy form zone III waters. A small portion of HTDS, measurement of the steam. We estimated the however, ascends through conduits without isotopic composition of the liquid in the reser mixing with RGW aquifer and condenses at a voir to be SD = -26%o and 8180 = +3%0, shallower and cooler place to form a liquid assuming the isotopic exchange equilibrium reservoir at the temperature around 240* C between the liquid and steam with the average beneath the Owakudani area. The isotopic observed isotopic composition of SD = -25% composition of HTDS may be taken to be the and 6180 = +I%, with the help of the liquid same as that of the 240°C reservoir which has vapor fractionation factors at 240* C given by been estimated above. Most of the dissolved Bottinga and Craig (1968). This set of values NaCI in HTDS should be left behind during the for the liquid lies just on the extention of the ascent owing to the temperature drop from thermal water (liquid) line in Fig. 2. Let us 390 to 240'C to form the final liquid reservoir. assume here that the liquid water with 6D = So that chlorine-bearing compounds in the final -26 %o and 8180 = +3 %o is a subsurface con reservoir should be mostly HCI. On the other densate of HTDS derived from a depth. hand, the concentration of volatile components We can now set the isotopic ratios of the such as HC1 and H2S in HTDS is almost kept two extremes, i.e., HTDS and RGW to be 5D = unchanged before the condensation occurs. -26 %o, 8180 = +3 %o and 8 D = -51 %o, 8180 = The steam from the subsurface liquid reservoir -8 .5 %o, respectively. As mentioned already reaches the surface to reveal fumarolic activity. zone III waters are the mixture of HTDS and The acid gases explain low pH of 1 to 2 of the RGW. The contribution of HTDS to zone III steam condensate, CW-2. waters is then calculated to be 25%, ranging from 20 to 36%, reflecting a slight change in Origin of HTDS The estimated values of isotopic ratios of zone III water. This estimate 8D = -26%o and 8180 = +3%o of HTDS can be agrees well with the HTDS contribution of 30% compared with 5D = -25 to -30%o and 8180 estimated by Oki and Hirano (1970) on the = +8 %o of the deep-lying hot brine of the Arima basis of the chemistry of zone III waters. When hot spring area, Hyogo Pref., Japan, estimated the chloride content of zone III waters and by Matsubaya et al. (1974), a narrow range mixing ratio of HTDS to zone III waters are centered at SD -25%o and 8180 = +7%o of taken to be respectively 2,500ppm and 25%, high temperature fumarolic gas condensates of 40 S. Matsuo et al.

Satsuma-Iwojima, off south end of Kyushu, water. Japan, measured by Matsuo et al. (1974) and Why is the 6D value of HTDS of the Hakone Matsubaya et al. (1975), and 6D = -28 to +6 %o geothermal system -26%o? We will present a and 6180 = +7%o of steam condensates from model which may account for the origin of White Island fumaroles, New Zealand (Stewart HTDS. The Hakone volcano (Pleistocene) is and Hulston, 1975). underlain by thick piles composed mostly of When we assume that the difference in submarine pyroclastic sediments of the lower to 6180 values of the above cited thermal waters middle Miocene (the Yugashima group). Com is due to the difference in the extent of rock mon alteration products of this group are calcite, water interaction at different temperatures, zeolites and chlorite (Kuno, 1950). When a 8180 of the deep thermal waters gives us no fluid reaches the Yugashima group at depths, the definite information on the ultimate origin of fluid interacts with pre-existing hydrous silicates the deep thermal waters. On the other hand, and changes its original D/H ratio. The final the similarity in 8 D values for the above cited 8 D value of the fluid should be -26 %o. systems may be an interesting problem to be In order to check this possibility, rock solved. The similarity in 6D values, however, samples of the Yugashima group were analyzed does not imply the direct contribution of for D/H ratio. Description of the samples and "juvenile water" , since there is no positive analytical results are given in Table 6. The 6D reason to imagine that the upper mantle beneath values of samples 1 and 2 in Table 6 can be con active zones such as Japanese and New Zealand sidered to represent those of typical pre-existing island arcs has retained the juvenile water hydrous silicates in the Yugashima group, and throughout its histories and the mantle water in this sense these samples are regarded as is estimated to have 6D values between -90 and "fresh". Samples 3 and 4 with lower 6D values, -75%0 (Matsuo et al., 1978). The facts that on the other hand, can be considered to have high 'He/'He ratio of the gas from CW-1 of 8.6 been subjected to later interaction with recent X 10-6 is about six times as high as the atmos thermal waters pertinent to the system. Thus pheric value of 1.4 X 10-6 and that He/Ne ratio samples 3 and 4 can be regarded as "altered" is 140 times higher than that of the atmosphere ones. Since chlorite is the predominant hydrous (Craig et al., 1978), however, indicate that most silicate mineral in the rock samples, chlorite of the helium must have derived from the can be considered to represent the hydrous juvenile helium retained in the mantle. This fact silicates in the samples on the first approxima is a clear indication of the least contamination tion. for the mantle helium during the degassing of A close look at Table 6 reveals that the water mantle volantiles, in contrast to the overwhelm content of the altered samples to be almost ingly large contribution of local surface water doubled as compared with that of the fresh that erases the isotopic signature of (juvenile?) Table 6. Total water content and D/H ratio of hydrous silicatesfrom the Yugashimagroup

Sample Sample description Water content 6D number (bulk, %) NO) 1 Typical Yugashima group at Atsugi (80m thick) 1.63 -55.0 2 Core sample from the basement rock of the Hakone area. Taken by Kuno which is located in the zone IVa 1.27 -53.1 209m deep 3 Core sample from the Yugashima group interacted with the Hakone thermal water. Taken from Suzawa which 3.22 -62 .4 is located in the middle of zone III, 293m deep

4 Yugashima group interacted with the Yugawara thermal water. Taken from Fudo-taki, Yugawara geothermal 3.28 -71 .5 area Origin of thermal waters 41

samples. This indicates that during water-rock or present-day) cannot be considered to be the interaction a part of water was incorporated into candidate for the initial water. For the same the hydrous mineral phases to give a net increase reason, the 1: 1 mixture of sea water and RGW, in chlorite content of the altered samples. In a which gives chloride content of 10,000ppm closed system we get the following isotopic (the estimated chloride content of HTDS), material balance equation for the water-rock can also be excluded. If we assume RGW as the interaction, initial water whose SD value is -51 %o, however, C/W ratio is calculated to be 0.5, which roughly WSW+ C6 i = (W- AW)SW corresponds to a weight ratio of rock to water of 14 for the rock containing 30% chlorite by +(C+AW)S1 (1), weight. The rock/water weight ratio (R/W)W where W and C are the number of hydrogen estimated above is higher than those for the atoms of water and chlorite, respectively, AW Salton Sea ((R/W), = 2, Clayton et al., 1968) the increase of chlorite hydrogen at the expense and Wairakei ((R/W), = 0.2, Clayton and of the equivalent amount of water, and i, f, c Steiner, 1975) geothermal systems, indicating and w denote initial, final, chlorite and water, the Hakone geothermal system to be more rock respectively. If we assume attainment of hydro dominated. gen isotopic exchange equilibrium, the follow The above results show that RGW can be a ing relation holds, source water of HTDS. If this is the case, SD value of the altered chlorite should be -66%o, 6c 6W = 1031na (2), sinceS~ SW= 103lna= -40. The hydrous silicates in the Yugashima group which have where a is the fractionation factor between interacted with the thermal waters give 6D chlorite and water. Since the water content of values of -62%o (sample 3 in Table 6) and the altered samples is twice as much as that of -72%o (sample 4 in Table 6) , which are in the fresh samples, good agreement with the estimated value. Oxygen isotopic material balance can be C+AW expressed in a form simpler than the case of C 2 (3). hydrogen isotopes as. follows,

Combining equations (1), (2) and (3), we get WSW+RSi=WSW +RSr (5),

C _ S f -S1 _ W W (4) . where W and R denote respectively the number W S' -SW-2X1031na of oxygen atoms of water and rock including not only hydrous silicates but also other rock We can now put the following values into equa forming minerals, Sr stands for initial 5"0 tion (4); 6 i = -54%o for the fresh chlorite value of the rock, and other notations are the (averageof samples1 and2) andSW = -26%o same as those in equation (1) but hydrogen is for SD of HTDS. 103 lna for the chlorite-water replaced by oxygen. Equation (5) can be re system can be found in Taylor (1974) and written as follows using oxygen isotope frac Marumo et al. (1980), and the value of -40 for tionation factor, 1031na, between rock an water, the temperature around 400'C is tentatively

taken for the purpose of calculation. R _ S f S i Since the C/W ratio is positive,S., SW W Sr SW 103lna (6) shouldalso be positive,i.e., SWshould be more negative than -26%o. From This, the water with The numeratorof equation(6), SW S. is so relatively high SD values such as seawater (fossil called oxygen isotopic shift of water, and the

C 42 S. Matsuo et al.

value of 11.5 %o corresponding to the isotopic clusion is shown in Fig. 6. The present results difference between HTDS and RGW can be suggest that thermal and volcanic waters with allotted. If we further assume 6' = +7%o for 6D values as high as -30 to -20%0 often ob the initialrock, S w = +3%o for the finalwater served for thermal water system (see the begin (HTDS) and 103lna = 2 which is the fractiona ning part of section "Origin of HTDS") may tion factor for the system plagioclase (Ab50An50) have been evolved from hydrothermal interac water at 400°C (O'Neil and Taylor, 1967) in tion of meteoric water with rocks containing equation (6), then R/W is calculated to be 5.8. hydrous silicates. This value corresponds roughly to rock/water weight ratio of 11.5, which is close to the value Acknowledgement-This paper has been influenced by estimated from the hydrogen isotopic evidence discussion with many colleagues and associates. The and shows that the Hakone geothermal system authors feel particularly indebted to Dr. 0. Matsubaya is rock-dominated, probably owing to the of Akita University and Dr. H. Sakai of Ocean Research Institute, Tokyo University for invaluable comments. limited supply of RGW in the caldera which is Thanks are extended to Mrs. Y. Matsuhisa, Miss K. Fuwa an independent hydrological unit (Matsuo et and Mr. Y. Tsutaki for their help in laboratory experi al., 1979). ments and to staff members of the Hot Springs Research On these bases, we conclude that the ulti Institute of Kanagawa Pref. for their help and coopera mate source of HTDS is meteoric in origin of tion during the field work. Isotopic analyses of sulfate were carried out at the Institute of Nuclear Sciences, which isotopic composition is similar to that of D.S.I.R., New Zealand while one of the authors (M.K.) RGW. A schematic presentation of our con was staying there. The authors are grateful to Dr. I. Friedman of U.S. Geological Survey, Denver, U.S.A., Dr. M. K. Stewart of Institute of Nuclear Science, (~ Steam D.S.I R., New Zealand, Dr. E. Mazor of Weizmann Ci~ l1 OD= -25 Institute of Science, Israel, and Dr. T. M. Gerlach of s'$0=+ 1 the Sandia National Laboratories, Albuquerque, U.S.A. for critical reading of the earlier version of the manu script. The authors thank Miss M. Kambayashi and Y. Matsuzawa for their cooperation during the preparation I NaCI y NOT liquid of the manuscript. removed reservoir Major aquifer' A part of the expense of this study was defrayed Zone by the Grant in Aid for Scientific Research from the 21D 5 8. 5: ...I.11--& IV waters Ministry of Education, Science and Culture of Japan, I 1Y and the Grant from Kanagawa Prefecture.

/ HTDS i with RGW NaCI+HCI REFERENCES Tertiary base ment rocks Bigeleisen, J., Perlman, M. L. and Prosser, H. C. (1952) 6D=-54 y `•Isotopic Exchange Conversion of hydrogenic materials to hydrogen for 1 at 400°C isotope analysis. Anal Chem. 24, 1356-1357. 6D,.~,ck=-62----71 Ii / Bottinga, Y. and Craig, H. (1968) High temperature Rock/Water 10 liquid-vapor fractionation factors for H2O-HDO H2180. Trans. Am. Geophys. Union 49,356-3S7. Clayton, R. N., Muffler, L. J. P. and White, D. E. (1968) Oxygen isotope study of calcite and silicates of the

x X X Heat Source " River Ranch No. 1 well, Salton Sea geothermal field,

X x X X California. Am. J. Sci. 266,968-979. Clayton, R. N. and Steiner, A. (1975) Oxyten isotope Fig. 6. A schematic presentation of a genetic model for studies of the geothermal system at Wairakei, New the thermal waters in the Hakone geothermal system. Zealand. Geochim. Cosmochim. Acta 39, 1179 See text for details. 1186.

i Origin of thermal waters 43

Craig, H., Boato, G. and White, D. E. (1956) Isotopic adjacent areas, Part II. J. Fac. Sci., Univ. Tokyo. geochemistry of thermal waters. Proc. 2nd Conf. Sect. 117, 351-402. Nucl. Proc. Geol. Settings, 29-38. Kusakabe, M., Wada, H., Matsuo, S. and Horibe, Y. Craig, H., Lupton, J. E. and Horibe, Y. (1978) A (1970) Oxygen and hydrogen isotopic ratios of mantle helium component in circum-Pacific volcanic monthly collected waters from Nasudake volcanic gases: Hakone, the Marianas, and Mt. Lassen. Ter area, Japan. J. Geophys. Res. 75, 5941-5951. restrial rare gases. Edited by E. C. Alexander, Jr. and Lloyd, R. M. (1967) Oxygen-18 composition of M. Ozima, 3-16, Center for Academic Publications oceanic sulfate. Science 156, 1228-1231. Japan. Marumo, K., Nagasawa, K. and Kuroda, Y. (1980) Deuser,W. G. and Hunt, J. M. (1969) Stable isotope Mineralogy and hydrogen isotope geochemistry of ratios of dissolvedinorganic carbon in the Atlantic. clay minerals in the Onuma geothermal area. Earth Deep-SeaRes. 16,221-225. Planet. Sci. Lett. 47, 255-262. Friedman, I. and Smith, G. I. (1970) Deuterium Matsubaya, 0., Sakai, H. and Tsurumaki, M. (1974) content of snow cores from the Sierra Nevada area. Hydrogen and oxygen isotopic ratios of thermal Science 169,467-470. and mineral springs in Arima area. Okayama Daigaku Fiedman, I. and Smith, G. I. (1972) Deuterium con Onsen Kenkyusho Hokoku No. 43, 15-28 (Japa tent of snow as an index to winter climate in the nese). Sierra Nevada area. Science 176, 790-793. Matsubaya, 0., Sakai, H., Kusachi, I. and Satake, H. Hirano, T. and Oki, Y. (1971) Geochemistry of (1973) Hydrogen and oxygen isotopic ratios and groundwaters of Hakone caldera. Bull. Hot Spring major element chemistry of Japanese thermal water Res. Inst., Kanagawa Pref. 2, 89-108 (Janaese). systems. Geochem. J. 7, 123-151. Hirano, T. and Tajima, Y. (1969) Geochemistry of Matsubaya, 0., Ueda, A., Kusakabe, M., Matsuhisa, Y., the Ubako hot springs, Hakone volcano. Bull. Hot Sakai, H. and Sasaki, A. (1975) An isotopic study spring Res. Inst., Kanagawa Pref. No. 8, 13-22 of the volcanoes and the hot springs in Satsuma (Japanese). Iwojima and some areas in Kyushu. Chishitsu Cho Hirano, T. and Tajima, Y. (1970) Tables of chem sasho Geppo 26, 375-392 (Japanese). ical analysis of mineral waters (1966-1969) from Matsuo, S., Kuroda, Y., Suzuoki, T., Aoki, K. and the Hot Spring Research Institute of Kanagawa Hariya, Y. (1978) Mantle water based on the Prefecture, Part 2, Bull. Hot Spring Res. Inst., Ka hydrogen isotopic ratios of hydrous silicates in the nagawa Pref No. 10, 1-54 (Japanese). mantle. Short Papers 4th Int. Conf Geochronology, Hirano, T., Oki, Y. and Awaya, T. (1972) Geochem Cosmochronology and Isotope Geology, 278-280. istry of hydrothermal system of the Yumoto-Tonosa Matsuo, S., Kusakabe, M., Niwano, M., Hirano, T. and wa Area, the eastern foot of Hakone volcano. Bull. Oki, Y. (1979) Water budget in the Hakone Hot Spring Res. Inst., KanagawaPref. 3, 109-130 caldera using hydrogen and oxygen isotope ratios. (Japanese). Isotopes in lake studies, 131-144. IAEA, Vienna. Hirano, T., Oki, Y. and Tajima, Y. (1968) Effect of Matsuo, S., Suzuoki, T., Kusakabe, M., Wada, H. and anomalous temperature on the chemical composi Suzuki, M. (1974) Isotopic and chemical compo tion of thermal water, Gora hot springs, the Hakone sitions of volcanic gases from Satsuma-Iwojima, volcano. Bull. Hot Spring Res. Inst., Kanagawa Pref. Japan. Geochem. J. 8, 165-173. No. 6, 51-62 (Japanese). Mizutani, Y. (1971) An improvementin the carbon Hirota, S. and Odaka, S. (1975) Continuousobserva reduction method for the oxygen isotopic analysisof tion of temperatures and water-levelsin the Hakone sulfates. Geochem.J. 5, 69-77. volcano, 1974. Bull.Hot SpringRes. Inst., Kanagawa Mizutani, Y., Asai, H. and Hamasuna, T. (1975) Origin Pref. 6, 79-86 (Japanese). of neutral chloride thermal waters from the south Ivanov, M. V., Grinenko, V. A. and Lein, A. Yu. (1968) eastern part of Izu peninsula, Japan. Kazan 19, 139 Exogenic formation of sulfuric acid in fumarole fields 150 (Japanese). of volcanoes of the Kurile Islands. Geochem. Int. 9, Mizutani, Y. and Rafter, T. A. (1969) Oxygen iso 1133-1140. topic composition of sulphates-Part 3. Oxygen Kroopnick, P. and Craig, H. (1972) Atmospheric isotopic fractionation in the bisulphate ion-water oxygen: isotopic composition and solubility frac system. N. Z. J. Sci. 12, 54-59. tionation. Science 175, 54-55. Mizutani, Y. and Rafter, T. A. (1973) Isotopic Kuno, H. (1950) Geology of Hakone volcano and behaviour of sulphate oxygen in the bacterial reduc adjacent areas, Part I. J. Fac. Sci., Univ. Tokyo, tion of sulphate. Geochem. J. 6, 183-19 1. Sect. II 7, 257-279. Moser, H. and Stichler, W. (1970) Deuterium mea Kuno, H. (1951) Geology of Hakone volcano and surement on snow samples from the Alps. Isotope 44 S. Matsuo et al.

hydrology, 43-57. IAEA, Vienna. hydrogen isotope geochemistry of clay minerals. Mosser, J. L., Mosser, A. G. and Brock, T. D. (1973) Geochim. Cosmochim. Acta 34, 25-42. Bacterial origin of sulfuric acid in geothermal hab Schoen, R. and Rye, R. 0. (1970) Sulfur isotope itats. Science 179, 1323-1324. distribution in solfataras, Yellowstone National Oki, Y. and Hirano, T. (1:970) The geothermal system Park. Science170, 1082-1084. of the Hakone volcano. Geothermics, Special Issue 2, Shinohara, H., Iiyama, J. T. and Matsuo, S. (1984) 1157-1166. Comportement du chlore dans le systeme magma Oki, Y. and Hirano, T. (1974) Hydrothermalsystem granitique-eau. C. R. Acd. Sci. Paris 298, 741-743. and seismic activity of Hakonevolcano. The utiliza Sourirajan, S. and Kennedy, G. C. (1962) The system tion of volcano energy. Edited by J. L. Colp and A. H20-NaCI at elevated temperatures and pressures. S. Furumoto, 13-40. Am. J. Sci. 260,115-141. Oki, Y., Ogino, K., Hirota, S., Ohguchi, T. and Morita, Stewart, M. K, and Hulston, J. R. (1975) Stable M. (1968) Anomalous temperature encountered in isotope ratios of volcanic steam from White Island, the Gora hydrothermal system of the Hakone volcano New Zealand. Bull. Volcanol. 39, 28-46. and its hydrological explanation. Bull. Hot Spring Tajima, Y. and Hirano, T. (1966) Tables of chem Res. Inst., Kanagawa Pref. No. 6, 1-20 (Japanese). ical analysis of mineral waters (1961-1966) from O'Neil, J. R. and Taylor, H. P., Jr. (1967) The the Hot Spring Research Institute of Kanagawa oxygen isotope and cation exchange chemistry of Prefecture. Bull. Hot Spring Res. Inst., Kanagawa feldspars. Am. Mineral. 52, 1414-1437. Pref. No. 3, 1-55 (Japanese). Rafter, T. A. (1957) Sulphur isotopic variation in Taylor, H. P., Jr. (1974) The application of oxygen nature. Part 2, A quantitative study of the reduction and hydrogen isotope studies to problems of hydro of barium sulphate by graphite for recovery of sul thermal alteration and ore deposits. Econ. Geol. 69, phate-sulphur for sulphur isotopic measurements. 843-883. N. Z. J. Sci., Technol. Sect. B 38, 955-968. Watanuki, K. (1966) Chemical composition of Owa Robinson, B. W. and Kusakabe, M. (1975) Quan kudani hot spring water, Hakone. Onsen Kagaku 17, titative preparation of sulfur dioxide, for 345/325 22-27 (Japanese). analyses, from sulfides by combustion with cuprous White, D. E. (1957) Thermal waters of volcanic oxide. Anal Chem. 47,1179-1181. origin. Bull. Geol. Soc. Am. 68,1637-1658. Sakai, H. (1957) Fractionation of sulphur isotopes in White, D. E., Muffler, L. J. P. and Truesdell, A. H. nature. Geochim. Cosmochim. Acta 12, 150-169. (1971) Vapor-dominated hydrothermal systems Sakai, H. and Matsubaya, 0. (1974) Isotopic geo compared with hot-water systems. Econ. Geol. 66, chemistry of the thermal waters of Japan and its 75-97. bearing on the Kuroko ore solutions. Econ. Geol. Yuhara, K., Kodai, K., Abe, K., Kotoda, K. and Hosono, 69,974-991. Y. (1966) Hydrothermal system of Owakudani Sakai, H. and Matsubaya, 0. (1977) Stable isotopic Gora zone, Hakone volcano. Rep. Cooperative Res. studies of Japanese geothermal systems. Geothermics Disaster Prevention No. 8, 29-42 (Japanese). 5,97-124. Yuhara, K. (1968) Quantitative estimation of hydro Sato, K. (1962) On the thermal springsand geology thermal system of Owakudani and Sounzan geother of Izu-Hakone district. Onsen Kagaku 13, 41-51 mal areas, Hakone volcano. Kazan 13, 74-83 (Japanese). (Japanese). Savin, S. M. and Epstein, S. (1970) The oxygen and