**3. Intensities of gamma-radiation in seawater and on seafloor**

Due to the effective shield by seawater, total count rate in water is rather lower than that of sub aerial measurement (Yoshida and Tsukaraha, 1987). According to the geological environment, total count rate of sub aerial environment is 20-100 nGy/h in Japan, which is equivalent to approx. 80-400 counts per second (cps) in our system: 1 nGy/h is equivalent to 4cps. Figure 2 is an example of the time-series record of total count rate in submersible dive:

Environmental Gamma-Ray Observation in Deep Sea 59

vents or cold seeps associated with active fault showed significantly anomalous count-rate, up to 104 cps; the highest total count rate was recorded at Izena Hole caldron, >8900 cps. Other hydrothermally active areas also showed rather high count-rates, e.g. Iheya ridges, Hatoma-knoll, Ishigaki-knolls of Nansei-shoto and Kagoshima-bay (Southwest Japan), Myojin-knoll of Shichito-Iwojima Ridge. To judge the apparent anomaly from the above described general distribution, here after 100 cps is adopted as a tentative threshold to identify geological activities. Here also pseudo-logarithmic scales were applied in Figure 3b to show the entire variation of total count-rate. In this figure, distribution peak of the γ-ray

Fig. 3. Maximum total count rate distribution recorded at individual dives of submersibles: (a) Linear scale classification of every 10 cps in intervals, (b) Data classification under logalithmic-like scale. Blue bars: frequency in the class; Red sequential line: cumulative

Hereafter, some examples of the intensities of environmental γ-ray obtained in various geological settings are discussed. Figure 4 summarizes the measured distributions of total count rate maxima of γ-radiation around Japan. It is clear that the anomalous values were recorded all the area around Japan even in the area rather old geologic edifices; e.g. Komahashi-daini Knoll of Kyushu-Palau Ridge, or Annei Smt. of Nishi-Shichito Ridge.

The localities where very high count-rates were observed, > 1000 cps, were limited to the active hydrothermal sites developing on the arc-backarc volcanoes: volcanism developing above trench-arc system relating to seafloor subduction. In contrast to the hydrothermal sites, fore-arc cold seepages showed moderately high cont-rates, up to 500 cps; mostly not exceeding to 200 cps. It is notable that four of five localities where the very high count-rate recorded (>1000 cps) were in Okinawa Trough. It is a back arc basin developing between Nansei-Shoto and Asian continent in East China Sea, where thick terrigeneous sediments have been accumulated. Generally, major radioactive elements, K, Th and U1, are rich in

1 All these elements are geochemically classified to incompatible elements that concentrated into continental crust due to their incompatibility to the rock forming minerals. Thus, oceanic crust or

**3.1 Relationships between tectonic settings and gamma-ray signature** 

relative frequencies from lower count rates.

magmas in ocean are relatively poor in such elements.

intensity is on the class 50-100 cps, which also support this tentative threshold.

(a) (b)

Fig. 2. A typical time sequential record obtained by submersible dives; 895th dive of *Shinnkai2000* at Myojin Knoll, Izu-Bonin Arc. Original Fig. was in Hattori & Okano (1999).

collected at the 895th dive of Shinkai2000. The cosmic-ray induced high energy gamma ray, > 3MeV as its energy level, still penetrate into shallow water, which raises total count rate approx. 20 cps. Descending to deeper, the total count rate decreases down to 10 cps or lower in typical; in this example, the minimum count rate is approx.14 cps. In the area with little suspended material in water column, such a minimum decreases down to 5 cps, e.g. in the vicinity of Mid Atlantic Ridges (Hattori et al., 2001). Approaching to the seafloor, total count rate raises up again to a few tens cps in most cases, which indicate significant contribution from seabed. Figure 3 shows the distributions of the maximum total count rates recorded when submersibles or ROVs were on bottom, touching on the seabed. The mode of the recorded value is in the class of 10-20 cps in linear scale with 10 cps in intervals, however, the frequency of the count rate gradually decreases to much higher count rate; there is no significant gap in the distribution. Even in this context, the area around the hydrothermal

Fig. 2. A typical time sequential record obtained by submersible dives; 895th dive of *Shinnkai2000* at Myojin Knoll, Izu-Bonin Arc. Original Fig. was in Hattori & Okano (1999).

collected at the 895th dive of Shinkai2000. The cosmic-ray induced high energy gamma ray, > 3MeV as its energy level, still penetrate into shallow water, which raises total count rate approx. 20 cps. Descending to deeper, the total count rate decreases down to 10 cps or lower in typical; in this example, the minimum count rate is approx.14 cps. In the area with little suspended material in water column, such a minimum decreases down to 5 cps, e.g. in the vicinity of Mid Atlantic Ridges (Hattori et al., 2001). Approaching to the seafloor, total count rate raises up again to a few tens cps in most cases, which indicate significant contribution from seabed. Figure 3 shows the distributions of the maximum total count rates recorded when submersibles or ROVs were on bottom, touching on the seabed. The mode of the recorded value is in the class of 10-20 cps in linear scale with 10 cps in intervals, however, the frequency of the count rate gradually decreases to much higher count rate; there is no significant gap in the distribution. Even in this context, the area around the hydrothermal vents or cold seeps associated with active fault showed significantly anomalous count-rate, up to 104 cps; the highest total count rate was recorded at Izena Hole caldron, >8900 cps. Other hydrothermally active areas also showed rather high count-rates, e.g. Iheya ridges, Hatoma-knoll, Ishigaki-knolls of Nansei-shoto and Kagoshima-bay (Southwest Japan), Myojin-knoll of Shichito-Iwojima Ridge. To judge the apparent anomaly from the above described general distribution, here after 100 cps is adopted as a tentative threshold to identify geological activities. Here also pseudo-logarithmic scales were applied in Figure 3b to show the entire variation of total count-rate. In this figure, distribution peak of the γ-ray intensity is on the class 50-100 cps, which also support this tentative threshold.

Fig. 3. Maximum total count rate distribution recorded at individual dives of submersibles: (a) Linear scale classification of every 10 cps in intervals, (b) Data classification under logalithmic-like scale. Blue bars: frequency in the class; Red sequential line: cumulative relative frequencies from lower count rates.

#### **3.1 Relationships between tectonic settings and gamma-ray signature**

Hereafter, some examples of the intensities of environmental γ-ray obtained in various geological settings are discussed. Figure 4 summarizes the measured distributions of total count rate maxima of γ-radiation around Japan. It is clear that the anomalous values were recorded all the area around Japan even in the area rather old geologic edifices; e.g. Komahashi-daini Knoll of Kyushu-Palau Ridge, or Annei Smt. of Nishi-Shichito Ridge.

The localities where very high count-rates were observed, > 1000 cps, were limited to the active hydrothermal sites developing on the arc-backarc volcanoes: volcanism developing above trench-arc system relating to seafloor subduction. In contrast to the hydrothermal sites, fore-arc cold seepages showed moderately high cont-rates, up to 500 cps; mostly not exceeding to 200 cps. It is notable that four of five localities where the very high count-rate recorded (>1000 cps) were in Okinawa Trough. It is a back arc basin developing between Nansei-Shoto and Asian continent in East China Sea, where thick terrigeneous sediments have been accumulated. Generally, major radioactive elements, K, Th and U1, are rich in

<sup>1</sup> All these elements are geochemically classified to incompatible elements that concentrated into continental crust due to their incompatibility to the rock forming minerals. Thus, oceanic crust or magmas in ocean are relatively poor in such elements.

Environmental Gamma-Ray Observation in Deep Sea 61

radiation dose rate or concentrations of γ-ray source nuclei in the environment could be calculated. Hereafter the radiation dose rates are applied because it is not obvious the spatial distribution of the source nuclei at the various locations; geometric complexities of the seafloor having tall hydrothermal chimneys abruptly standing up > 10m in some cases and contribution from vigorous flow out from hydrothermal chimneys or seepages may cause

Figure 5 shows the statistics of γ-ray dose rate from three possible sources, K, U-series and Th-series, respectively. K is major element in seafloor sediment; the recorded maximum of dose rate was 7.0 μR/h regardless of its mode at 0.28 μR/h. Entire variation was within three orders of magnitude, which was narrower than those of U- and Th-series radiations

Contrary to K, dose-rate of U-series varies more than five orders of magnitude caused by its maximum of ~200 μR/h regardless of its relatively low mode (0.54 μR/h; Figure 5(b)). Dose rate distribution of Th-series shows the intermediate nature between K and U-series, ranging within four orders of magnitude (Figure 5(c)). It is notable that its mode of distribution is 1.38μR/h, which is much higher than that of U-series. It may relate to the rather broad

Under comparison with the recorded enormous total count-rate and the large variation of U-series dose rate, high concentration of U-series in environment should cause very high intensities of γ-ray, > 1000 cps, equivalent to >250 nGy/h. To confirm this view, the relationship of total count rates with U-series dose rate is plotted in Figure 6. It is clear that the observed high count rates were tightly associated with the high dose rate of U-series, > 100 cps as total count rate. Considering the progeny radio-nuclei of U-series and their half lives, Ra fed from hydrothermal vent fluids causes such radio activities2. Even in the lower count-rate environment, rough relationship between total count rate and U-series dose rate still found, which suggests a ubiquity of U-series controlled γ-radiation environment; that is represented by the areas of hydrothermal activities in the higher dose rate end of correlation line. In addition, the trends was slightly above from the extrapolation defined in >100cps of total count rate. It is suggested significant contributions caused from Th-series nuclei, which

To investigate such Th-series contribution, measured dose rates of U-series and Th-series were plotted (Figure 7). As predicted the above described relationship between total count rate and U-series dose rate, the data were dominantly plotted around the correlation line of slope 1. This also supports the view of a ubiquity of U-series controlled γ-radiation environment. Above the trend, a few tens of data obtained from the fore-arc seepage areas were plotted; e.g. in Suruga Bay or in Sagami Bay (both are in southern coast of mainland Japan), or off Kamaishi-city near the Japan Trench. In such areas, the U-series dose rate is not so high regardless of the moderately high total count rate, up to 300 cps. Instead, dose rate of Th-series reaches approx. 5 μR/h regardless of the dose rate of U-series of <1 μR/h. As these areas are in the vicinities of active faults, thus, the contribution from Th-series nuclei is significant in the tectonics controlled environment. The data points from other tectonic active area either on southern coast of mainland Japan (e.g. Zenisu-ridge) or in

2 Ra is in the group of an alkaline earth elements, which shows high solubility to water. Ra also

frequently replaces Ba in minerals as the same group element.

distribution tailing to the higher side of dose rate. This nature will be discussed later.

unexpected increase of gamma radiation.

(Figure 5(a)).

built up the total count-rate.

such terrigeneous sediments, which potentially cause the high count-rate of γ-radiation. The hydrothermal circulation within the sedimentary layer enhanced by the magmatic heat sources may scavenge and concentrate such radioactive species from thick sediments in those areas. Therefore, it is natural that the hydrothermal activities in oceanic environment far from continents or large land masses, e.g. at Mid-Ocean Ridges, did not show any very high count-rates. The vicinity of Hawaiian Island is also the area of low γ-ray intensities. In those areas, maximum count rates did not exceed 200 cps.

Fig. 4. The regional distribution of anomalous total count rate maxima, 100 cps, in the vicinity of Japan. Bathymetric contours are drawn as 1000 m in intervals. Original figure was taken from Hattori & Okano (2002) and redrawn.

Even by using NaI(Tl) scintillator with 7% resolution, the sources of γ-ray could be resolved by its energies. In this purpose, three characteristic γ-ray of 609 keV of 214Bi, 1460 keV of 40K and 2614 keV of 208Tl were used in this study. Under some reasonable assumptions, either

such terrigeneous sediments, which potentially cause the high count-rate of γ-radiation. The hydrothermal circulation within the sedimentary layer enhanced by the magmatic heat sources may scavenge and concentrate such radioactive species from thick sediments in those areas. Therefore, it is natural that the hydrothermal activities in oceanic environment far from continents or large land masses, e.g. at Mid-Ocean Ridges, did not show any very high count-rates. The vicinity of Hawaiian Island is also the area of low γ-ray intensities. In

Fig. 4. The regional distribution of anomalous total count rate maxima, 100 cps, in the vicinity of Japan. Bathymetric contours are drawn as 1000 m in intervals. Original figure was

Even by using NaI(Tl) scintillator with 7% resolution, the sources of γ-ray could be resolved by its energies. In this purpose, three characteristic γ-ray of 609 keV of 214Bi, 1460 keV of 40K and 2614 keV of 208Tl were used in this study. Under some reasonable assumptions, either

taken from Hattori & Okano (2002) and redrawn.

those areas, maximum count rates did not exceed 200 cps.

radiation dose rate or concentrations of γ-ray source nuclei in the environment could be calculated. Hereafter the radiation dose rates are applied because it is not obvious the spatial distribution of the source nuclei at the various locations; geometric complexities of the seafloor having tall hydrothermal chimneys abruptly standing up > 10m in some cases and contribution from vigorous flow out from hydrothermal chimneys or seepages may cause unexpected increase of gamma radiation.

Figure 5 shows the statistics of γ-ray dose rate from three possible sources, K, U-series and Th-series, respectively. K is major element in seafloor sediment; the recorded maximum of dose rate was 7.0 μR/h regardless of its mode at 0.28 μR/h. Entire variation was within three orders of magnitude, which was narrower than those of U- and Th-series radiations (Figure 5(a)).

Contrary to K, dose-rate of U-series varies more than five orders of magnitude caused by its maximum of ~200 μR/h regardless of its relatively low mode (0.54 μR/h; Figure 5(b)). Dose rate distribution of Th-series shows the intermediate nature between K and U-series, ranging within four orders of magnitude (Figure 5(c)). It is notable that its mode of distribution is 1.38μR/h, which is much higher than that of U-series. It may relate to the rather broad distribution tailing to the higher side of dose rate. This nature will be discussed later.

Under comparison with the recorded enormous total count-rate and the large variation of U-series dose rate, high concentration of U-series in environment should cause very high intensities of γ-ray, > 1000 cps, equivalent to >250 nGy/h. To confirm this view, the relationship of total count rates with U-series dose rate is plotted in Figure 6. It is clear that the observed high count rates were tightly associated with the high dose rate of U-series, > 100 cps as total count rate. Considering the progeny radio-nuclei of U-series and their half lives, Ra fed from hydrothermal vent fluids causes such radio activities2. Even in the lower count-rate environment, rough relationship between total count rate and U-series dose rate still found, which suggests a ubiquity of U-series controlled γ-radiation environment; that is represented by the areas of hydrothermal activities in the higher dose rate end of correlation line. In addition, the trends was slightly above from the extrapolation defined in >100cps of total count rate. It is suggested significant contributions caused from Th-series nuclei, which built up the total count-rate.

To investigate such Th-series contribution, measured dose rates of U-series and Th-series were plotted (Figure 7). As predicted the above described relationship between total count rate and U-series dose rate, the data were dominantly plotted around the correlation line of slope 1. This also supports the view of a ubiquity of U-series controlled γ-radiation environment. Above the trend, a few tens of data obtained from the fore-arc seepage areas were plotted; e.g. in Suruga Bay or in Sagami Bay (both are in southern coast of mainland Japan), or off Kamaishi-city near the Japan Trench. In such areas, the U-series dose rate is not so high regardless of the moderately high total count rate, up to 300 cps. Instead, dose rate of Th-series reaches approx. 5 μR/h regardless of the dose rate of U-series of <1 μR/h. As these areas are in the vicinities of active faults, thus, the contribution from Th-series nuclei is significant in the tectonics controlled environment. The data points from other tectonic active area either on southern coast of mainland Japan (e.g. Zenisu-ridge) or in

<sup>2</sup> Ra is in the group of an alkaline earth elements, which shows high solubility to water. Ra also frequently replaces Ba in minerals as the same group element.

Environmental Gamma-Ray Observation in Deep Sea 63

Fig. 6. The relationship between maximum total count rate and maxim dose rate from U-

Fig. 7. The relationship of maximum dose rate from U-series and Th-series nuclei. Above the dense data distribution well correlate with U-series increase, a few tens of relatively high

Th-series dose rate data were found.

series nuclei. Most of data were tightly scattered around slope-1 correlation line.

Fig. 5. The maximum dose rate distribution of individual sources; (a): K, (b): U-series, and (c): Th-series.

Japan sea-side (e.g. off Rebun Isl.) are also plotted above the line regardless of rather lower total count rates. These signatures may relate to squeezing of pore fluids by the compaction of sediments, which supplies Ra or other soluble species in seafloor environment.

Fig. 5. The maximum dose rate distribution of individual sources; (a): K, (b): U-series, and

Japan sea-side (e.g. off Rebun Isl.) are also plotted above the line regardless of rather lower total count rates. These signatures may relate to squeezing of pore fluids by the compaction

of sediments, which supplies Ra or other soluble species in seafloor environment.

(c): Th-series.

Fig. 6. The relationship between maximum total count rate and maxim dose rate from Useries nuclei. Most of data were tightly scattered around slope-1 correlation line.

Fig. 7. The relationship of maximum dose rate from U-series and Th-series nuclei. Above the dense data distribution well correlate with U-series increase, a few tens of relatively high Th-series dose rate data were found.

Environmental Gamma-Ray Observation in Deep Sea 65

Island through the electro-optical submarine cable. The energy spectra can be obtained by automated calculation every ten minutes at the shore station, i.e. each dataset of energy spectrum is the summation of ten minute measurement. The gamma-ray sensor unit is installed to touch its scintillator side on seabed in which scintillator is attached downward

Plate 2. Gamma ray sensor of cabled observatory off Hatsushima Island.

Plate 3. Cabled observatory off Hatsushima Island and gamma ray sensor

It is known that output signal of NaI(Tl) scintillator is affected by temperature variation. However, since the water temperature at the observation site on deep seafloor is approx. 3 °C and shows very small perturbation, the influence with temperature is negligible. On the other hand, some kind of signal drift associated with aging could occur. Fig. 8 shows the spectra obtained on January 1st in 2003, 2005, 2007, 2009 and 2011. Each spectrum is

(denoted by circle) deployed on seafloor.

to maximize its sensitivity (Plate 3).
