Long range gas-geochemical anomalies of a remote earthquake
Transcription
Long range gas-geochemical anomalies of a remote earthquake
Geochemical Journal, Vol. 45, pp. 137 to 156, 2011 Long range gas-geochemical anomalies of a remote earthquake recorded simultaneously at distant monitoring stations in India HIROK CHAUDHURI ,1* WASEEM B ARI,2 NASEER IQBAL,2 R AKESH K. BHANDARI,1 DEBASIS GHOSE ,1 PRASANTA S EN3 and BIKASH S INHA1 1 2 Variable Energy Cyclotron Centre, 1/AF Bidhannagar, Kolkata - 700 064, India Physics Department, University of Kashmir, Hazratbal, Srinagar -190 006, India 3 Saha Institute of Nuclear Physics, 1/AF Bidhannagar, Kolkata - 700 064, India (Received June 5, 2010; Accepted November 7, 2010) Continuous He, 222Rn concentrations and gamma dose rate were simultaneously monitored for pre-seismic signatures at two thermal springs in India that are separated at a distance of approximately 1612 km. The recordings of six month monitoring period (September 15, 2007–March 15, 2008) are presented here. In this paper, we make a cross correlation study of the simultaneously recorded multi-component gas-geochemical data from the two thermal springs. In the monitoring period a number of prominent fluctuations beyond the regular diurnal variations in the gas composition were recorded. We attempt to find out the linkage between the recorded distinct variations with local and remote seismic activities. Anomalous fluctuations in the spring gases were observed during the period December 24–27, 2007 at both the thermal springs and these anomalies may be correlated to the distant (~1000 km) China earthquake of magnitude M = 6.3 that occurred on January 09, 2008. Based on the obtained sequence of data points a time series analysis to relate earthquake magnitude, epicentral distance and precursor time through statistical methods and empirical equations related to the zone of influence is made. Concurrent monitoring of multi-component gas-geochemical anomalies recorded at a number of distant sites appears to be a potential tool to deal with the commonly debated question of earthquake precursors. Keywords: thermal spring, gas-geochemical monitoring, He, quake precursor 222 Rn, gamma dose, pre-seismic anomalies, time series analysis, earth- mechanisms are well understood (Weinlich et al., 2006). Seismicity as well as associated geochemical precursory changes in ground waters, mineral spring gases (ambient temperature), thermal spring gases and soil gases are both intrinsically linked to the time dependant interaction between tectonic stress, fluid flow, pore-pressure changes in fluid-rock interactions and rock deformation (Parotidis et al., 2003). Consequently the subject matter of the study of seismo-geochemical precursors has drawn worldwide attention of a large base of scientific community over the last several decades to carry out and experimentally improve gas-geochemical monitoring methods as a means of deriving pre-seismic information (Scholz et al., 1973; Sugisaki, 1978; King, 1980, 1986; Hauksson and Goddard, 1981; Kawabe, 1984, 1987; Barsukov et al., 1984; Satake et al., 1985; Sato et al., 1986; Sugisaki and Sugiura, 1986; Wakita et al., 1988; Thomas, 1988; Reimer, 1990; Nagamine and Sugisaki, 1991; Virk and Baljinder, 1994; Virk, 1995; Sugisaki et al., 1996; Toutain and Baubron, 1999; Heinicke and Koch, 2000; Ito, 2000; Biagi et al., 2000; Virk et al., 2001; Gülec et al., 2002; Planinic et al., 2004; Zmazek et al., 2005; Hartmann, 2005a, b; Yang et al., 2006; King et al., 2006, 2008; Das et al., 2009; Cicerone et al., 2009; Chaudhuri et al., 2010b). Pre- INTRODUCTION “Potential earthquake precursory signals”, often called “anomalies” have been well documented over the last few decades. Anomalies in the gas-geochemical and watergeochemical signals preceding major seismic events have been recognized as short-term earthquake precursors (Talwani et al., 1980; Mogro-Campero et al., 1980; Fleischer, 1981; Hauksson, 1981; Sugisaki, 1981; King, 1986; Heinicke et al., 1995; Igarashi et al., 1995; Igarashi and Wakita, 1995; Tsunogai and Wakita, 1995, 1996; Virk, 1995; Wakita, 1996; Virk et al., 1997; Toutain and Baubron, 1999; Ito et al., 1999; Heinicke and Koch, 2000; Favara et al., 2001; Chyi et al., 2005; Yang et al., 2005; Hartmann, 2005a; Hartmann and Levy, 2005, 2006; Weinlich et al., 2006; Cicerone et al., 2009). Changes in the chemical composition of ground waters, deep earth gases and ascending fluids are linked with seismic activities and could be used as earthquake precursors if the *Corresponding author (e-mail: [email protected]) Copyright © 2011 by The Geochemical Society of Japan. 137 seismic geochemical changes are observed in the ground water (Tsunogai and Wakita, 1995; Favara et al., 2001) and to a greater degree in the mineral spring gases (Weinlich et al., 2006), thermal spring gases (Das et al., 2006a; Chaudhuri et al., 2007; Horiguchi and Matsuda, 2008) and soil gases (Virk and Baljinder, 1994; Hartmann, 2005a, b) monitored in a tectonically sensitive region. Seismic induced gas-geochemical fluctuations are often superimposed on diurnal fluctuations that are caused by the effects of earth’s tides. The criteria adopted to attribute an observed gas-geochemical signal as genetically due to a seismic occurrence, is by observing anomalous characteristics in gas concentration amplitude, form and duration of stay that occur preceding earthquakes (Toutain and Baubron, 1999). Explicit rules to unequivocally distinguish earthquake related signals in a monitoring time series have not been developed as yet. However, a variety of rules to identify and organize earthquake related signals have been proposed by several workers (Dobrovolsky et al., 1979; Kawabe, 1987; Igarashi and Wakita, 1990; Hartmann and Levy, 2005). Concurrently quantitative changes in gas-geochemical time series data exceeding 2σ (σ is the standard deviation of the time series data) spread over durations of several hours to two, three or more days are often considered as precursory indicative parameters and are treated to be of seismic origin (Hauksson and Goddard, 1981; King, 1986). The maximum distance at which geochemical precursors are observed are governed by power law dependant algorithms (Giuseppe et al., 1997). Therefore, it is likely that for earthquakes of magnitude M > 6 the precursory signals may be observed at a monitoring site which is few thousand km away from the epicenter (Dobrovolsky et al., 1979; Fleischer and Mogro-Campero, 1985; Rikitake, 1988; Morgounov and Malzev, 2007; Walia et al., 2005; Cicerone et al., 2009). However, local and regional settings play an important role in influencing and determining the occurrence and pattern of geochemical anomalies. Although geochemical techniques appear to be a prospective tool for earthquake precursor study, however, analysis of geochemical signals related to earthquake still poses a challenging task, since geochemical markers pointing unambiguously to a definite earthquake occurrence, out of long duration data received from a single site, are hard to come by. This is an inherent natural restriction that illustrates the limitation of single site geochemical observation. To get out of this difficulty, in the present paper major focus has been made towards multi-component and multi-station approach to record and analyze the pre-seismic gas-geochemical anomalies. Conventionally, the “one earthquake-one signal” hypothesis is often used to relate the pre-seismic signals with the impending earthquake. However, this approach neglects the continuum of stress-strain changes (Hartmann, 2005a). 138 H. Chaudhuri et al. Based on the “multiple signal-one earthquake” and/or “one signal-multiple earthquake” hypotheses, comprehensive monitoring of multi-component and multi-station data-sets has begun to result in the detection of more reliable precursory signals. Often these proposed precursors simultaneously appear at several monitoring locations and contain two or more observations (Igarashi and Wakita, 1995; Hartmann, 2005a; Hartmann and Levy, 2005). In this paper attempt has been made to verify the feasibility of “multiple signal-one earthquake” and/or “one signal-multiple earthquake” hypotheses. Consequently the aim is to carry out direct inter-comparison and cross-correlation study of the pre-seismic gasgeochemical anomalies recorded simultaneously at two identically equipped monitoring stations placed in totally dissimilar geological environments and in different seismic zones. The choice of monitoring station is a crucial factor for the pre-seismic gas-geochemical monitoring. The essential condition for successful monitoring of preseismic gas-geochemical signal is the appropriate choice of a location for the emplacement of a monitoring station (Weinlich et al., 2006). According to King et al. (2006) the sensitivity of a site to pre-seismic (and post-seismic) change depends upon its proximity to active faults. The monitoring station may be overlying in (i) a thermal spring located in a highly fractured, extensively faulted and tectonically sensitive region (say, geothermal region with high heat flow) or (ii) in the vicinity of an active fault or (iii) directly in a seismically active region. These regions serve as easy escape channels for the deep earth gas emission due to high permeability and porosity of the underlying bed rock and the presence of a large number of gas emanation vents that are deeply rooted, possibly into the in-situ bedrock. Seismic induced variations in thermal spring gases and soil gases are best examined if the time series data are taken in a tectonically sensitive, highly fractured and extensively faulted region, against benefit of background statistics recorded over a long duration (Das et al., 2006a; Chaudhuri et al., 2007). With this in view, we set-up a laboratory at Bakreswar thermal spring in the eastern part of India during December 2004 (Das et al., 2005) to measure continuously hourly concentration changes in He, CH4, 222Rn and gamma dose rate in thermal spring gases. Time series data obtained thereafter revealed several large anomalies that were subsequently correlated to possible major seismic disturbances (local and/or remote) that occurred in and around India (Das et al., 2005, 2006a, 2006b, 2007, 2009; Chaudhuri et al., 2007, 2010b). With an aim to obtain better seismic correlation and to enhance the reliability, we established a second monitoring station at Tatta Pani thermal spring of the north-western part of the country during August 2007. The instrumentation and on-line, continuous [24×7] Fig. 1. Geological settings of Bakreswar (modified after Majumdar et al., 2005) and Tatta Pani (modified after Hassan and Baig, 2006) geothermal areas along with the location of the China earthquake. monitoring as well as data recording facility in both the laboratories are identical. GEOLOGICAL DESCRIPTION OF THE M ONITORING STATIONS The monitoring stations are placed in dissimilar geological environments and in different seismic zones of the country and are separated by a distance of 1612 km. One of the monitoring stations is located at thermal spring site at Bakreswar (23°52′30″ N, 87°22′30″ E) which lies in the eastern part of the country in a moderate risk zone— “seismic zone III”. The other monitoring station is situated at Tatta Pani thermal spring site (33°14′16″ N, 74°24′43″ E) which falls in the north-western part of the country in a high risk zone—“seismic zone IV”. The monitoring spring at Bakreswar lies close to the extinct (115 Ma) Rajmahal volcanism of the Chotanagpur gneissic complex in West Bengal while the spring at Tatta Pani is located in a non-volcanic geothermal area that lies within the mountain folds of Jammu and Kashmir, situated in proximity to the Main Boundary Thrust (MBT) of the Lesser Himalayas. The locations and geological settings of the two monitoring sites are shown in Fig. 1. Bakreswar Bakreswar geothermal area is situated in a geologically mixed faulted area about 250 km north-west of Kolkata and consists of a cluster of seven thermal springs with differing water temperatures (Majumdar et al., 2009). Long range pre-seismic gas-geochemical anomalies 139 The thermal springs lie in an area with Precambrian rocks and is associated with a 1.2 km long shear zone linked to the extinct (115 Ma) Rajmahal volcanic activity (Nagar et al., 1996). The geological conditions of the region maintain a constant flow of hot waters that issue out of fractures in a reactivated composite mass comprising predominantly Precambrian granitic rocks (Majumdar et al., 2000). This geothermal zone has been subjected to different cycles of plate movements with intervening periods of iso-static readjustments during the Precambrian to Cenozoic era (Sarkar, 1982; Ravi Shankar, 1991a). The spring gas along with the soil air in the region is characterized by the presence of high helium as well as high radon flux (Ghose et al., 1998, 2002; Chaudhuri et al., 2010a, b). The Bakreswar geothermal province is characterized by high average heat flow of ~200 mW/m 2 (Chandrasekharam, 2000). This is about three times the recent estimates of the mean heat flow 65 ± 1.6 mW/m2 for all continents and twice for all oceans 101 ± 2.2 mW/ m 2 (Gilat and Vol, 2005). The monitoring spring “Agnikunda” has a water temperature of ~69°C at surface, with water discharge rate of ~790 slpm. Bakreswar thermal spring water is alkaline with a pH value of 9.2. The alkaline water contains few water dissolved minerals like F– (0.17 ppm), Cl– (21.80 ppm), SO42– (26.0 ppm), B (0.21 ppm) and water dissolved 222Rn (38.90 kBq/m3) at ambient temperature and pressure. Spring gases ascend to the earth’s surface through the intricate network of fissures and faults with a gas flux of ~3.5 slpm and at 1.6 bar pressure. Bakreswar spring gas is primarily composed of He (~1.45 vol%), O 2 + Ar (~3.00 vol%), N2 (~92.00 vol%), CO 2 (~0.10 vol%) and CH4 (~3.40 vol%). The spring gas is highly radioactive with an average 222Rn concentration of 828.85 kBq/m 3. The geothermal area connects the eastern edges of two major fault systems that cut across the northern and central parts of India—the “SONATA” (by name of the Indian rivers Sone, Narmada and Tapti) and the “ONGC” (by name of the Oil and Natural Gas Corporation, India) faults (Gupta et al., 1975; Ravi Shankar, 1991a; Majumdar et al., 2005) incising through the central India state of Madhya Pradesh. The crustal thickness at this area is only 24 km as compared to an average of 38 km of the Precambrian craton (Mukhopadhyay et al., 1986; Majumdar et al., 2000; Ghose et al., 2002). Tatta Pani Tatta Pani thermal springs are about 35 km south-east of Rajouri town in the state of Jammu and Kashmir. Tatta Pani is located on the Himalayan Pir-Panjal mountain range and lies in a region of considerable structural complexity that is typical of the highly heterogeneous geotectonics of the Lesser-Himalayas (Wadia, 1928; Gokam et al., 2002). Tatta Pani lies close to the Main 140 H. Chaudhuri et al. Boundary Thrust (MBT) at an altitude of about 800 m above mean sea level. The rocks of this region have ages ranging from Precambrian to recent. There are two thermal springs at Tatta Pani with temperatures of 55°C and 46°C. The higher temperature spring is used by pilgrims for bathing. The monitoring spring temperature is ~46°C with water discharge rate of 360 slpm. The spring water at Tatta Pani is acidic having a pH value of 6.41. The spring water contains F – (2.11 ppm), Cl – (9.90 ppm), SO42– (55.00 ppm), B (0.46 ppm) and water dissolved 222 Rn (46.90 kBq/m 3) at ambient temperature and pressure. Spring gases flow out of the spring vents with a gas flux of ~1.5 slpm and at 1.3 bar pressure. The spring gas comprises of He (~1.15 vol%), O2 + Ar (~3.30 vol%), N2 (~89.50 vol%), CO2 (~4.90 vol%) and CH4 (~0.40 vol%). The 222Rn activity of the spring gas is reasonably high (899.21 kBq/m3). The thermal springs fall in the vicinity of geologically conjectured fault and contact zones of the mountain formation. These thermal springs issue out of the deep rooted faults of Tertiary age in the MBT which lies across the north of Jammu and regionally separates the sandstone conglomerate rock formations known as the “Shiwaliks” from the other sedimentary sequences of the Lesser-Himalayas and the Indo-Gangetic alluvials (Ravi Shankar, 1991b; Hassan and Baig, 2006). The heat flow in this geothermal area is around 180 mW/m 2 (Ravi Shankar, 1988). THEORY ON EMPIRICAL ALGORITHMS A number of empirical algorithms, given in the literature, facilitate to identify the patterns of precursory signals that characteristically occur prior to earthquakes (Scholz et al., 1973; Yoshino et al., 1992; Giuseppe et al., 1997; Toutain and Baubron, 1999; Walia et al., 2005; Cicerone et al., 2009). The empirical algorithms have been proposed on the basis of statistical comparison of precursory data with seismic ones, from an existing data base recorded by means of geochemical, hydrological, seismological and electromagnetic techniques. A few of the existing empirical algorithms also illustrate the relationships between the (i) earthquake magnitude-epicentral distance, (ii) magnitude-precursor time, (iii) magnitudesignal duration, (iv) precursor time-signal duration, (v) precursor time-epicentral distance and (vi) signal duration-epicentral distance (Hartmann and Levy, 2005). The relationship between magnitudes and epicentral distances The preparation area (or the zone of influence) of an earthquake scales directly with the magnitude M. The maximum limit (Rmax) of this zone of influence (or the maximum epicentral distance) is essentially a geodesic distance (in km) that includes the epicenter of the earth- quake and the monitoring stations where significant anomalies can be expected due to seismic activity. The maximum distances at which precursory signals may be observed was estimated by Dobrovolsky et al. (1979) through an empirical algorithm and expressed as ship between precursor time (T) and earthquake magnitude (M) was first suggested by Tsubokawa (1973) entirely on an empirical basis (Rikitake, 1975). Scholz et al. (1973) proposed a similar relationship and expressed as RDob = 100.43M for M ≥ 3 T Sch = 10(0.685M–1.57). (1) (7) where, M (in Richter scale) is the magnitude of the earthquake and R (in km) is the radius of the earthquake preparation area, which essentially is the zone of influence. Hauksson (1981) proposed a relationship as follows The precursor time (T) was estimated by Talwani (1979) through an empirical algorithm and expressed as RHau = 10(M+0.43)/2.4. Subsequently, Rikitake (1987) proposed a relationship between precursor time and earthquake magnitude as (2) The formulation put forward by Rikitake (1988) is given by RRkt = 10 (0.38M+0.33). (3) Virk (1996) proposed the following four relations RVrk = 100.32M when 10 < RVrk < 50 RVrk = 10 0.43M RVrk = 10 0.56M (4b) when 100 < RVrk < 500 (4c) (4d) Relatively recently a multiple-fracture model of preseismic phenomena developed by Morgounov and Malzev (2007) gives RMog = 10 (0.5M–0.27). TRkt = 10[(–0.47±0.73)+(0.28±0.12)M]. (8) (9) Out of the various empirical relations to evaluate precursor time (T), the formulation put forward by Sultankhodzhayev et al. (1980) offers the maximum value Tmax and given by (4a) when 50 < RVrk < 100 RVrk = 100.63M when 500 < RVrk < 1250. TTal = 10[(M+0.07)/2]. TSul,M≥3 = [10(0.63M+0.15)]/R for M ≥ 3 (10a) T Sul,M<3 = [10(0.63M–0.15)]/R for M < 3. (10b) T, M and R are expressed in days, Richter scale and km respectively. Equations (10a) and (10b) give the maximum value Tmax as Tmax,Sul,M≥3 = [10 (0.63M+0.15)]/Rmax,Fls,M≥3 for M ≥ 3 (11a) (5) Tmax,Sul,M<3 = [10 (0.63M–0.15)]/Rmax,Fls,M<3 for M < 3 (11b) Out of the various empirical relations that connect M to R, the equation where R max,Fls confers the largest value to the radius of influence and can be estimated using equations (6a) for M ≥ 3 and (6b) for M < 3. Relatively recently, Hartmann and Levy (2005) proposed an empirical formula to evaluate the maximum precursor time, expressed as Rmax,Fls,M≥3 = 10 0.480M for M ≥ 3 (6a) Rmax,Fls,M<3 = 100.813M/16.6 for M < 3 (6b) and put forward by Fleischer (Fleischer and Mogro-Campero, 1985; Fleischer, 1981) confer the largest value (Rmax) to the radius of influence (R). The relationship between magnitudes, precursor time and epicentral distances The precursory time of a pre-seismic signal is the time interval between the recorded anomaly at a station and the actual occurrence of an earthquake. A linear relation- T max,Hrt = 43.6975M – 0.3046R – 32.7731 (12) where T max is the maximum precursor time in days, M is the magnitude of the earthquake in Richter scale and R is the epicentral distance in km. EXPERIMENTAL TECHNIQUE Thermal springs are a steady source of hot water and gases that bubble intermittently out of various places at the base of the spring. The spring pool is 2.5 m deep. In Long range pre-seismic gas-geochemical anomalies 141 order to minimize the influence of meteorological parameters and associated noise in the data-sets thermal spring gases are trapped under a column of hot water. The method is a potential way to avoid masking of weak gasgeochemical variations by various kinds of signals come from the earth’s tidal effect and associated noise. Consequently a stainless steel (SS) funnel about 1m diameter and 0.5 m height is placed inverted under water at the location where maximum gas bubbles are expelled. Gas trapped under the funnel is led through a stainless steel pipe into the monitoring laboratory located at a distance within 5 m from the spring edge. The moisture laden gas stream is led into a series of glass traps filled with chemical desiccants (KOH and anhydrous CaCl2). The dry gas leaving the traps is directly led to a set of instruments that have comprehensive measurement functions. The experimental hardware section consists of two low power automated microprocessor controlled field-deployable equipment—(i) a gas chromatograph and (ii) a radon monitor. Both instruments operate on-line in continuous mode. The schematic diagram of the installed experimental arrangement is shown in Fig. 2. A programmable gas chromatograph type Varian CP 4900 (manufactured by Varian Inc., The Netherlands), provided with narrow bore 20 m long 15N silica capillary column coated with 5A molecular sieve and micro thermal conductivity detector, analyses the spring gas for its stable constituents such as He, O2 + Ar, N2 and CH4. Ultra pure (>99.998 vol%) hydrogen gas is used as the carrier gas in the gas chromatograph. The setting of the sampling interval (1 hour) and data acquisition of the chromatograph is controlled by the on-line chromatographic data managing software “Star Workstation”. The gas chromatograph column is periodically regenerated to maintain the separation efficiency in the chromatograph column. A radon monitor type Alpha-GUARD PQ2000 PRO (manufactured by Saphymo, GmbH; formerly Genitron GmbH, Germany) registers the spring gas radon ( 222Rn) concentration. This instrument consists of an open type pulse ionization chamber and uses alpha spectroscopy to detect 222Rn. The two common isotopes of radon (222Rn and 220Rn) are discriminated through energy-specific windows. The signal generated from 222Rn is converted to a digital output and the data are recorded on an hourly basis. The radon monitor can determine concentrations of 222 Rn from 2 Bq/m3 to 2 × 106 Bq/m3 with a resolution of 1 Bq/m 3. On account of high radon activity level in the spring gas, this instrument is yearly sent to the manufacturer to remove contamination and for recalibration. Besides measuring 222Rn, this instrument is equipped with an unshielded Geiger-Muller (GM) counter which records the total gamma ray activity on account of the combined effects of (a) short-lived gamma emitting 222Rn decay 142 H. Chaudhuri et al. Fig. 2. Integrated experimental set-up of the automated, continuous and online monitoring system to detect gas-geochemical precursory signals. products 214Pb and 214 Bi, present in the spring gas, (b) terrestrial gamma background radiation generated by the decay of β emitting decay products 214Bi, 208Tl and 40K of the three primordial radio nuclides viz. U, Th and K, respectively, present in the rocks and soils of the spring area and (c) gamma environment in proximity of the spring primarily on account of the radioactivity of the mineral waters due to the presence of dissolved radon and radium salts. The Alpha-GUARD registers the gamma dose rate in units of nSv/hr. The 222Rn concentration and gamma dose rate data recorded in the Alpha-GUARD is automatically logged on the control computer using the on-line and automatic data management professional software “Data-Expert”. To produce analytically consistent data-sets, both monitoring equipment are configured to perform with low influence of external noise. Numeric data received from the installed instruments are automatically stored in a local server computer (IBM make). A back-up power pack supply ensures continuous and stable power supply in case of a power failure. To carry out proximate analysis of the time series, all data stored in the server computer installed at field laboratory are transmitted automatically by means of secured broad band wireless internet connectivity (VSAT, ERNET India) to our centre at Variable Energy Cyclotron Centre (VECC) and Saha Institute of Nuclear Physics (SINP) situated in the same campus at Kolkata to prepare update Fig. 3. Helium time series plot of Bakreswar thermal spring recorded during the six month period of September 15, 2007 to March 15, 2008. geochemical precursory signals databank. A seismological data sheet is also prepared routinely by collecting diurnal seismic updates from the India Meteorological Department (IMD) and the US Geological Survey (USGS) website reports (IMD & USGS websites) for seismic events that occurred in India and its neighborhood. RESULTS The hourly data of He, 222Rn and gamma dose rate time series span six month from September 15, 2007 to March 15, 2008. Figure 3 shows the helium concentration time series recorded at Bakreswar during this period. The average helium abundance was 1.38 vol%. The helium time series show a fluctuating pattern having standard deviation of 0.11. In Fig. 4 we show concurrent spring gas 222 Rn and gamma activity patterns recorded at Bakreswar and Tatta Pani during the same time period. The mean 222 Rn concentrations at Bakreswar and Tatta Pani thermal springs were 828.85 kBq/m3 and 899.21 kBq/ m3, respectively. Throughout the entire monitoring period temperatures of the monitoring springs at Bakreswar and Tatta Pani appeared to be constant and were well above the maximum value (40°C) of ambient temperatures recorded during the said six month monitoring period. Besides, the hydrostatic pressures at which the bubbling gases emerge from the spring vents are 1.6 bar and 1.3 bar for Bakreswar and Tatta Pani, respectively which are above the atmospheric pressure. Therefore, meteorological parameters have little effect, if any, on the diurnal as well as seismic induced variations of the monitoring spring gases. The nature of the time series data-sets recorded at both monitoring stations reveal an oscillating mode for 222Rn and gamma activity. Figure 5 shows the temporal variations of helium, 222 Rn concentrations and gamma dose rates monitored at both the monitoring springs during the one-month period of December 14, 2007 to January 18, 2008. Noteworthy fluctuations in helium, 222Rn concentrations and gamma dose rates were observed at Bakreswar laboratory during December 22–27, 2007. Comparison of data from identical measurements recorded at the spring at Tatta Pani displayed similar anomalies for 222Rn and gamma at about the same time period. The logged time series data at Bakreswar show a large helium fluctuation during December 22–24, 2007 as seen in the lower window of Fig. 5. Starting December 22, the helium concentration scaled Long range pre-seismic gas-geochemical anomalies 143 Fig. 4. Time series plots of 222Rn concentrations and gamma dose rates recorded at Bakreswar and Tatta Pani thermal springs during the six month period of September 15, 2007 to March 15, 2008. to 1.96 vol% against its six-month baseline value of 1.38 vol% and remained at the elevated level for over 44 hours. This concentration rise is 42%, and is well above 2σ . The helium concentration profile for the corresponding pe144 H. Chaudhuri et al. riod at Tatta Pani could not be recorded on account of a temporary failure of the on-line gas chromatograph installed there. The recorded data for 222Rn concentrations and gamma Fig. 5. Temporal variations of helium, 222Rn concentrations and gamma activities of Bakreswar and Tatta Pani thermal springs monitored during the one month period of December 14, 2007 to January 18, 2008 with significant anomalies observed during December 22–27, 2007 prior to the China earthquake (M = 6.3) that occurred on January 9, 2008. Long range pre-seismic gas-geochemical anomalies 145 Fig. 6. Pre-seismic anomalies in 222Rn concentrations and gamma dose rates recorded at Bakreswar and Tatta Pani in the period of December 24–27, 2007. 146 H. Chaudhuri et al. dose rates during the same time period, as seen in the upper window of Fig. 5, at both Bakreswar and Tatta Pani show prominent fluctuations. To focus more closely on the anomalous part of the total time series we have taken a short-time period of seven days from December 23–29, 2007 during which significant anomalies were recorded as shown in Fig. 6. Characteristic concentration of 222Rn at Bakreswar scaled up to 888.00 kBq/m3 on December 24, 2007 over its resident mean of 818.38 kBq/m3. The anomaly remained for about 16 hours. A similar pattern of increase in 222Rn, but having a time delay by about 20 hours was observed at Tatta Pani on December 25, 2007. The broad time scale (~18 hours) anomaly in 222Rn concentration at Tatta Pani was 945.00 kBq/m3 against its mean 888.13 kBq/m3. Besides 222Rn, well-defined gamma dose rate anomalies were recorded at both sites during the same time span. Anomalous levels for 222Rn as well as gamma dose rate at the two sites were in excess of 2σ with occupying durations between one to two days. An approximate 24-hour time lag in gamma dose rate anomaly with respect to the occurrence of radon anomaly was noted at each of the monitoring stations. ANALYTICAL TECHNIQUE In order to assign the probable seismic events that sourced the observed spring gas perturbations, we take all the 20 earthquakes (magnitude > 2 M) that occurred in India and its neighborhood during the two month time period December 1, 2007 to January 31, 2008 encompassing a time of about two weeks on either side of the time window (December 14, 2007–January 15, 2008) during which anomalous fluctuations were recorded. The regional occurrences of the listed earthquakes were within a radius of 3500 km from the center of the line joining the monitoring sites. Table 1 shows the seismological data (earthquake magnitude, depth, location and date of occurrence) collected from the India Meteorological Department (IMD) and the US Geological Survey (USGS) website reports (IMD & USGS websites) for the given earthquake events. Table 1 also includes precursory time in days T B and TT for Bakreswar and Tatta Pani, respectively. The geodesic distance in km, which essentially is epicentral distance between the earthquake epicenters and the monitoring stations at Bakreswar and Tatta Pani, are given in Table 1 (DB and DT respectively in our case) as well. We consider the possibility whether the observed gasgeochemical anomalies could have arisen on account of any one, or a combination thereof, of the earthquakes listed in Table 1, to determine the possible existence of a direct relationship of the recorded anomalies to the listed earthquake events. Therefore, based on the experimental data recorded at Bakreswar and Tatta Pani we make a time series analysis to relate magnitude and epicentral distances and precursor time through various empirical relationships mentioned earlier. Among these, Eqs. (6a), (6b), (11a) and (11b) spell the conditions whether an impending earthquake of a certain magnitude can be viewed against an occurring geochemical anomaly at a thermal spring. These equations are also valid for intra-plate earthquakes and long distance seismic events (Lomnitz and Lomnitz, 1978; Fleischer, 1981; Igarashi and Wakita, 1990; Virk, 1996; Shalimov and Gokhberg, 1998; Das et al., 2009). To make a comparative study of the (i) maximum epicentral distance Rmax and (ii) maximum precursor time Tmax observed at the two monitoring sites dispersed at locations about 1612 km apart, we employ the theoretical expressions (6a), (6b), (11a) and (11b). Accordingly, we contest the maximum radius of influence R max (=Rmax,Fls) and the maximum precursor times T max (=T max,Sul) of all the 20 earthquakes given in Table 1 against the possibility whether any of the listed seismic events could have generated an anomaly at any one or both the Bakreswar and Tatta Pani thermal springs. We put forward the argument that an impending earthquake will generate anomalies in spring gas concentrations provided the distance between the monitoring spring and the earthquake epicenter (D B and DT respectively in our case) is approximately close to or less than Rmax i.e., (i) D B and/or DT ~ Rmax or (ii) D B and/or DT ≤ Rmax. Table 1 also shows DB and DT for the listed earthquakes vs. corresponding calculated maximum radius of influence Rmax. D ISCUSSION In this section we have discussed (a) the causes to the oscillating nature of the spring gas, (b) the reasons behind the significant anomalies, (c) the seismogeochemical mechanism, (d) feasibility of the (i) “one signal-one earthquake”, (ii) “multiple signal-one earthquake” and (iii) “one signal-multiple earthquake” hypotheses and (e) the validity as well as the effective range of interaction of the existing empirical algorithms by analyzing the recorded gas-geochemical precursory signals using two distant reference sampling locations at Bakreswar and Tatta Pani. Diurnal variations As seen in Figs. 3 and 4 diurnal variations occur in the monitored helium, 222Rn concentration and gamma activity time series data sets independent of the seasons and the weather conditions. Gas-geochemical parameters recorded at Bakreswar and Tatta Pani revealed fluctuating patterns from daily minima to daily maxima. However, these fluctuations were below 2σ. During the monitoring period the helium concentration at Bakreswar scaled upto daily maximum value of 1.61 vol% from mini- Long range pre-seismic gas-geochemical anomalies 147 148 H. Chaudhuri et al. Occurrence Location date Latitude; Longitude Magnitude (M) (in Richter Scale) December 14, 2007 December 14, 2007 December 22, 2007 6 7 8 Sumatra Indonesia Koyna, Maharashtra India Nanded, Maharashtra India India (Kashmir) −China Border 35.4°N 77.2°E 19.3°N 77.7°E 17.1°N 73.7°E 1.5°N 98.0°E 5.8 2.9 2.7 5.3 2738 10.0 January 04, 2008 January 05, 2008 January 09, 2008 January 12, 2008 January 14, 2008 12 13 14 15 16 Andaman Islands India India (Mizoram) −Bangadesh Border China N. Sumatra Indonesia S. Sumatra Indonesia 10.4°N 92.5°E 22.7°N 92.6°E 5.8 5.0 6.3 84.8°E 5.4 94.8°E 32.0°N 6.0 5.9°S 103.8°E 5.3°N 33.0 33.0 96.0 28.6 15.0 21.0 19.0 16.0 12.0 11.0 1593 549 938 2213 3760 1783 1610 5.0 8.0 1122 5.0 1747 1613 8.0 747 33.0 2233 2123 188.7 4.0 3179 33.0 35.0 2803 Distance between Bakreswar and epicenter (DB) (in km) Precursor time (TB) (in days) 10.0 Depth (d) (in km) Seismo-geochemical features at Bakreswar 20.0 18.0 15.0 11.0 10.0 7.0 7.0 3.0 Precursor time (TT) (in days) 3141 2131 945 3755 5342 1420 1390 3805 4302 1796 1584 352 2255 510 4746 4367 Distance between Tatta Pani and epicenter (DT) (in km) Seismo-geochemical features at Tatta Pani Anomalous fluctuations (>2σ) observed during Dec. 1, 2007−Jan. 29, 2008 Bakreswar: Dec. 24−26, 2007 (>2σ); Tatta Pani: Dec. 25−27, 2007 (>2σ); (Fluctuations >> 2σ) 9 December N. Sumatra 5.4°N 5.2 2.7 28, 2007 Indonesia 95.8°E 10 January Junagarh, Gujarat 21.2°N 3.5 10.5 01, 2008 India 70.6°E 11 January Junagarh, Gujarat 21.0°N 3.6 5.0 01, 2008 India 70.3°E December 11, 2007 5 Bakreswar: No anomaly (~2σ); Tatta Pani: No anomaly (~2σ); (Fluctuations << 2σ) 1 December N. Sumatra 1.9°N 5.9 01, 2007 Indonesia 98.3°E 2 December S. Sumatra 1.9°S 5.3 02, 2007 Indonesia 100.1°E 3 December Hindukush 37.0°N 4.8 02, 2007 Afganisthan 71.2°E 4 December Myanmar 23.5°N 5.0 07, 2007 94.7°E Event No. Earthquake data obtained from IMD & USGS website 608 251 1057 391 759 54 48 313 608 14 10 350 251 201 350 680 Maximum radius of Rmax,Fls (in km) 10.47 7.94 12.45 9.12 11.22 4.89 4.73 8.51 10.47 3.47 3.76 8.81 7.94 7.41 8.81 10.84 Maximum precursor time Tmax,Sul (in days) Theoretical estimated values of the seismic events Table 1. Seismo-geochemical features (precursor time T and epicentral distance D) of the recorded precursory signals and seismological data of the earthquakes that occurred in and around India during the period December 1, 2007 to January 29, 2008 along with the theoretical estimated values of the parameters—maximum radii of influence R max and maximum precursor time Tmax India (Arunachal Pradesh) −China Border January 29, 2008 20 Gajapati, Orissa India January 27, 2008 19 Rudraprayag, Uttaranchal India January 25, 2008 18 The shaded row indicates the earthquake that has been correlated with the observed geochemical anomalies recorded at geochemical monitoring laboratories at Bakreswar as well as at Tatta Pani. 6.03 104 4.2 10.0 14.0 690 1864 4.57 43 3.4 15.0 12.0 623 1790 4.73 48 1101 10.0 13.9 3.5 2736 7.0 6.0 1.3°N 97.5°E 30.6°N 79.2°E 19.0°N 84.4°E 27.7°N 92.8°E N. Sumatra Indonesia January 22, 2008 17 Bakreswar: Jan. 15−17, 2008 (<2σ); Tatta Pani: No anomaly; (Fluctuations < 2σ or ~2σ) 10.0 Depth (d) (in km) Magnitude (M) (in Richter Scale) Latitude; Longitude Occurrence Location date Event No. 572 11.22 759 4291 Maximum precursor time Tmax,Sul (in days) Maximum radius of Rmax,Fls (in km) Distance between Tatta Pani and epicenter (DT) (in km) Precursor time (TT) (in days) Distance between Bakreswar and epicenter (DB) (in km) Precursor time (TB) (in days) Theoretical estimated values of the seismic events Seismo-geochemical features at Tatta Pani Anomalous fluctuations (>2σ) observed during Dec. 1, 2007−Jan. 29, 2008 Seismo-geochemical features at Bakreswar Earthquake data obtained from IMD & USGS website Fig. 7. The power spectrum of the helium time series recorded at Bakreswar during the one month period of December 14, 2007 to January 18, 2008. mum value of 1.35 vol% with a mean value of 1.38 vol%. The 222Rn concentrations recorded at Bakreswar and Tatta Pani showed the similar diurnal variations from daily maxima of 872 kBq/m 3 and 888 kBq/m 3 from daily minima of 782 kBq/m3 and 790 kBq/m3 with mean values of 828 kBq/m3 and 848 kBq/m3 for Bakreswar and Tatta Pani, respectively. The fluctutaions (<2σ ) observed are in part short term oscillations caused by changes in lithological strain pattern of the confining rock masses owing to aquifer dilation due to earth’s tides (Sugisaki, 1981). Figure 7 shows the power spectrum of helium time series recorded at Bakreswar during the one month period December 14, 2007–January 15, 2008. The power spectrum exhibits relative maxima close to periods of 24 hr (and small indications at 12 hr), which coincide with the earth’s tide constituents (e.g., O1, S1, OO1, M2 and S2). The O1, S1, OO1, M2 and S2 are the Darwin symbol of the different frequencies (in cpd) of the diurnal and semi-diurnal earth’s tides; O1 = 0.9295357, S1 = 1.0000000, OO1 = 1.0759401, M2 = 1.9322736 and S2 = 2.0000000 (Hartmann and Wenzel, 1995; Kudryavtsev, 2004). Weinlich et al. (2006) reported similar S1, P1, J1, O1, OO1, S2 and M2 peaks in the 222Rn and CO2 spectra during 222Rn and CO2 measurements in mineral springs. A similar prominent S1 peak in the radon spectrum of radon measurements inside a gypsum mine was also reported by Millich et al. (1999). Therefore, we may conclude that the 24 hr cycles observed in the helium data are based on earth’s tides. Barnet et al. (1997) also observed a correlation between diurnal 222Rn variation and earth’s tide in mines. Gas-geochemical time series data are further complicated by the influence of meteorological parameters, rainfall, tectonic processes and other factors (Binchard and Libby, 1980; Sugisaki, 1981; Igarashi and Wakita, 1991; Finkelstein et al., 1998). As mentioned earlier meteorological parameters have little effect on the fluctuating pattern of the Bakreswar and Tatta Pani thermal spring gas emanation and therefore the oscillating pattern observed in the time series data is due to intrinsic influence of the tidal effect. However, if fluctuations Long range pre-seismic gas-geochemical anomalies 149 (<2σ) of state of stress caused by tides affect the deep earth fluids, a complete coincidence between tide and gas geochemical data does not necessarily have to exist. The stress caused by tectonic movements has to be considered in addition, and that will affect the gas emanation process. Beyond the diurnal variations (<2σ ) caused by earth’s tides, anomalous changes (>2 σ) in the He, 222Rn concentrations and gamma activity occur which clearly exceed these variations and are thought to be of seismic origin (Weinlich et al., 2006). Seismic induced signals Several distinct fluctuations (>2σ) beyond the diurnal variations (<2σ ) in the helium and 222Rn time series occur at both the monitoring stations as shown in Figs. 5 and 6. These anomalies cannot be explained by the tidal effects and/or by the changes in meteorological parameters. These alterations inevitably reveal stress variations caused by the seismic activities. In most cases these anomalies are noteworthy and may be accompanied by local seismic activities and/or by the seismic signals of remote earthquakes (Weinlich et al., 2006). Although a remote earthquake will not result in a long-term redistribution of stress, it can cause a pressure-pulse on the deep earth fluid systems (Weinlich et al., 2006). Pressure changes at the Aigio fault, Gulf of Corinth, Greece induced by remote earthquakes (Rat Islands, 2003; Sumatra, 2004) were reported by Doan and Cornet (2005) which clearly indicate an effect on the state of stress even in remote fault systems. This coincidence indicates a relationship between the changes in the gas-geochemical data and the distant seismic activities. Pre-seismic excess release of helium and radon are caused by the changes in tectonic stress-strain pattern during the critical phase preceding an earthquake event. This phase is accompanied by permeability changes due to micro-crack formation in rocks prior to the seismic event. The anomalous fluctuations observed may have resulted from (i) relative increase in heat flow that precedes earthquakes, (ii) stress-induced pore-collapse and (iii) stress-induced micro-fracturing. The observed effect could be attributed to any one of the three or a possible combination thereof (Monnin and Seidel, 1988; Ghose et al., 1996; Chaudhuri et al., 2010b). Rock masses usually hold systems of overlapping cracks and their interconnections in a three dimensional system. Elastic deformation due to hydrostatic compression influences the process of helium and radon release. Effectively large transformations in the gas concentrations are known to occur prior to major seismic events in response to alterations in rock properties, changes in aquifer chemistry and inputs from mantle sources (Kennedy et al., 1997). The build up of stresses preceding an earthquake or their release thereafter can cause changes in the strain field within the 150 H. Chaudhuri et al. Fig. 8. Earthquake magnitude (M) vs. the theoretically derived maximum precursory time (T max) and experiential precursor times TB and TT recorded at the two monitoring sites Bakreswar and Tatta Pani, respectively. earth. Because the crust largely acts as an elastic continuum the stress field domain may be large (Ouchi et al., 1985). When the ground is squeezed, the pore-spaces are narrowed non-linearly resulting in extrusion of helium, radon and other deep earth gases (Fleischer, 1981). As mentioned earlier the observed gamma ray activity is due to combined effects of (a) gamma radiation present in the spring gas, (b) terrestrial gamma background radiation and (c) ambient gamma radiation. Each of the above contributory gamma source is responsive to the earth stress changes (Tsvetkova et al., 2001; Karangelos et al., 2002; Steinitz et al., 2003; Das et al., 2006a). Accordingly the registered gamma activity would be roughly synchronized with 222Rn and helium changes as seen in Fig. 5. During the anomalous fluctuations period December 22–24, 2007 an approximate 24 hour time lag was observed in both monitoring stations in gamma dose rate anomaly with respect to the occurrence of radon anomaly. This is purely an experimental observation and may have a link with the geological disposition of the two monitoring sites. During the same time period a time delay of 20 hours was also observed in the appearance of the radon anomaly at Tatta Pani as compared to that at Bakreswar. The time delay may be the outcome of natural geological inhomogeneities, differences in the heat flow rate and regional settings at the two test sites that affect strain propagation and consequently spring gas emanation response. Fig. 9. Overlapping circles of radii increasing by 150 km with the monitoring sites as the center showing in the earthquake risk zone map of India. Monitoring observations Table 1 shows that magnitude distribution (M; in Richter scale) of the listed earthquakes that occurred at depth 2.7 km to depth about 188 km during the referred twomonth period, ranged between 2.7 M to 7.4 M covering a spatial area within latitude 01.0°N to 26.3°S and longitude 139.4°E to 177.4°W. The values of Rmax and Tmax vary between 10 km to 1057 km and 3.47 days to 12.45 days, respectively in our case as shown in Table 1. Out of the 20 earthquakes there is only one instance (event No. 14 in our case) the China earthquake of magnitude 6.3 M (January 9, 2008), where Rmax (=1057 km) exceeded DB (=938 km) as well as D T (=945 km) simultaneously. The remaining earthquakes have DB and/or D T >> Rmax which violates our given condition. Precursor signal of this China earthquake was clearly recorded as gas-geochemical anomalies (helium, 222Rn and gamma dose rate) from both the springs, as shown in Fig. 5. The locations of the monitoring stations Bakreswar, Tatta Pani and the epicenter (32°00′ N; 84°48′ E) of the said 6.3 M China earthquake is shown in Fig. 1. In the case of event No. 5, the 5.3 M earthquake in India (Kashmir)–China border (December 11, 2007), Rmax (350 km) ~ DT (352 km) but << D B (1613 km). No noticeable concentration change (>2σ ) either for helium or radon was observed at either of the monitoring sites one week before or after the said event. In the case of event No. 20 the 4.2 M earthquake in India (Arunachal Pradesh)–China border (January 29, 2008), an evident anomaly but somewhat lower than 2σ did take place as seen in Figs. 3 and 4. Helium as well as 222Rn anomalies (~2σ) indeed occurred during the period January 15–17, 2008 at Bakreswar which was absent at Tatta Pani. The Long range pre-seismic gas-geochemical anomalies 151 calculations show that the theoretically derived precursory time Tmax falls appreciably short of the experiential precursor times TB and T T recorded at the two monitoring sites Bakreswar and Tatta Pani, respectively. Further, all the experiential precursor times TB and TT do not exactly fall on the M-T curve (for earthquakes M > 3) of the magnitude (M) vs. precursor time (T) graphical representation are plotted against the empirical relationships (6a) and (11a) as shown in Fig. 8. It seems that the equations applied in deriving the theoretical values for Tmax need a revision. CONCLUSION Correlations between gas-geochemical anomalies and distant earthquakes have attracted renewed interest in India following large scale disasters caused by some major earthquakes in recent years. The helium, 222Rn and gamma dose rate anomalies were registered during December 22–27, 2007 in India at two widely separated (1612 km) and extensively faulted geothermal regions prior to the 6.3 M earthquake that occurred in China on January 9, 2008. This suggests that nucleation effects in the form of pre-seismic changes of large earthquake events (>6 M) may occur at tectonically sensitive regions that may be considerably far away (~1000 km) from the eventual epicenter of that earthquake. In the present case, the estimated maximum radius of influence is large enough (~1057 km) and the two monitoring thermal springs at Bakreswar and Tatta Pani, which lie in different geological areas, fall within the theoretically calculated precursory zone of the mentioned China earthquake. The observations may be interpreted as evidence that the gasgeochemical time series data monitored at a thermal spring located in a fault network plays a significant role to monitor the pre-seismic phase even for a distant earthquake, as it moves towards criticality. However, the results presented here are quite preliminary but they do indicate that the area of precursory regions may be large for large earthquakes. Therefore, pre-seismic geochemical signals for strong earthquakes (M > 6) may be identified by means of suitable sensitive instrumentation even if they are positioned at distant (~1000 km) monitoring stations. The given data are results of a time series measurement obtained from two stations. Addition of other such monitoring centers along with long term data recording is expected to bring about greater correlations between geochemical data sets and seismicity. Further, the “one earthquake-one signal” hypothesis which is frequently used in earthquake research, neglects the continuum of stress-strain changes (Hartmann, 2005a). It is therefore necessary to adopt the “multiple signal-one earthquake” and/or “one signal-multiple earthquake” hypotheses by installing a network of multi-component monitoring sta- 152 H. Chaudhuri et al. tions along with a simultaneous analysis of the multicomponent and multi-station data. The geochemical signals of a definite seismic event monitored in more than one station with a possible time delay could provide an understanding of the effects of seismic shocks on the gas bearing fluid systems. With this view in mind, we have more recently, set-up a third monitoring laboratory at the Baratang mud volcano (12°07′44″ N; 92°47′29″ E) in the Andaman and Nicobar Islands, India, having identical instrumentation as at Bakreswar/Tatta Pani. Figure 9 shows the overlapping and concentric circles drawn upon the earthquake risk zone map of India with radii increasing by 150 km, taking into consideration the above three monitoring sites as individual centers. Earthquake preparation activity that occurs within the overlaps may induce similar variations in gas concentrations in each site. A similar concept was proposed by Hartmann and Levy (2005) earlier. It is expected that such reasoning would facilitate better analytical correlation and reinforce conclusions on time series data derived from all the three laboratories relating pre-seismic gas-geochemical anomalies with regional earthquakes. Besides, a systematic approach for analysing earthquake precursory signals and a standardised reporting scheme may be established. It is also suggested that the empirical algorithms applied in deriving the theoretical value for maximum epicentral distance and maximum precursor time need revision. This in turn, may help identify earthquake precursory signals in a more reliable way and may assist in better understanding the earthquake preparation phenomena. Acknowledgments—The authors would like to express their appreciation to the Department of Atomic Energy (DAE), the Department of Science and Technology (IS-STAC, DST), the Ministry of Earth Sciences (Seismology Division, MoES), Government of India, for financial support. The support of the University of Kashmir, Srinagar is highly appreciated, for extending organizational help at the Tatta Pani field laboratory, Jammu (J&K). The authors would like to thank the reviewers for their critical and constructive comments and suggestions. Special thanks are expressed to Dr. Takuya Matsumoto for his careful reading and suggestions for improving the clarity and readability of the manuscript. Finally the authors would like to acknowledge the assistance of Prof. A. N. Sekar Iyengar in improving the power spectrum plot. REFERENCES Barnet, Y., Kies, A., Skalský, L. and Procházka, J. (1997) Radon variations and Earth tides in unventilated underground spaces. Bull. Czech Geol. Surv. 72(2), 105–114. Barsukov, V. L., Varshal, G. M. and Zamokina, N. S. (1984) Recent results of hydrogeochemical studies for earthquake prediction in the USSR. Pure Appl. Geophys. 122(2–4), 143– 156. Biagi, P. F., Ermini, A., Cozzi, E., Khatkevich, Y. M. and Gordeev, E. I. (2000) Hydrogeochemical precursors in Kamchatka (Russia) related to the strongest earthquakes in 1988–1997. Nat. Haz. Earth Sys. Sci. 21(2–3), 263–276. Birchard, G. F. and Libby, W. F. (1980) Soil radon concentration changes preceding and following four magnitude 4.2– 4.7 earthquakes on the San Jacinto fault in Southern California. J. Geophys. Res. 83(B6), 3100–3106. Chandrasekharam, D. (2000) Geothermal energy resources of India: country update. Proc. World Geothermal Congress, Kyushu-Tohoku, Japan, 133–145. Chaudhuri, H., Das, N. K., Bhandari, R. K., Ghose, D., Sen, P. and Sinha, B. (2007) Radon and helium fluctuations prior to seismic events in thermal spring gas. Proc. Electromagnetic Phenomenon Related to Earthquakes and Volcanoes (Singh, B., ed.), Narosa Publishing House, 147–155, ISBN: 978-81-7319-858-8. Chaudhuri, H., Das, N. K., Bhandari, R. K., Ghose, D., Sen, P. and Sinha, B. (2010a) Radon activity measurements around Bakreswar thermal springs. Radiat. Meas. 45, 143–146. Chaudhuri, H., Ghose, D., Bhandari, R. K., Sen, P. and Sinha, B. (2010b) The enigma of helium. Acta Geod. Geophys. Hung. Dec. (to be published). Chyi, L. L., Quick, T. J., Yang, T. F. and Chen, C. H. (2005) Soil gas radon spectra and earthquakes. Terr. Atoms. Oceanic Sci. 16, 763–774. Cicerone, R. D., Ebel, J. E. and Britton, J. (2009) A systematic compilation of earthquake precursors. Tectonophysics 476, 371–396. Das, N. K., Bhandari, R. K., Ghose, D., Sen, P. and Sinha, B. (2005) Anomalous fluctuation of radon, gamma dose and helium emanating from thermal spring prior to earthquake. Curr. Sci. 89(8), 1399–1404. Das, N. K., Choudhuri, H., Bhandari, R. K., Ghose, D., Sen, P. and Sinha, B. (2006a) Continuous monitoring of 222Rn and its progeny at a remote station for seismic hazard surveillance. Radiat. Meas. 41, 634–637. Das, N. K., Bhandari, R. K., Ghose, D., Sen, P. and Sinha, B. (2006b) Explosive helium burst in thermal spring emanations. Appl. Radiat. Isot. 64, 144–148. Das, N. K., Bhandari, R. K., Ghose, D., Sen, P. and Sinha, B. (2007) Imprints of seismic activity on helium and radon fluctuations in spring emanations. Proc. Geochemical Precursors for Earthquakes, Advanced Research Series (Sen, P. and Das, N. K., eds.), 12–21, Macmillan India Ltd. Das, N. K., Bhandari, R. K., Ghose, D., Sen, P. and Sinha, B. (2009) Significant anomalies of helium and radon ahead of 7.9 M China Earthquake. Acta Geod. Geophys. Hung. 44, 357–365. Doan, M. L. and Cornet, F. H. (2005) Hydraulic anomalies triggered close to an active fault by far field earthquakes. Geophys. Res. Abst. 7, EGU 05. Dobrovolsky, I. P., Zubkov, S. I. and Miachkin, V. I. (1979) Estimation of the size of earthquake preparation zones. Pure Appl. Geophys. 117, 1025–1044. Favara, R., Grassa, F., Inguaggiato, S. and Valenza, M. (2001) Hydrogeochemistry and stable isotopes of thermal springs: earthquake erelated chemical changes along Belice Fault (Western Sicily). Appl. Geochem. 16, 1–17. Finkelstein, M., Brenner, S., Eppelbaum, L. and Ne’Eman, E. (1998) Identification of anomalous radon concentrations due to geodynamic processes by elimination of Rn variations caused by other factors. Geophys. J. Int. 133, 407–412. Fleischer, R. L. (1981) Dislocation model for radon response to distant earthquakes. Geophys. Res. Lett. 8, 477–480. Fleischer, R. L. and Mogro-Campero, A. (1985) Association of subsurface radon changes in Alaska and the northeastern United States with earthquakes. Geochim. Cosmochim. Acta 49, 1061–1071. Ghose, D., Das, N. K. and Sinha, B. (1996) Anomalous helium emission: Precursor to earthquakes. Curr. Sci. 71, 56–58. Ghose, D., Paul, D. and Sastri, R. C. (1998) Radon geochemical anomaly at Bakreswar thermal springs. Radiat. Phys. Chem. 51(4–6), 613–614. Ghose, D., Chowdhury, D. P. and Sinha, B. (2002) Large scale helium escape from earth surface around Bakreswar, Tantloi geothermal area in Birbhum District, West Bengal and Dumka district, Jharkhand, India. Curr. Sci. 82(8), 993–996. Gilat, A. and Vol, A. (2005) Primordial hydrogen-helium degassing, an overlooked major energy source for internal terrestrial processes. HAIT J. Sci. Eng. 2, 125–167. Giuseppe, E., Massimo, C. and Fedora, Q. (1997) Seismological algorithms for earthquake predictions: an overview. Annali di Geofisica. XL(6), 1483–1492. Gokam, S. G., Rao, C. K. and Gupta, G. (2002) Crustal structure in the Siwalik Himalayas using magnetotelluric studies. Earth Planets Space 54, 19–30. Gülec, N., Hilton, D. R. and Mutlu, H. (2002) Helium isotope variations in Turkey: relationship to tectonics, volcanism and recent seismic activities. Chem. Geol. 187, 129–142. Gupta, M., L., Narain, H. and Saxena, V. K. (1975) Geochemistry of thermal waters from various geothermal provinces of India. Proc. of the Grenoble Symposium, August 1975—International Association of Hydrological Sciences 119, 47–58. Hartmann, J. (2005a) Long-term seismotectonic influence on the hydrochemical composition of a spring located at Koryaksky-Volcano, Kamchatka: deduced from aggregated earthquake information. Int. J. Earth Sci. (Geol. Rundsch.) 95, 649–664. Hartmann, J. (2005b) Difference information criterion for the analysis of a seismotectonic influence on a radon time series at KSM-site, Japan. Geophys. J. Int. 160(3), 891–900. Hartmann, J. and Levy, J. K. (2005) Hydrogeological and gasgeochemical earthquake precursors: a review for application. Nat. Haz. 34(3), 279–304. Hartmann, J. and Levy, J. K. (2006) The influence of seismotectonics on precursory changes in groundwater composition for the 1995 Kobe earthquake, Japan. Hydrogeology J. 14(7), 1307–1318. Hartmann, T. and Wenzel, H. G. (1995) The HW95 tidal potential catalogue. Geophys. Res. Lett. 22, 3553–3556. Hassan, M. U. M. and Baig, M. S. (2006) Paleogene biostratigraphy of Tattapani, Kotli Azad Kashmir, northwest sub-Himalayas, Pakistan. J. Himalayan Earth Sci. 39, 39– 48. Hauksson, E. (1981) Radon content of groundwater as an earthquake precursor: evaluation of worldwide data and physical basis. J. Geophys. Res. 86, 9397–9410. Long range pre-seismic gas-geochemical anomalies 153 Hauksson, E. and Goddard, J. G. (1981) Radon earthquake precursor studies in Iceland. J. Geophys. Res. 86, 7037–7054. Heinicke, J. and Koch, U. (2000) Slug flow—a possible explanation for hydrogeochemical earthquake precursors at Bad Brambach, Germany. Pure Appl. Geophys. 157, 1621– 1641. Heinicke, J., Koch, U. and Martinelli, G. (1995) CO 2 and radon measurements in the Vogtland area (Germany)—a contribution to earthquake prediction research. Geophys. Res. Lett. 22, 771–774. Horiguchi, K. and Matsuda, J. (2008) On the change of 3He/ 4 He ratios in hot spring gases after the Iwate–Miyagi Nairiku earthquake in 2008. Geochem. J. 42, e1–e4. Igarashi, G. and Wakita, H. (1990) Groundwater radon anomalies associated with earthquakes. Tectonophysics 180, 237– 254. Igarashi, G. and Wakita, H. (1991) Tidal responses and earthquake-related changes in the water level of deep wells. J. Geophys. Res. 96(B3), 4269–4278. Igarashi, G. and Wakita, H. (1995) Geochemical and hydrological observations for earthquake prediction in Japan. J. Phys. Earth 43, 585–598. Igarashi, G., Saeki, S., Takahata, N., Sumikawa, K., Tasaka, S., Sasaki, Y., Takahashi, M. and Sano, Y. (1995) Groundwater radon anomaly before the Kobe earthquake in Japan. Science 269, 60–61. India Meteorological Department (IMD) website reports. available at http://www.imd.gov.in/section/seismo/dynamic/ welcome.htm Ito, T. (2000) Seismo-geochemical anomalies of He/Ar ratio of gas bubbles at Hoshina spa near Matsushiro, Nagano Prefecture, central Japan. J. Earth Planet. Sci., Nagoya Univ. 47, 37–47. Ito, T., Nagamine, K., Yamamoto, K., Adachi, M. and Kawabe, I. (1999) Preseismic hydrogen gas anomalies caused by stress-corrosion process preceding earthquakes. Geophys. Res. Lett. 26, 2009–2012. Karangelos, D. J., Petropoulos, N. P., Anagnostakis, M. J., Hinis, E. P. and Simopoulos, S. E. (2002) Data leading to the investigation of a relation between seismic activity and radon daughters concentration outdoors. NRE VII, International Symposium, May 20–22, Rhodes, Greece (http:// arcas.nuclear.ntua.gr/public/page1.html). Kawabe, I. (1984) Anomalous changes of CH 4/Ar ratio in subsurface gas bubbles as seismogeochemical precursors at Matsuyama, Japan. Pure Appl. Geophys. 122, 194–214. Kawabe, I. (1987) Identification of seismogeochemical anomalies in subsurface gas CH 4/Ar ratio: Geochemical filtering of earthquakes. Geochem. J. 21, 105–117. Kennedy, B. M., Kharaka, Y. K., Evans, W. C., Ellwood, A., DePaolo, D. J., Thordsen, J., Ambats, G. and Mariner, R. H. (1997) Mantle fluids in the San Andreas fault system, California. Science 278, 1278–1281. King, C. Y. (1980) Episodic radon changes in subsurface soil gas along active faults and possible relation to earthquakes. J. Geophys. Res. 85, 3065–3078. King, C. Y. (1986) Gas geochemistry applied to earthquake prediction: An overview. J. Geophys. Res. 9(B12), 12269– 12281. 154 H. Chaudhuri et al. King, C. Y., Zhang, W. and Zhang, Z. (2006) Earthquakeinduced groundwater and gas changes. Pure Appl. Geophys. 163(4), 633–645. King, C. Y., Azuma, S., Ohno, M., Asai, Y., He, P., Kitagawa, Y., Igarashi, G. and Wakita, H. (2008) In search of earthquake precursors in the water-level data of 16 closely clustered wells at Tono, Japan. Geophys. J. Int. 143(2), 469– 477. Kudryavtsev, S. M. (2004) Improved harmonic development of the Earth tide-generating potential. J. Geodesy. 77, 829– 838. Lomnitz, C. and Lomnitz, L. (1978) Tangshan 1976, A case history in earthquake prediction. Nature 271, 109–111. Majumdar, N., Majumdar, R. K., Mukherjee, A. L., Bhattacharya, S. K. and Jani, R. A. (2005) Seasonal variations in the isotopes of oxygen and hydrogen in geothermal waters from Bakreswar and Tantloi, Eastern India: implications for groundwater characterization. J. Asian Earth Sci. 25, 269–278. Majumdar, N., Mukherjee, A. L. and Majumdar, R. K. (2009) Mixing hydrology and chemical equilibria in Bakreswar geothermal area, Eastern India. J. Volcanol. Geotherm. Res. 183(10), 201–212. Majumdar, R. K., Majumdar, N. and Mukherjee, A. L. (2000) Geoelectric investigations in Bakreswar geothermal area, West Bengal, India. J. Appl. Geophys. 45(3), 187–202. Millich, E., Lenzen, M. and Neugebauer, H. J. (1999) The influence of fluid flow induced by earth tides on radon transport. Geodynamik-Phys. Litossphäre, Univ. Bonn, Nußalke 8, 93–96. Mogro-Campero, A., Fleischer, R. L. and Likes, R. S. (1980) Changes in subsurface radon concentration associated with earthquake. J. Geophys. Res. 85, 3053–3057. Monnin, M. and Seidel, J. L. (1988) Sur une hypothétique émission intense de radon avant un événement géophysique majeur: une analyse théorique. C.R. Acad. Sci. Paris 307, 1363–1368. Morgounov, V. A. and Malzev, S. A. (2007) A multiple fracture model of pre-seismic electromagnetic phenomena. Tectonophysics 431, 61–72. Mukhopadhyay, M., Verma, R. K. and Ashraf, M. H. (1986) Gravity field and structures of the Rajmahal Hills: Example of the Paleo–Mesozoic continental margin in eastern India. Tectonophysics 131, 353–367. Nagamine, K. and Sugisaki, R. (1991) Coseismic changes of subsurface gas compositions disclosed by an improved seismo-geochemical system. Geophys. Res. Lett. 18, 2221– 2224. Nagar, R. K., Vishwanathan, G., Surendra, S. and Sankaranaraanan, A. (1996) Geological, geophysical and geochemical investigations in Bakreswar–Tantloi thermal field, Birbhum and Santhal Paragana districts, West Bengal and Bihar, India, Geothermal Energy in India. Geol. Surv. India Spl. Pub. 45, 349–360. Ouchi, T., Goriki, S. and Ito, K. (1985) On the space-time pattern formation of the earthquake strain field. Tectonophysics 113, 31–48. Parotidis, M., Rothert, E. and Shapiro, S. A. (2003) Porepressure diffusion: A possible triggering mechanism for the earthquake swarms 2000 in the Vogtland/NW Bohemia, central Europe. Geophys. Res. Lett. 30, 2075. Planinic, J., Radolic, V. and Vukovik, B. (2004) Radon as an earthquake precursor. Nucl. Instr. Meth. A 530, 568–574. Reimer, G. M. (1990) Soil gas helium increase preceding the Loma Prieta Earthquake. Eos, Trans.-Am. Geophys. Union 71, 289. Rikitake, T. (1975) Dilatancy model and empirical formulas for an earthquake area. Pure Appl. Geophys. 113, 141–147. Rikitake, T. (1987) Earthquake precursors in Japan: precursor time and delectability. Tectonophysics 136, 265–282. Rikitake, T. (1988) Earthquake prediction: an empirical approach. Tectonophysics 148, 195–210. Sarkar, A. N. (1982) Precambrian tectonic evolution of eastern India: a model of converging microplates. Tectonophysics 86, 363–366. Satake, H., Ohashi, M. and Hayashi, Y. (1985) Discharge of H 2 from the Atotsugawa and Ushikubi faults, Japan, and its relation to earthquakes. Pure Appl. Geophys. 122, 185–193. Sato, M., Sutton, A. J., McGee, K. A. and Russell-Robinson, S. (1986) Monitoring of hydrogen along the San Andreas and Calaveras faults in central California in 1980–1984. J. Geophys. Res. 91, 12,315–12,326. Scholz, C. H., Sykes, L. R. and Aggarwal, Y. P. (1973) Earthquake prediction: A physical basis. Science 181, 803–810. Shalimov, S. and Gokhberg, M. (1998) Lithosphere-ionosphere coupling mechanism and its application to the earthquake in Iran on June 20, 1990, A review of ionospheric measurements and basic assumptions. Phys. Earth Planet. Int. 105, 211–218. Shankar, R. (1988) Heat-flow map of India and discussion on its geological and economic significance. Indian Miner. 42, 89–110. Shankar, R. (1991a) The thermal and crustal structure of “Sonata”. A zone of mid continental rifting in Indian shield. J. Geol. Soc. India 37, 211–220. Shankar, R. (1991b) Distribution of thermal springs in NorthWestern India (J&K, Himachal Pradesh, Parts of Uttar Pradesh and Haryana), Geothermal Atlas of India. Geol. Surv. India Spl. Pub. 19, 10–13. Steinitz, G., Begin, Z. B. and Gazit-Yaari, N. (2003) A statistically significant relation between radon flux and weak earthquakes in the Dead Sea rift valley. Geology 31, 505–508. Sugisaki, R. (1978) Changing He/Ar and N2/Ar ratios of fault air may be earthquake precursors. Nature 275, 209–211. Sugisaki, R. (1981) Deep-seated gas emission induced by the Earth tide: A basic observation for geochemical earthquake prediction. Science 212(4500), 1264–1266. Sugisaki, R. and Sugiura, T. (1986) Gas anomalies at three mineral springs and a fumarole before an inland earthquake, central Japan. J. Geophys. Res. 91, 12,296–12,304. Sugisaki, R., Ito, T., Nagamine, K. and Kawabe, I. (1996) Gas geochemical changes at mineral springs associated with the 1995 southern Hyogo earthquake (M = 7.2), Japan. Earth Planet. Sci. Lett. 139, 239–249. Sultankhodzhayev, A. N., Latipov, S. U., Zakirov, T. Z. and Zigan, F. G. (1980) Dependence of hydrogeoseismological anomalies on the energy and epicentral distance of earthquakes. Dokl. AN Uzb. SSR 5, 57–59. Talwani, P. (1979) An empirical earthquake prediction model. Phys. Earth Planet. Int. 18, 288–302. Talwani, P., Moore, W. S. and Chiang, J. (1980) Radon anomalies and microearthquakes at Lake Jocassee, South Carolina. J. Geophys. Res. 85, 3079–3088. Thomas, D. M. (1988) Geochemical precursors to seismic activity. Pure Appl. Geophys. 126, 241–266. Toutain, J. P. and Baubron, J. C. (1999) Gas geochemistry and seismotectonics: A review. Tectonophysics 304, 1–27. Tsubokawa, I. (1973) On relation between duration of precursory geophysical phenomena and duration of crustal movement before earthquake. J. Geod. Soc. Japan 19, 116–119. Tsunogai, U. and Wakita, H. (1995) Precursory chemical changes in groundwater: Kobe earthquake, Japan. Science 269, 61–63. Tsunogai, U. and Wakita, H. (1996) Anomalous changes in groundwater chemistry-possible precursors of the 1995 Hyogo-ken Nanbu earthquake, Japan. J. Phys. Earth 44, 381–390. Tsvetkova T., Monnin, M., Nevinsky, I. and Perelygin, V. (2001) Research on variation of radon and gamma-background as a prediction of earthquakes in the Caucasus. Radiat. Meas. 33, 1–5. US Geological Survey (USGS) website reports. available at http://earthquake.usgs.gov Virk, H. S. (1995) Radon monitoring of microseismicity in the Kangra and Chamba Valleys of Himachal Pradesh, India. Nucl. Geophys. 9, 141–146. Virk, H. S. (1996) A critique of empirical scaling relationship between earthquake magnitude, epicentral distance and precursor time for interpretation of radon data. J. Earthquake Prediction. Res. 5, 547–583. Virk, H. S. and Baljinder, S. (1994) Radon recording of Uttarkashi earthquake. Geophys. Res. Lett. 21, 737–740. Virk, H. S., Sharma, A. K. and Walia, V. (1997) Correlation of alpha-logger radon data with microseismicity in N-W Himalaya. Curr. Sci. 72, 656–663. Virk, H. S., Walia, V. and Kumar, N. (2001) Helium/radon precursory anomalies of Chamoli earthquake, Garhwal Himalaya, India. J. Geodyn. 31, 201–210. Wadia, D. N. (1928) The geology of Poonch State (Kashmir) and adjacent portions of the Punjab. Mem. Geol. Surv. India 51, 185–370. Wakita, H. (1996) Geochemical challenge to earthquake prediction. Proc. Natl. Acad. Sci. USA, Colloquium Paper 93, 3781–3786. Wakita, H., Nakamura, Y. and Sano, Y. (1988) Short-term and intermediate-term geochemical precursors. Pure Appl. Geophys. 126(2–4), 267–278. Walia, V., Virk, H. S., Yang, T. F., Mahajan, S., Walia, M., Bajwa, B. S. (2005) Earthquake Prediction Studies using Radon as a Precursor in N-W Himalayas, India: A case study. Terr. Atoms. Oceanic Sci. 16(4), 775–804. Weinlich, F. H., Faber, E., Boušková, A., Horálek, J., Teschner, M. and Poggenburg, J. (2006) Seismically induced variations in Mariánské Lázně fault gas composition in the NW Bohemian swarm quake region, Czech Republic—A continuous gas monitoring. Tectonophysics 421, 89–110. Yang, T. F., Walia, V., Chyi, L. L., Fu, C. C., Wang, C. C., Chen, Long range pre-seismic gas-geochemical anomalies 155 C. H., Liu, T. K., Song, S. R., Lee, C. Y. and Lee, M. (2005) Variations of soil radon and thoron concentrations in a fault zone and prospective earthquakes in SW Taiwan. Radiat. Meas. 40, 496–502. Yang, T. F., Fu, C. C., Walia, V., Chen, C. H., Chyi, L. L., Liu, T. K., Song, S. R., Lee, M., Lin, C. W. and Lin, C. C. (2006) Seismo-geochemical variations in SW Taiwan: multiparameter automatic gas monitoring results. Pure Appl. Geophys. 163, 693–709. 156 H. Chaudhuri et al. Yoshino, T., Tomizawa, I. and Sigimoto, T. (1992) Results of statistical analysis of LF seismogenic emissions as precursors to the earthquake and volcanic eruptions. Res. Lett. Atmos. Electr. 12, 203–210. Zmazek, B., Zivcic, M., Todorovski, L., Dzeroski, S., Vaupotic, J. and Kobal, I. (2005) Radon in soil gas: how to identify anomalies caused by earthquakes. Appl. Geochem. 20, 1106– 1119.