Fracture Pattern and Low-Temperature Geothermal ...

九大地熱・火山研究報告 第 16 号 (2007) 7-13 頁

Fracture Pattern and Low-Temperature Geothermal Systems in Fukuoka Prefecture, Southwestern Japan Hakim SAIBI*, Jun NISHIJIMA*, Tomohiro HIRANO**, Yasuhiro FUJIMITSU* and Sachio EHARA* ABSTRACT

Fukuoka area is located in the southwestern part of Japan. This area has some low-temperature geothermal systems. From 1996 to 2007, Fukuoka area was covered by gravity surveys using Scintrex CG-3 and CG-3M gravimeters in an attempt to delineate the subsurface structure. The gravity data has been analyzed using some integrated gradient interpretation techniques such as: Horizontal Gradient (HG), Tilt Derivative (TDR), and Euler deconvolution method. The results of this study will hopefully lead the good understanding of the relation between the interpreted faults and the location of the lowtemperature geothermal systems and possible future geothermal exploration in the area. 1. INTRODUCTION

Karakida et al., 1994). The basement rock is composed essentially of granite (Matsushita et al., 1971).

Some low temperature hot springs are localized in the southern part of Fukuoka city, exactly in Yokote-Ijiri area. The generation of the hot springs in this area is not related with a special heat source (Karakida et al., 1994). Fujimitsu et al. (2003) stated the possibility of generation by a specific subsurface structure. The Kego fault, one of the active faults in Fukuoka area was expected to pass through Yokote-Ijiri area (Karakida et al., 1994) but at that time no scientific investigations have been conducted to locate the Kego fault. The studied area was struck by a strong earthquake on March 20, 2005 (MJMA7.0) at depth (9 kmJMA) and a MJMA5.8 aftershock on April 20. From previous geosciences’studies, the earthquake occurred along the extension of Kego fault under the sea of Genkai, running from the northwest to

Figure 1: Location of Fukuoka area in Kyushu Island,

southeast.

Japan.

Therefore, the aim of this study is to delineate the subsurface structure of Fukuoka area and its relation with

Fukuoka area was covered by gravity surveys using

the low-temperature geothermal systems by applying some

Scintrex CG-3 and CG-3M gravimeters during the period

gradient interpretation methods to the available gravity

1996-2007. The total number of gravity stations is 1590

data.

over the survey area covering approximately 300 km 2. The average spacing between gravity stations is 50 m to

2. GRAVITY DATA

2 km. The Bouguer gravity data has been analyzed using

Fukuoka area is located in the southwestern part of Japan,

some integrated gradient interpretation techniques. The

between latitudes 33˚30 ' -33˚45 ' N and longitudes 130˚ 15 ' -130˚30 ' E (Figure 1). Geologically, Fukuoka area is

gravity data is used in this study in an attempt to delineate

composed of Paleozoic Sangun metamorphic rocks, Late

2470 kg/m3 (Hirano et al., 2006) was used to yield the

Mesozoic granitic rocks, Paleogene, Neogene basaltic

Bouguer anomaly map of the study area as shown in Figure

rocks and Quaternary sediments (Matsushita et al., 1971;

2-A. From visual inspection in Figure 2-A, this area is

the subsurface structure of Fukuoka area. A density of

* Laboratory of Geothermics, Department of Earth Resources Engineering, Faculty of Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka, 819-0395, Japan **Laboratory of Geothermics, Department of Earth Resources Engineering, Graduate School of Engineering, Kyushu University, 744 Motooka, Nishi-ku, Fukuoka, 819-0395, Japan 7

Hakim SAIBI, Jun NISHIJIMA, Tomohiro HIRANO, Yasuhiro FUJIMITSU and Sachio EHARA

characterized by the gravity value distribution which has

overturned structure affecting this area. This fracture zone

a trend of increasing in the northern, and western parts of

is believed to represent the Kego fault.

the map area. Two low Bouguer gravity anomalies can be seen in the central part of the area extending in NW-SE direction. The hot springs are located at the border between very-low and low Bouguer gravity anomalies (Figure 2-B). This region can reflect a small basin elongated in the NWSE direction.

Figure 3: Stratigraphic log of two drill-holes (Well 1 and Figure 2: A) Bouguer gravity map of Fukuoka area. Black

10) in Yokote-Ijiri area (Matsushita et al., 1971). The two

circles indicate the locations of the gravity stations. The

drill-holes are used as control points in the forward gravity

black rectangle indicates the area of Figs. 5, 6 and 7. B) Hot

modeling (see Figure 10).

spring location is overlapped on the Bouguer gravity map. 3. METHODOLOGY Based on petrographic and macroscopic analysis of

In our study three gravity interpretation techniques were

drill cuttings and core samples (Matsushita et al., 1971),

used: HG, TDR, and Euler deconvolution. The combination

the stratigraphic column of Yokote-Ijiri, southern part of

of the three methods will enhance the structural definition

Fukuoka city can be divided into three layers: Layer 1:

of the study area. The TDR has the advantage of responding

gravel and sand, Layer 2: weathered granite and Layer 3

well to both shallow and deep sources and the map of TDR

corresponding to the Mesozoic granite, which represents

recognizes the horizontal location and extent of sources.

the basement rock in this area (Figure 3). The Layer 1

The location of the maximum HG may be used as an

is absent in the Well No. 1. The Paleogene sandstone is

indicator of the location of edges of the source. The Euler

detected at the bottom of Well No. 10, beneath the fracture

solutions give the depths of sources. The HG and TDR

zone. Matsushita et al. (1971) explained this structure by an

techniques are illustrated in Figure 4. 8

Fracture Pattern and Low-Temperature Geothermal Systems in Fukuoka Prefecture, Southwestern Japan

horizontal derivatives of the field (Phillips, 1998). The method is also robust in delineating both shallow and deep sources, in comparison with the vertical gradient method, which is useful only in identifying shallower structures. The amplitude of the horizontal gradient (Cordell and Grauch, 1985) is expressed as:

HG g

g x

2

g y

2

(1)

where (∂g/∂x) and (∂g/∂y) are the horizontal derivatives of the gravity field in the x and y directions, respectively. The amplitude of the horizontal gradient of Fukuoka area was calculated in the frequency domain and is illustrated in Figure 5. The area may be dissected by major faults striking in the N-S and NW-SE directions.

Figure 4: Gravity anomaly (mGal) over one block. Density contrast is 1000 kg/m3. TDR= Tilt angles (Degree), HG= Horizontal Gradient (mGal/m), VD= Vertical Derivative (mGal/m).

Figure 5: Horizontal gradient map of gravity data.

4. HORIZONTAL GRADIENT OF GRAVITY DATA

5. TILT DERIVATIVE OF GRAVITY DATA

The horizontal gradient method has been used intensively

The TDR is used to enhance and sharpen the potential

to locate boundaries of density contrast from gravity data

field anomalies. The advantage of the TDR is that the

or pseudogravity data stating that the horizontal gradient

zero contour line located on or close to the contacts. The

of the gravity anomaly caused by a tabular body tends to

operating definition is (Miller and Singh, 1994):

overlie the edges of the body if the edges are vertical and well separated from each other (Cordell, 1979; Cordell and Tilt

Grauch, 1985).

tan

The greatest advantage of the horizontal gradient method is that it is least susceptible to noise in the data, because it only requires the calculations of the two first-order

where 9

1

vertical component of gradient horizontal component of gradient

tan

1

g z g h

(2)

Hakim SAIBI, Jun NISHIJIMA, Tomohiro HIRANO, Yasuhiro FUJIMITSU and Sachio EHARA

g h

2

g x

g y

where

2

       (3)

(x0, y0, z0) are the position of a source whose total gravity is detected at (x, y, z), b is the regional value of the gravity, and η is the structural index (SI) which can be defined as

The result of TDR is presented in Figure 6.

the rate of attenuation of the anomaly with distance. SI must be chosen according to a prior knowledge of the source geometry. For example, in gravity case, SI=2 for a sphere, SI=1 for a horizontal cylinder, SI=0 for a fault, and SI=-1 for a contact (FitzGerald et al., 2004). The two horizontal gradients (∂g/∂x, ∂g/∂y) and a vertical derivative (∂g/∂z) are used to compute the anomalous source locations. By considering four or more neighboring observations at a time (an operating window), source location (x0, y0, z0) and b can be computed by solving a linear system of equations generated from the Equation (5). Then by moving the operating window from one location to the next over the anomaly, multiple solutions for the same source are obtained. In our study, a technique of Euler deconvolution has been applied to the gravity data using a moving window of 250 m X 250 m. The grid cell size is 500 m. We have assigned several structural indices values and found that SI of (0) gives good clustering solutions. Reid et al. (1990; 2003), and Reid (2003) presented a structural index equal to zero for the gravity field for detecting faults. The results of the Euler deconvolution for gravity data are shown in

Figure 6: Tilt derivative of gravity data. The zero value

Figure 7.

matches well with the HG map of gravity data.

The interpretation of Euler solutions (Figure 7) indicates that the NE-SW, NW-SE, E-W, and N-S trends characterize

6. EULER DECONVOLUTION OF GRAVITY DATA

the structural setting of Fukuoka area. The depth of faults

Euler deconvolution is used to estimate depth and

ranges from less than 500 m to more than 1000 m.

location of the gravity source anomalies. The method was established by Thompson (1982) and applied essentially for real magnetic data along profiles. Reid et al. (1990) followed up a suggestion in Thompson’s paper and developed the equivalent method operating on gridded magnetic data. The application of Euler deconvolution to gravity data has been carried out by several authors, e.g., Wilsher (1987), Corner and Wilsher (1989), Klingele et al. (1991), Marson and Klingele (1993), Fairhead et al. (1994), and Huang et al. (1995). The 3D equation of Euler deconvolution given by Reid et al. (1990) is: g x

x xo

y yo

g y

g z

z zo

g

(4)

Equation (4) can be rewritten as:

x

g x

y

g y

z

g z

g

xo

g x

yo

g y

zo

g z

(5) Figure 7: Euler solutions for a structural index of zero. 10

Fracture Pattern and Low-Temperature Geothermal Systems in Fukuoka Prefecture, Southwestern Japan

8. DISCUSSION AND CONCLUSION

The maximum relative error is 10 %. Solutions are selected using standard criteria. The solutions of Euler

Figure 9 shows a tentative qualitative interpretation of

deconvolution are consistent with the results of TDR and

the Horizontal Gradient data and the Tilt Derivative data.

HG. Background is the image of Bouguer gravity map.

The Euler deconvolution method provides the depth of the

The depth of Kego fault is less than 500 m as shown by the

faults; however the HG and TDR methods determine the

good clustering of the Euler solutions.

location of the faults. The results show a relation between the structural pattern and locations of the hot springs at

7. HOT SPRINGS IN FUKUOKA AREA

Fukuoka area. The hot spring waters emerge along fault

Eleven low temperature hot springs are located in Yokote-

lineaments in Fukuoka area. Fukuoka area is dissected by

Ijiri area in southern part of Fukuoka city. This hot spring

major faults striking in the E-W, and NW-SE direction. The

area is characterized by a low resistivity less than 10 Ωm

depth of these faults is less than 500 m, which is obtained

and it was explained that the low resistivity was caused by

from the Euler deconvolution method. The faults located at

hot water (Matsushita et al., 1971). Fujimitsu et al. (2003)

the eastern side of the hot springs are deeper than the faults

mentioned that the high temperature distribution of the hot

of the western side. The rose diagram (Figure 9) indicates

springs in Yokote-Ijiri area has an extension of NW-SE

that there are four major faults patterns (N-S, E-W, NE-SW,

direction (Figure 8). The hot springs are located in a dense

NW-SE) characterizing the study area.

fractured area. The characteristics of these hot springs are mentioned in Table 1. The hottest hot spring is Well No. 1 with 49 ºC.

Figure 8: Hot spring temperature distribution. Numbers are hot springs. Table 1: Hot spring waters characteristics in Fukuoka area (Matsushita et al., 1971). Well number 1 2 3 4 5 6 7 8 9 10 11

Temperature in oC 49 43 37.5 46 31 42 27 46 34.3 33.1 43

Depth in meter 72 70 100 78 100 150 100 150 150 130 100

Figure 9: Faults interpretation map of Fukuoka area from

Amount in l/min. 100 54.6 120 50 26 110 20 120 41 34.2 120

HG and TDR maps. Eleven hot springs are plotted with their numbers. A rose diagram of faults extracted from gravity data at Fukuoka area is presented. The interpreted faults are statistically analyzed using the GIS software (ER Mapping ver. 7.0) (Table 2). From Table 2 some important information about the tectonic setting of the study area were obtained: 11

Hakim SAIBI, Jun NISHIJIMA, Tomohiro HIRANO, Yasuhiro FUJIMITSU and Sachio EHARA

Table 2: Statistical analysis of interpreted lineaments in Fukuoka area. The total of lineaments is 112. Using the count of number of lineaments, we have two main groups: the first group (11 lineaments) in the direction of -40ºW, and the second group (11 lineaments) in the direction of 90ºE. From the point of view of length, the first group has a long length, while the second group has a short length. The maximum length is observed at -30ºW with 2421 m, while the shortest lineament is observed at 60ºE with 648 m. Minimum

Maximum

Mean

Total

Mean

Length

Length

Length

Length

Angle

(m)

(m)

(Degree)

560.14

2800.72

-78.8

858.8

6011.57

-70.6

1019.5

541.77

4334.18

-58.5

316.4

2315.41

922.46

9224.6

-50.5

11

387.83

2169.53

1003.11

11034.2

-40.9

-30

5

286.71

2421.49

1041.38

5206.88

-30.7

-20

9

208.18

1355.82

682.96

6146.65

-21.4

-10

6

223.17

2236.61

880.24

5281.45

-11.7

Angle

Number of

(Degree)

Lineaments

(m)

(m)

-80

5

194.41

840.29

-70

7

192.58

1579.5

-60

8

261.19

-50

10

-40

Figure 10: 2-D model by using two wells 10 and 1 (No. 1

0

7

228.21

1580.04

666.68

4666.75

-2

and 10, see Figure 2-B) as control point based on forward

10

8

349.27

705.73

531.95

4255.58

7.8

20

4

213.24

922.29

554.79

2219.16

22.4

modeling of the gravity data using Talwani’s algorithm

30

3

303.91

863.27

562.63

1687.88

29.6

(Talwani et al., 1959). The hot water flows from downward

40

4

238.41

1442.43

714.12

2856.48

41.6

to the surface through the fault structure. Kego fault could

50

4

271.44

1361.05

839.73

3358.93

53.5

60

1

648.54

648.54

648.54

648.54

62.1

be the main flow path of the geothermal fluids.

70

2

184.09

665.25

424.67

849.35

70.9

80

7

170.47

1213.9

575.83

4030.81

79.9

90

11

118.31

1433.1

528.15

5809.62

88.5

ACKNOWLEDGMENTS

1- The lineaments in NW-SE direction are of high length

The first author acknowledges the financial support of

(2420 m), the lineaments in718.07 E-W direction are of Lineaments 112while 118.31 2421.49 80423.34

Japan Society for the Promotion of Science (JSPS) for

short length. From the cross-cutting relationships, the

the research activities in Japan. This study is supported

lineaments in NW-SE direction are newer (geological

by KAKENHI (Grant-in-Aid for Exploratory Research

time) than the lineaments in E-W direction. In general,

by JSPS) No. 17651100 (Principal Researcher: Yasuhiro

a newer fault cuts an older fault (cited by Warvelle,

Fujimitsu)

All

1968 in his book, page 23). REFERENCES

2- In the NW-SE direction, many faults in this direction with 10 degrees of shift toward the north or to the

Cordell, L. and V. J. S. Grauch (1985) Mapping basement

west due to the tectonic reactivation were observed.

magnetization zones from aeromagnetic data in the San

3- The faults in the NW-SE direction were tectonically

Juan Basin, New Mexico, in Hinze, W. J., Ed., The utility of

reactivated. It can be seen in recent earthquakes

regional gravity and magnetic anomaly maps, Soc. Explor.

happening in this area, caused mainly by the Kego

Geophys., 181-197.

fault. Cordell, L. (1979) Gravimetric expression of graben Figure 10 shows the 2-D model along a line A-B (Figure

faulting in Santa Fe Country and the Espanola Basin,

2-B). Two geological layers with different densities are

New Mexico, in Guidebook to Santa Fe Country, 30th

included in the model using the drilling data of Well 1

Field Conference, Edited by R. V. Ingersoll, New Mexico

and 10. The model presents a basin with 90 m of depth,

Geological Society, pp. 59-64.

bordered by faults. The hot water flows to the surface through the faults. The Kego fault can be the main flowing

Corner, B. and W. A. Wilsher (1989) Structure of the

path of the geothermal water. The location and depth of

Witwatersrand basin derived from interpretation of the

faults agree with the results of the HG, TDR, and Euler

aeromagnetic and gravity data, in Garland, G. D., Ed.,

deconvolution method.

Proceedings of exploration '87, third decennial international

The obtained structural map is suitable for planning of

conference on geophysical and geochemical exploration for

geothermal research in Fukuoka area.

minerals and groundwater, Ontario Geol. Survey, Special, 3, 12

Fracture Pattern and Low-Temperature Geothermal Systems in Fukuoka Prefecture, Southwestern Japan

532-546.

overturned structure observed in the southern part of Fukuoka city, The Science reports of the Shimabara Volcano

Fairhead, J. D., K. J. Bennet, R. H. Gordon, and D. Huang,

Observatory, the Faculty of Science, Kyushu University,

Euler (1994) Beyond the‘Black Box’, 64th Ann. Internat.

No. 7, 1-8. (in Japanese with English abstract)

Mtg., Soc. Expl. Geophys., Expanded Abstracts, 422-424. Miller, H. G. and V. Singh (1994) Potential field tilt

a

FitzGerald, D., A. Reid, and P. McInerney (2004) New

new concept for location potential field sources, Applied

discrimination techniques for Euler deconvolution,

Geophysics, 32, 213-217.

Computers & Geosciences, 30, 461-469. Phillips, J. D. (1998) Processing and interpretation of Fujimitsu, Y., J. Nishijima, Y. Oba, and S. Ehara (2003)

aeromagnetic data for the Santa Cruz Basin-Patahonia

Low temperature hydrothermal system in the southern area

Mountains area, South-Central Arizona, U. S. Geological

of Fukuoka city, Japan. Proceedings of the 38th Conference

Survey Open-File Report 02-98.

of Societe Internationale des Techniques Hydrothermales and the 56th Annual Meeting of the Balneological Society

Reid, A. B., J. M. Allsop, H. Granser, A. J. Millet, and

of Japan, Beppu, Japan, 2003, pp. 119-124.

I. W. Somerton (1990) Magnetic interpretation in three dimensions using Euler deconvolution, Geophysics, 55,

Hirano, T., J. Nishijima, and Y. Fujimitsu (2006) Estimation

80-91.

of fault structure using densed gravity survey data at Fukuoka city. Geothermal and Volcanological Research

Reid, A. B. (2003) Short note, Euler magnetic structural

Report of Kyushu University, No. 15, 16-22. (in Japanese

index of a thin bed fault, Geophysics, Published

with English abstract)

electronically.

Huang, D., D. Gubbins, R. A. Clark, and K. A. Whaler (

Reid, A., D. FitzGerald, and P. McInerney (2003) Euler

1995) Combined study of Euler’s homogeneity equation

deconvolution of gravity data. SEG annual meeting, Dallas.

for gravity and magnetic field, 57th Conf. & Tech. Exhib., Euro. Assoc., Expl. Geophys, Extended Abstracts, 144.

Talwani, M., J. L. Worzel, and M., Landisman (1959) Rapid gravity computations for two-dimensional bodies with

Karakida, Y., S. Tomita, S. Shimoyama, and K. Chijiwa (

applications to the Mendocino submarine fracture zone, J.

1994) Geology of the Fukuoka district, Quadrangle series,

Geophys. Res., 64, 49-59.

Scale 1:50000, Fukuoka (14) No. 51, 1-192. (in Japanese with English abstract)

Thompson, D. T. (1982) EULDPH A new technique for making computer-assisted depth estimates from magnetic

Klingele, E. E., I. Marson, and H. G. Kahle (1991)

data, Geophysics, 47, 31-37.

Automatic interpretation of gravity gradiometric data in two dimensions: vertical gradient, Geophysical Prospecting, 39,

Warvelle, J. (1968) Stratigraphy and geologic time, edited

407-434.

by W. C. Brown Co, 113 pp.

Marson, I. and E. E. Klingele (1993) Advantages of using

Wilsher, W. A. (1987) A structural interpretation of

the vertical gradient of gravity for 3-D interpretation,

the Witwatersrand basin through the application of

Geophysics, 58. No. 11, 1588-1595.

the automated depth algorithms to both gravity and aeromagnetic data, M. Sc. thesis, Univ. of Witwatersrand.

Matsushita, H., T. Miki, and A. Yamashita (1971) An

13