Content uploaded by Toshihiko Takemura
Author content
All content in this area was uploaded by Toshihiko Takemura on Oct 16, 2014
Content may be subject to copyright.
335
Geochemical Journal, Vol. 46, pp. 335 to 339, 2012
*Corresponding author (e-mail: sebastian.d.aa@m.titech.ac.jp)
Copyright © 2012 by The Geochemical Society of Japan.
among its minor constituents sulfur-35 isotope (
35
S).
35
S is a radionuclide with a half life of 87 days pro-
duced in the atmosphere by cosmic ray spallation of
Argon-40 (Lal and Peters, 1967). Once produced,
35
S is
rapidly oxidized to
35
SO
2
(~1 sec.) and by further oxida-
tion this radionuclide containing molecule forms chemi-
cally stable sulfates (
35
SO
4
=
). Since all chemical proper-
ties of stable and radio isotopes are similar to the most
abundant one (
32
S, in this case), the aerosol formation
process followed by wet and dry removal from the at-
mosphere is the same for all isotopic species. Due to its
short half life, most of
35
SO
2
and
35
SO
4
=
produced in the
stratosphere does not readily reach the troposphere be-
fore decaying to
35
Cl except during stratosphere-
troposphere air exchange. The study of radionuclides pro-
duced by nuclear detonations over sea water has shown
the presence of
35
S produced by slow neutron capture of
35
Cl via
35
Cl(n,p)
35
S reaction (Dyrssen and Nyman, 1955).
35
S is also produced by slow neutrons via
34
S(n,
γ
)
35
S.
However, this process is 200 times slower than
35
Cl(n,p)
35
S (Love and Sam, 1962).
35
S production from
both pathways are likely to have taken place during the
injection of sea water within the reaction vessels of units
1, 2 and 3.
An estimation of the radioactive
35
S emitted into the atmospheric
from the Fukushima Daiichi Nuclear Power Plant
by using a numerical simulation global transport
SEBASTIAN O. DANIELACHE,
1
* CHISATO YOSHIKAWA,
2
ANTRA PRIYADARSHI,
3
TOSHIHIKO TAKEMURA,
4
YUICHIRO UENO,
1
MARK H. THIEMENS
3
and NAOHIRO YOSHIDA
1
1
Department of Earth and Planetary Science, Tokyo Institute of Technology, Meguro-ku, Tokyo 152-8551, Japan
2
Department of Environmental Science and Technology, Interdisciplinary Graduate School of Science and Engineering,
Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8502, Japan
3
9500 Gilman Drive, University of California San Diego, La Jolla, California 92093, U.S.A.
4
Research Institute for Applied Mechanics, Kyushu University, Fukuoka 816-8560, Japan
(Received February 29, 2012; Accepted June 3, 2012)
We present a numerical study carried out with the SPRINTARS model modified to account for the radioactive decay of
35
S compounds emitted from the Fukushima Daiichi nuclear power plant station after the hydrogen and vapor blast. The
transport dynamics of the released material reproduced previous field observations. Four different emission scenarios
were compared to the measurements of atmospheric
35
S in sulfate collected in La Jolla, Tsukuba, Kashiwa and Yokohama.
Linear regressions of the relation between emitted and transported material that reached the sampling sites were used to
estimate the amount of
35
S atoms and the amount of neutrons released in to the atmosphere. We estimate that a lower limit
of 1.9 × 10
16
35
S atoms sec
–1
were released after the events in March and this flux dropped to 4–39 × 10
14
35
S atoms
sec
–1
at the end of the month. Based on this calculations we estimated a lower limit of 5.2 × 10
21
slow neutrons m
–2
sec
–1
were emitted from the nuclear fuel rods to the sea water injected in the reactors after the events in March.
Keywords: Fukushima Daiichi nuclear power plant, radioactive isotopes, sulfate, global transport model, SPRINTARS
INTRODUCTION
The massive earthquake that affected the northeast
coast of Japan on March 11th released a powerful Tsu-
nami which hit the shores of Fukushima prefecture. The
power shortage occasioned by this natural disaster pro-
duced the meltdown in the reaction vessels of units 1, 2
and 3 of the Fukushima Daiichi nuclear power plant pro-
duced a hydrogen fueled vapor blasts that lasted for a
period of several days until March 15. Consequently, sev-
eral radioactive species were released into the boundary
layer, an event of such magnitude that has not been seen
since the Chernobyl disaster and the extent, significance
and consequences of such tragedy will remain a topic of
research and monitoring for many years. The material
released into the boundary layer includes a series of ra-
dioactive materials including radio nuclei of Iodine-131
(
131
I), cesium-137 (
137
Cs) (Chino et al., 2011; Yoshida
and Takahashi, 2012; Yoshida and Kanda, 2012) and
336 S. O. Danielache et al.
Priyadarshi et al. (2011) measured
35
S activity in
sulfate collected at Scripps Institution of Oceanography
(SIO), La Jolla California during March–April 2011. They
observed a distinct increase in
35
S activity (1500 atoms
m
–3
) from the natural background (450 atoms m
–3
) on 28
March 2011. Based on measurements and mathematical
modeling, they explained that the observed enhancement
in
35
SO
4
=
was due to the long range transport of radioac-
tive
35
SO
4
=
produced within the nuclear reactor in
Fukushima nuclear power plant.
35
S activity measured at
La Jolla peaked after the hydrogen blasts at the Fukushima
nuclear power plant. Their results complemented with a
moving box model study estimated a concentration of 2
× 10
5
molecules m
–3
of
35
S in the marine boundary layer
over Fukushima. An additional report by the same group
(Priyadarshi et al., in prep.) presented the fluctuation of
radioactive species
35
SO
2
and
35
SO
4
=
monitored at the
Tokyo Institute of Technology (TITECH), Atmosphere
and Ocean Institute (AORI) and National Institute of
Environmental Studies (NIES) in the Kanto area of Ja-
pan. At these monitoring sites, a maximum of 11778,
61402 and 5019 atoms m
–3
were respectively measured
during March 25–April 7.
A numerical simulation that models the global trans-
port of atmospheric particles after the major emission of
radioactive materials has been reported by Takemura et
al. (2011). Their results replicate the transport and dissi-
pation patterns reported by numerous measurements and
simulations carried out throughout the planet (See
Takemura et al. (2011) and references there in). Here we
present the results of using a modify version of the code
employed by Takemura et al. (2011) where radioactive
decay of
35
SO
4
=
is added to the simulation. Our results
are compared to the timing and amount of
35
S material
detected in California and Tokyo. Using field measure-
ments as constraining conditions, we estimate the bulk
amount of
35
S material released from the nuclear power
plant.
MODEL DESCRIPTION
The model employed for this numerical study is a spec-
tral radiation-transport model for aerosol species
(SPRINTARS) developed by Takemura and co-workers
(Takemura et al., 2000, 2002, 2005). This code has been
developed to calculate physical processes involved in the
transport of particles including emission, molecular and
turbulent diffusion, advection by inter-grid air motion,
while deposited species are conformed by gravity, wet
and dry deposition. The dynamic core of SPRINTARS is
based on the atmospheric component of a global climate
model MIROC (Watanabe et al., 2010) and therefore is
capable of calculating global atmospheric state, such as
air motion, temperature, moisture, clouds and precipita-
tion. The model grid consists of 20 vertical levels where
the lowermost 4 are located below 1 km altitude while
the horizontal resolution is approximately 0.56° by 0.56°
in latitude and longitude (T213 spectral truncation in the
dynamical core). In order to achieve higher accuracy, the
atmospheric state generated by the dynamic core was as-
similated to a 6-hourly data based on dynamic observa-
tions.
The release of material was approximated in the model
by injecting particles in the lowest layer at the grid point
where the nuclear power plant is located (Fig. 1). The
emission of particles was kept constant from 1200 UTC,
March 14 when measured gamma radiation was above 2
mSv/h at the plant (TEPCO, 2011), until the end of the
simulation on April 10. This assumption is taken on the
bases that the
35
S emissions are not known and the mea-
sured amounts of gamma radiation are a proxy of the re-
leased material, therefore from a model point of view the
simulation represents the ratio in concentration between
the emission domain and any given point at considerable
distance. A more detailed presentation of the emission
scenario is presented in Takemura et al. (2011). In this
experiment the authors considered the emission to be com-
posed of
137
Cs and other radioactive materials. In our
study, emitted material is composed of sulfate aerosols
since
35
S formed in the reactor is rapidly converted to
35
SO
4
=
, since the generated
35
S atom is highly unstable
and is likely to be oxidized to H
2
35
SO
4
(Giulianelli and
Willard, 1974). The relation between mode radii (
µ
m) and
relative humidity is a deciding factor in the simulation of
dry, wet and gravitational settling. In this study, the radii
values for a 0–99% calculated relative humidity range was
0.0695–0.231
µ
m. For a detailed description and formu-
lation of the computations of aerosols and their proper-
Fig. 1. Geographical position of Fukushima Daiichi nuclear
power plant (37.4
°
N, 141.8
°
E) and
35
SO
4
=
sampling sites
TITECH (35.3
°
N, 139.3
°
E), AORI (35.9
°
N, 140.0
°
E), NIES
(36.1
°
N, 140.1
°
E) and SIO (32.8
°
N, 117.3
°
W).
Estimation of
35
S emitted from the Fukushima Daiichi Nuclear Power Plant 337
ties, see Takemura et al. (2000, 2002, 2005). The code
has been modified to take into account the radioactive
decay (half life 87 days) of emitted
35
SO
4
=
. In this model
the mass balance of the emitted material consist of wet,
dry, gravitational deposition and radioactive decay (81%,
11%, 0.02% and 0.7%, respectively).
The emission scenario for this study was set to 1.43 ×
10
16
estimated from the numerical values presented by
Priyadarshi et al. (2011) and expanded to 3 different emis-
sion scenarios (1.43 × 10
14
, 1.43 × 10
15
, and 1.43 × 10
17
atoms s
–1
). Once the linearity of the relation between re-
leased material and the material that reached SIO, NIES,
AORI and TITECH was confirmed, the number of atoms
of
35
S emitted after the considered events (hydrogen and
vapor blasts) were estimated.
RESULTS AND DISCUSSION
Figure 2 shows the simulated distribution of
35
S con-
centration emitted from Fukushima Daiichi nuclear power
plant. The numerical data presented in this figure corre-
spond to the scenario where 1.43 × 10
16
atoms s
–1
were
released into the lowermost layer of the simulation grid.
Panel a and b of Fig. 2 shows the initial wave of
35
SO
4
=
species being transported eastwards. The simulation
shows that the initial wave (concentrations above 10
–13
atoms m
–3
) reached 120°W latitude on March 17. The
simulation presented in this report shows a transport time
of 4 days since the initial injection of radioactive mate-
rial. Panel c and d of Fig. 2 shows that the concentration
of the transported
35
SO
4
=
species arriving at La Jolla ex-
ceeded 1000 m
–3
on March 25, 12 days after the initial
blast. This maximum was also measured at La Jolla in a
sample retrieved during March 24–28 (Priyadarshi et al.,
2011). After this 12 days maximum, the emitted
35
SO
4
=
species is dispersed within 16 km altitude layer of the
northern hemisphere.
Figure 3 presents simulated temporal variations of
35
SO
4
=
compounds concentrations at 4 different sites
within Japan (TITECH, NIES and AORI) and US (SIO).
Since the emission flux of 1.43 × 10
16
atoms s
–1
was kept
constant, large peaks sampled within Tokyo area and La
Jolla suggest specific wind patterns of transported mate-
rial with high concentration levels of
35
SO
4
=
from the
emission source. In order to validate the inferred conclu-
sions, our calculations have been compared to field meas-
urements carried out at TITECH, NIES, AORI and SIO
and reported by Priyadarshi et al. (2011, in preparation).
The simulation presents peaks with concentrations of
35
SO
4
=
material of above 10000 atoms m
–3
in Tokyo area
Fig. 2. Simulated near surface concentration of
35
SO
4
=
material emitted continuously (1.43
×
10
16
emission scenario) presented
for the entire globe for March 17 (a) and March 25 (c). Panel b and d present latitudinal mean concentration in the northern
hemisphere on the longitudinal vs. altitude plot. The chromatic scale represents atoms m
–3
.
338 S. O. Danielache et al.
during the periods of March 14–15, March 20–24, March
28–April 1, April 3–8 and April 9 (Fig. 3a), and peaks of
above 1000 atoms m
–3
measured at SIO during March 25
and March 27–28 period (Fig. 3b). These modeled peaks
replicate the timing of appearance with high degree of
accuracy with the measurements of Priyadarshi et al.
(2011 and in prep.) carried out in Japan (March 23–25,
March 28–April 4) and US (March 24–28). The time evo-
lution of the released material from the power plant shows
that the measured peaks represent events where the emit-
ted material was transported by southeastward winds in
the case of Tokyo area and the North Pacific Jet stream
for the case of SIO.
From the 4 calculated case scenarios, a linear regres-
sion of the relation between emitted flux and transported
material concentration at each site was constructed and
the measured amounts of
35
S atoms were employed to
calculate the specific emitted flux of
35
S material from
the nuclear reactors. Calculated fluxes were then em-
ployed to estimate the amount of slow neutrons emitted
from the nuclear fuel rods that activated the sea water
within the reactors. The methodology of this conversion
was the same as the one employed by Priyadarshi et al.
(2011). Table 1 presents a summary of observed
35
SO
4
=
,
calculated
35
S emission flux and released neutron amount.
Since the background levels of
35
S atoms produced by
cosmic rays spallation have some fluctuation, data sets
with large concentrations of
35
S species (above 10000
atoms m
–3
) were employed for the case of Tokyo but back-
ground corrected values were employed for the measure-
ments at SIO. Backtrace analysis presented by Priyadarshi
et al. (2011) of
35
S species at SIO between March 24–28
shows that it took 10 days of transport since its, based on
this analysis we estimate that 183 × 10
14
atoms s
–1
were
emitted from the nuclear power plant between March 14
and 18 (Table 1). The measurements at AORI between
March 23 and March 25 were used to estimate an emis-
sion of 91 × 10
14
atoms s
–1
. Joint measurements at AORI,
NIES and TITECH between March 28 and April 1 esti-
mate an emission of 4–39 × 10
14
atoms s
–1
that were trans-
ported within a day from the source. Based on these re-
sults, the emission simulated between March 28 and April
1 represents approximately 8% of the initial emissions
simulated between March 14–18. Chino et al. (2011) study
on the time variation of released main radioactive nuclei
131
I and
137
Cs estimates a reduction in the order of 10–
100 from the emission on March 14 until sometime be-
tween March 28 and April 1. If this initial amount of
material is taken as representative of the occurred events
we estimate that 52 × 10
20
slow neutrons m
–2
sec
–1
were
released from the fuel rods into the sea water present
within the nuclear reactor. Priyadarshi et al. (2011) have
estimated a release amount of 4 × 10
11
slow neutrons
m
–2
. The large difference with our estimation comes from
the intrinsic limit of the box model study by Priyadarshi
Fig. 3. Simulated temporal variation of
35
SO
4
=
compounds
concentrations at 1.43
×
10
16
emission scenario for NIES,
TITECH and AORI Japan (a) and SIO, US (b).
Site Observed term Measured
35
SO
4
=
(atoms m
–3
)
35
S emission flux
(×10
14
atoms s
–1
)
Neutron flux
(×10
20
m
–2
s
–1
)
SIO 24–28 March 1050* 183 52
AORI 23–25 March 17074 91 26
AORI 28–30 March 13893 4 1
AORI 30 March–1 April 61402 39 11
NIES 29 March–1 April 18049 4 1
TITECH 28–30 March 11778 7 2
3/11/11 3/16/11 3/21/11 3/26/11 3/31/11 4/5/11 4/10/11
0
1,000
2,000
3,000
4,000
SIO
0.E+00
1.E+07
2.E+07
3.E+07
NIES
TITECH
AORI
(a)
(b)
[atoms*m
–3
]
[days]
Table 1.
35
S emission and calculated neutron flux at the Fukushima Daiichi
nuclear power plant estimated by the linearity of the relation between released
material and the material detected at each site
*This value is corrected for the background levels of
35
S measured between March 2 and 9.
Estimation of
35
S emitted from the Fukushima Daiichi Nuclear Power Plant 339
et al. (2011), which takes into account deposition and de-
cay during the trans-Atlantic transport but is limited to
estimate the actual amount of
35
S atoms present at inside
the reactor. By considering different scenarios our model
directly estimates the amount of material released from
the reactor core. The estimated initial
35
S material and
number of neutron represent a lower limit of the amount
of radiation emitted from the nuclear reactors at
Fukushima Daiichi nuclear power plant. These values can
be used as a proxy to the total amount of radiation emit-
ted since the melt down.
Acknowledgments—S. O. Danielache, C. Yoshikawa and N.
Yoshida would like to express their gratitude to the Global En-
vironmental Fund (A094) of the Japanese Ministry of Environ-
ment and the Global COE program “Earth to Earths”, Ministry
of Education, Culture, Sports, and Technology (MEXT), Japan
for their financial support. S. Danielache and Y. Ueno are sup-
ported by the NEXT program (GR033) of MEXT, Japan. T.
Takemura expresses his gratitude to the Funding Program for
Next Generation World-Leading Researchers by the Cabinet
Office, Government of Japan (GR079). The authors thank K.
Yamada and S. Toyoda for their comments and support.
REFERENCES
Chino, M., Nakayama, H., Nagai, H., Terada, H., Katata, G.
and Yamazawa, H. (2011) Preliminary estimation of release
amounts of
131
I and
137
Cs accidentally discharged from the
Fukushima Daiichi Nuclear Power Plant into the atmos-
phere. J. Nuclear Sci. Technol. 48, 1129–1134.
Dyrssen, D. and Nyman, P. O. (1955) Slow-neutron-induced
radioactivity of sea-water. Acta Radiol. 43, 421–427.
Giulianelli, J. L. and Willard, J. E. (1974) Chemical states of
sulfur-35 formed by the
35
Cl(n,p)
35
S process in potassium
chloride. J. Phys. Chem. 78, 372–374.
Lal, D. and Peters, B. (1967) Cosmic ray produced radioactiv-
ity in the earth. Hand Phys. 46, 551–612.
Love, D. L. and Sam, D. (1962) Radiochemical determination
of sodium-24 and sulfur-35 in seawater. Anal. Chem. 34,
336–340.
Priyadarshi, A., Dominguez, G. and Thiemens, M. H. (2011)
Evidence of neutron leakage at the Fukushima nuclear plant
from measurements of radioactive (35)S in California. Pro-
ceedings of the National Academy of Sciences of the United
States of America 108, 14422–14425.
Takemura, T., Okamoto, H., Maruyama, Y., Numaguti, A.,
Higurashi, A. and Nakajima, T. (2000) Global three-
dimensional simulation of aerosol optical thickness distri-
bution of various origins. J. Geophys. Res.-Atmos. 105,
17853–17873.
Takemura, T., Uno, I., Nakajima, T., Higurashi, A. and Sano, I.
(2002) Modeling study of long-range transport of Asian dust
and anthropogenic aerosols from East Asia. Geophys. Res.
Lett. 29, doi:10.1029/2002GL016251.
Takemura, T., Nozawa, T., Emori, S., Nakajima, T. Y. and
Nakajima, T. (2005) Simulation of climate response to aero-
sol direct and indirect effects with aerosol transport-
radiation model. J. Geophys. Res.-Atmos. 110, doi:10.1029/
2004JD005029.
Takemura, T., Nakamura, H., Takigawa, M., Kondo, H.,
Satomura, T., Miyasaka, T. and Nakajima, T. (2011) A nu-
merical simulation of global transport of atmospheric par-
ticles emitted from the Fukushima Daiichi Nuclear Power
Plant. Sola 7, 101–104.
TEPCO (2011) Available at http://www.tepco.co.jp/cc/press/
betu11_j/images/110528d.pdf
Watanabe, M., Suzuki, T., O’ishi, R., Komuro, Y., Watanabe,
S., Emori, S., Takemura, T., Chikira, M., Ogura, T.,
Sekiguchi, M., Takata, K., Yamazaki, D., Yokohata, T.,
Nozawa, T., Hasumi, H., Tatebe, H. and Kimoto, M. (2010)
Improved Climate Simulation by MIROC5; Mean states,
variability, and climate sensitivity. J. Climate 23, 6312–
6335.
Yoshida, N. and Kanda, J. (2012) Tracking the Fukushima
radionuclides. Science 336, 1115–1116.
Yoshida, N. and Takahashi, Y. (2012) Land-surface contamina-
tion by radionuclides from the Fukushima Daiichi Nuclear
Power Plant accident. Elements 8, 201–206.
