Journal of Environmental Radioactivity 144 (2015) 86e95

Contents lists available at ScienceDirect

Journal of Environmental Radioactivity

journal homepage: www.elsevier.com/locate/jenvrad

Effects of radiocesium inventory on 137Cs concentrations in river

waters of Fukushima, Japan, under base-flow conditions
Shinya Ochiai a, *, Shinji Ueda a, Hidenao Hasegawa a, Hideki Kakiuchi a,
Naofumi Akata a, b, Yoshihito Ohtsuka a, Shun'ichi Hisamatsu a
a
Department of Radioecology, Institute for Environmental Sciences, 1-7 Ienomae, Obuchi, Rokkasho, Kamikita, Aomori 039-3212, Japan
b
National Institute for Fusion Science, 322-6 Oroshi-cho, Toki City, Gifu 509-5292, Japan

a r t i c l e i n f o a b s t r a c t

Article history: To investigate the behavior of nuclear accident-derived 137Cs in river water under base-flow conditions,
Received 18 December 2014 concentrations of dissolved and particulate 137Cs were measured at 16 sampling points in seven rivers of
Received in revised form Fukushima Prefecture, Japan, in 2012 and 2013. The concentration of dissolved 137Cs was significantly
2 March 2015
correlated with the mean 137Cs inventory in the catchment area above each sampling point in both
Accepted 5 March 2015
Available online
sampling years. These results suggest that the concentration of dissolved 137Cs under base-flow condi-
tions is primarily determined by the 137Cs inventory of the catchment area above the sampling point.
However, the concentration of particulate 137Cs did not show a clear relationship with either the mean
Keywords: 137
Fukushima Dai-ichi Nuclear Power Plant
Cs inventory or the dissolved 137Cs concentration, thus indicating that particulate and dissolved forms
accident do not effectively interact in rivers. To evaluate the contribution of the 137Cs inventory within catchment
Radiocesium areas, we analyzed relations between the 137Cs concentration and the mean 137Cs inventory over the area
River water within certain flow path lengths that were traced along the river and slope above the sampling point.
Base-flow Coefficients of determination for dissolved 137Cs concentrations were highest for the longest flow path,
Catchment inventory i.e., the whole catchment area, and lower for shorter flow paths. Coefficients of determination for par-
ticulate 137Cs concentrations were only moderately high for the shortest flow path in 2012, whereas the
values were quite low for all flow paths in 2013. These results suggest that dissolved 137Cs can originate
from a larger area of the catchment even under base-flow conditions; however, particulate 137Cs did not
show such behavior. The results also show that under base-flow conditions, dissolved and particulate
137
Cs behave independently during their transport from river catchments to the ocean.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction
After being damaged by the tsunami that followed the giant
Tohoku earthquake of 11 March 2011, the Fukushima Dai-ichi Nu-
clear Power Plant (FDNPP) of Tokyo Electric Power Company
released a large amount of radiocesium (134Cs and 137Cs) into the
surrounding environment (Nuclear Emergency Response
Headquarters, 2011). The airborne radiocesium caused contami-
nation over a broad area in northern Japan that was especially
heavy in northeastern Fukushima Prefecture, where deposition of
137
Cs on the ground was estimated to exceed 3000 kBq m
2 in the
most contaminated areas (Ministry of Education, Culture, Sports,
* Corresponding author. Tel.: þ81 175 71 1426; fax: þ81 175 71 1492.
E-mail address: sochiai@ies.or.jp (S. Ochiai).
Science and Technology (MEXT), 2012, 2013; Mikami et al., 2015;
Saito et al., 2015).
Runoff carries 137Cs deposited on the ground into rivers; thus,
the discharge rates of 137Cs from catchments to rivers and from
rivers to downstream regions are important for evaluating the
impacts of radiocesium on human health in riverine and coastal
marine ecosystems. Long-term changes in the discharge rates of
radiocesium from river catchments were studied for global fallout
and the Chernobyl accident (e.g., Helton et al., 1985; Monte, 1995;
Garcia-Sanchez, 2008). These studies have shown that the sea-
sonal- and annual-scale discharge rate depends on the deposited
amount and a transfer function, which expresses the temporal
change in the discharge fraction of the deposited amount after
deposition. Based on these models, long-term changes in discharge
have been evaluated at the global scale (e.g., Smith et al., 2004).

http://dx.doi.org/10.1016/j.jenvrad.2015.03.005
0265-931X/© 2015 Elsevier Ltd. All rights reserved.
 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95
In Fukushima rivers, previous studies have investigated the
relationship between observed fluvial discharge data and radio-
cesium deposition in the catchment to clarify the dominant factors
controlling the radiocesium discharge from the catchment (Tsuji
et al., 2014; Yoshimura et al., 2015). These studies have demon-
strated that 137Cs concentrations are related to the 137Cs inventories
in the catchments; hence, it is useful to estimate the initial
discharge from the catchment. However, these relationships may
be affected by hydrological fluctuations, especially those on shorter
timescales, because the mobility of radiocesium depends on what
forms (i.e., particulate and dissolved forms) are being transported
and these forms are impacted differently by changes in discharge.
For the annual timescale, Yamashiki et al. (2014) estimated that
84e92% of the total radiocesium discharge was transported in
particulate form in the Abukuma River, which is the largest river in
the Fukushima area. However, observations of shorter-term fluc-
tuations show that particulate 137Cs largely fluctuates in accordance
with the hydrological conditions. Ueda et al. (2013) determined
that the percentage of particulate 137Cs fluctuated from 40% during
base-flow conditions to over 90% during high-flow conditions in
two small contaminated catchments. Similar observation results
have been reported in southeastern Fukushima rivers, where the
particulate percentages were 21eca.100% (Nagao et al., 2013), and
in the Abukuma River and its tributaries, where the particulate
percentages were 51e99% (Sakaguchi et al., 2015). Based on these
results, the transport behaviors are expected to be different be-
tween particulate and dissolved 137Cs in the region. Therefore, it
would be valuable to clarify the influences of such forms on the
relationship between 137Cs concentration and catchment deposi-
tion for proper estimation of 137Cs discharge from the catchments.
With this purpose in mind, we focused our efforts on fluvial
transport of 137Cs under base-flow conditions, which is when the
differences in behaviors of dissolved and particulate 137Cs are ex-
pected to be the most clear. Specifically, this study investigated the
distribution of 137Cs in river water in eastern Fukushima during
2012e2013 and determined its relation to the catchment inventory
based on a simple model and geographic analyses. From these data,
we then attempted to clarify the transport processes of 137Cs under
base-flow conditions.
2. Samples and methods
This study targeted the main streams and tributaries of seven
river systems (Uta, Mano, Niida, Ohta, Odaka, Ukedo, and Abukuma
rivers) in eastern Fukushima Prefecture (Fig. 1), which received
large amounts of radiocesium from the FDNPP accident. The
catchment of the Abukuma River (5400 km2) is separated from the
other six by the Abukuma Mountains. River water samples were
collected at 16 sampling points during JuneeNovember 2012 and at
13 points during JuneeAugust 2013, when streamflows were at
base level (Tables 1 and 2). Samples measuring 15e20 L were
collected with a polyvinyl chloride bucket from 1 to 5 days after the
latest rainfall (>0.5 mm/h). At the time of sample collection, water
temperature, electrical conductivity, and pH were measured at the
site with a water quality meter (DS4a, Hydrolab Co., USA). Water
level data of Abukuma and Niida rivers during sampling intervals
were provided by the Ministry of Land, Infrastructure, Transport
and Tourism, Japan (MLIT) (2014a) and Fukushima Prefecture,
respectively. Precipitation data observed in Fukushima City was
provided by Japan Meteorological Agency.
After the water samples were taken to the laboratory, sub-
samples of 1e4 L were filtered through membrane filters (0.45 mm
pore size, Millipore) before the dissolved radiocesium measure-
ment. Total radiocesium (dissolved and particulate) activities were
measured in unfiltered samples. Concentrations of suspended
87
solids were obtained by weighing the residue on the filters after
drying for more than 24 h at room temperature in a container
containing desiccants (OZO-Z, OZO Kagakugiken Co., Japan). Each
water sample, whether unfiltered or filtered, was passed through a
column filled with a mixture of Powdex PAO and PCH resins (Eco-
dyn Co., USA) to separate both ionic and particulate radionuclides
from the sample (Small et al., 1975; Kimura, 1980; Yamamoto et al.,
1998). The homogenized resin was oven-dried at 80  C for more
than 48 h and then packed into a plastic case (60 mm in diameter
and 37 mm high). The activity concentrations of 134Cs (T1/2 ¼ 2.06 y)
and 137Cs (T1/2 ¼ 30.17 y) were determined by a high-purity Ge
detector (relative efficiency of 59%; Seiko EG&G, Japan). The par-
ticulate radiocesium concentration in water samples was estimated
from the difference between the filtered and unfiltered samples.
Geographic analyses of river catchments and radiocesium dis-
tributions were performed using QGIS 2.2.0 software and Funda-
mental Geospatial Data (Geospatial Information Authority of Japan,
2014) and National Land Numerical Information (MLIT, 2014b).
Catchment areas upstream from each sampling point were delin-
eated on the basis of digital elevation models (DEM) (250-m mesh
for the Abukuma River and 10-m mesh for the other rivers), and
their areas were determined from boundaries that were defined on
the basis of inclinations between the map meshes. The 137Cs
deposition data were based on airborne surveys by MEXT (2012,
2013) and Nuclear Regulation Authority (2013). Mean radio-
cesium inventories were calculated for each catchment from the
catchment areas and the 137Cs deposition data.
3. Results and discussion
137
3.1. Concentration of Cs in river water
Fig. 2 shows the changes in water levels of the Abukuma River
(Kuroiwa observatory) and Niida River (Haramachi observatory)
and hourly precipitation at Fukushima City for 10 days before each
sampling date. The locations of the observatories are indicated in
Fig. 1. Although high-flow events were observed during these in-
tervals, especially in June 2012 just before the samples were
collected, the samples used here were collected after the water
level decreased to the base level. Therefore, hydrological conditions
were similar among the sampling dates.
Tables 3 and 4 list the 134Cs and 137Cs concentrations (dissolved
and particulate) in river water in 2012 and 2013, respectively. The
activity ratio 134Cs/137Cs of both dissolved and particulate forms,
corrected to the value on the day of the FDNPP accident on 11
March 2011, was approximately 1.0, which indicates that most of
the radiocesium in the river water was derived from the FDNPP
accident rather than from global fallout. Therefore, we focus only
on the 137Cs concentrations in the following discussion. Concen-
trations of dissolved 137Cs were 0.004e0.35 (arithmetic mean 0.16;
median 0.14) Bq L
1 in 2012 and 0.010e0.39 (mean 0.11; median
0.08) Bq L
1 in 2013. Concentrations of particulate 137Cs were
0.010e0.40 (mean 0.19; median 0.12) Bq L
1 in 2012 and
0.006e0.21 (mean 0.081; median 0.052) Bq L
1 in 2013. Electrical
conductivity was 47e200 mS cm
1, which indicates that the influ-
ence of seawater was negligible at these sampling points.
The geographic distributions of dissolved and particulate 137Cs
concentrations for each sampling point are shown in Fig. 3 for 2012
and in Fig. 4 for 2013, together with the outlines of the catchments
for each point and the 137Cs inventories based on airborne surveys.
The 137Cs concentration in river water was relatively high at sam-
pling points in the Niida, Ohta, Odaka, and Ukedo river systems,
which roughly correspond to areas with high 137Cs inventories, and
relatively low in the Uta, Mano, and Abukuma river systems, where
137
Cs inventories were likewise low.
88 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95

Fig. 1. River systems studied in the eastern Fukushima area, Japan. Inset shows the regional setting of the study area. Sampling points are indicated by open circles, and the location
of the FDNPP is shown by a star. Open and closed squares indicate the locations of the water levels (Kuroiwa and Haramachi) and precipitation (Fukushima) observatories,
respectively. Topography is based on Fundamental Geospatial Data (Geospatial Information Authority of Japan, 2014) and National Land Numerical Information (MLIT, 2014b).

137
3.2. Effects of Cs inventory on the concentration in river water
To analyze the spatial variation of 137Cs concentrations in river
water, we introduced a simple model of 137Cs transport in which
the catchment above a sampling point was divided into small
segments. Water and 137Cs derived from these segments flow into
a river and finally drain from the downstream end, corresponding
to the sampling point. We assumed that the water discharge qi
from the i-th segment was proportional to the area Ai of the
segment:

Table 1
Sampling locations and water quality of river water in 2012.

Sampling point Sampling date Sampling location Water Conductivity pH SS (mg L
1) Catchment 137Cs inventory in the catchmenta
(year/month/day) temperature ( C) (mS cm
1) area (km2) (kBq m
2)

Latitude Longitude Mean Median

Ukedo River
Ukedo-U e e e e e e e e e e
Ukedo-D 2012/9/26 37 290 53.000 141 000 24.800 e 78 7.2 4.6 151.6 2138 2000
Odaka River 2012/6/26 37 330 50.600 141 000 17.3000 20.5 93 6.7 7.3 55.4 555 370
Ohta River 2012/6/26 37 350 44.600 141 000 26.800 18.2 55 6.8 22.1 70.9 1064 990
Mano River 2012/6/25 37 430 07.600 140 560 31.900 16.4 80 6.8 5.2 106.3 350 325
Uta River 2012/6/25 37 470 41.200 140 550 10.500 16.0 87 6.8 4.3 102.4 148 140

Niida River
Ganbe Dam 2012/6/26 37 380 44.200 140 410 11.700 17.1 47 6.7 3.2 3.3 715 730
Iitoi River 2012/6/26 37 390 41.000 140 460 31.300 13.4 54 6.7 4.5 47.4 839 760
Niida-U 2012/6/26 37 400 01.200 140 460 51.300 14.8 65 6.7 2.8 52.9 638 670
Hiso (Nagadoro) 2012/6/26 37 360 50.100 140 480 03.500 13.6 50 6.9 2.1 29.4 1382 1500
Mizunashi-U 2012/6/26 37 370 35.100 140 520 33.800 13.3 59 6.8 2.3 8.6 706 855
Mizunashi-D 2012/6/26 37 380 23.200 140 560 04.500 16.9 54 6.8 13.8 18.1 706 700
Niida-D 2012/6/26 37 390 01.800 140 570 32.500 17.8 69 6.9 3.3 205.5 784 710
Niida River Mouth 2012/6/26 37 380 37.200 141 000 13.400 19.0 81 6.8 8.0 258.6 692 680

Abukuma River
Abukuma River 2012/6/27 37 430 05.700 140 290 46.300 18.5 151 6.7 8.1 2909.4 86 70
Hirose River 2012/11/28 37 410 39.400 140 360 35.600 4.5 136 7.73 1.6 61.8 182 150
Ishida River 2012/11/28 37 450 53.500 140 360 41.700 5.3 122 7.8 1.4 30.8 215 150

e, Not measured.
a
Based on MEXT (2012).
 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95 89

Table 2
Sampling locations and water quality of river water in 2013.

Sampling point Sampling date Sampling location Water Conductivity pH SS (mg L
1) Catchment 137Cs inventory in the catchmenta
(year/month/day) temperature ( C) (mS cm
1) area (km2) (kBq m
2)

Latitude Longitude Mean Median

Ukedo River
Ukedo-U 2013/6/3 37 330 30.700 140 450 12.000 15.9 63 7.2 1.9 25.4 625 361
Ukedo-D 2013/6/4 37 290 53.000 141 000 24.800 18.0 83 7.1 3.4 151.6 1743 1578
Odaka River 2013/6/4 37 330 50.600 141 000 17.3000 19.3 148 7.4 24.7 55.4 445 228
Ohta River 2013/6/4 37 350 44.600 141 000 26.800 18.3 130 6.8 8.5 70.9 883 855
Mano River 2013/8/29 37 440 08.300 140 550 12.900 22.1 118 7.7 1.9 101.3 300 281
Uta River 2013/8/29 37 470 35.500 140 550 07.000 23.6 133 7.7 4.3 102.4 134 130

Niida River
Ganbe Dam 2013/7/22 37 380 44.200 140 410 11.700 22.8 56 7.6 3.0 3.3 527 522
Iitoi River e e e e e e e e e e
Niida-U 2013/8/30 37 400 01.500 140 460 51.300 21.0 87 7.1 3.6 52.9 480 491
Hiso (Nagadoro) 2013/8/30 37 360 36.100 140 470 53.700 20.2 62 7.2 2.1 27.7 1080 1102
Mizunashi-U e e e e e e e e e e
Mizunashi-D 2013/8/29 37 380 23.200 140 560 04.500 22.0 86 7.0 0.9 18.1 574 551
Niida-D 2013/8/29 37 400 05.600 140 550 50.800 22.4 84 7.5 3.8 196.1 648 574
Niida River Mouth e e e e e e e e e e

Abukuma River
Abukuma River 2013/8/30 37 430 05.600 140 290 46.400 24.9 200 7.6 7.8 2909.4 69b 57b
Hirose River 2013/8/30 37 410 52.200 140 360 33.000 23.1 163 7.6 6.9 65.6 149 137
Ishida River e e e e e e e e e e

e, Not measured.
a
Based on Nuclear Regulation Authority (2013).
b
Based on MEXT (2013).

and particulate 137Cs are originating from every part of a catchment

qi ¼ k1 Ai ; (1) and do not interact each other during the transport.
We derived the catchment-average 137Cs inventory (arithmetic
where k1 is the proportionality constant. Removal flux of radionu- mean and median) by using all data points from the airborne sur-
clides from the catchment was assumed to be proportional to the vey inside each catchment, which resulted in a wide range of mean
inventory (e.g., Garcia-Sanchez, 2008). Based on this assumption, values: 0.086e2.1 MBq m
2 and 0.068e1.8 MBq m
2 in 2012 and
the flux of dissolved or particulate 137Cs fi from the i-th segment 2013, respectively (Tables 1 and 2). The arithmetic mean of the
was expressed as follows: inventory was larger than the median for the Ukedo-U, Ukedo-D,
and Odaka River sites, which is indicative of a skewed statistical
fi ¼ k2 Ii Ai ; (2)
distribution; this was due to the extremely high points in these
where k2 and Ii are the proportionality constant and inventory of catchments. Although the median is a suitable descriptor to avoid
the segment, respectively. Assuming that the river water and 137Cs biased influence from a few extreme values, the concentrations in
are conserved throughout the model, water discharge Q and 137Cs river water were related to the arithmetic mean of the inventory in
discharge at the sampling point CmQ are equal to the sum of com- the above model (Eq. (6)). Therefore, we also used the mean values
ponents from the segments making up the catchment: in the following discussion.
Fig. 5a shows that dissolved 137Cs concentration in river water is
X X
Q¼ qi ¼ k1 Ai (3) strongly correlated with catchment-mean 137Cs inventory, with
r2 ¼ 0.90 (P < 0.001) and 0.92 (P < 0.001) for 2012 and 2013,
X X respectively. We can conclude that the concentration of dissolved
Cm Q ¼ fi ¼ k2 Ai Ii ; (4) 137
Cs of river water under base-flow conditions was primarily
determined by the catchment-mean 137Cs inventory. Those results
where Cm is the 137Cs concentration at the sampling point. Based on are consistent with the conservation of dissolved 137Cs throughout
Eqs. (3) and (4), the dissolved or particulate 137Cs concentration at the river. This correlation between dissolved concentration and
the sampling point Cm can be rewritten as inventory of 137Cs has been reported previously. Tsuji et al. (2014)
P P related dissolved radiocesium concentrations to radiocesium in-
k2 Ai Ii k AI
Cm ¼ ¼ P 2 i i: (5) ventories in the catchment of the Abukuma River and its tributaries
Q k1 Ai in 2013. Yoshimura et al. (2015) documented the relationship be-
tween dissolved or particulate 137Cs concentrations and inventories
If the constants k1 and k2 are uniform in a catchment, Eq. (5) is in catchments of eastern Fukushima rivers in 2012. However, those
finally rewritten as papers reported data for only a single year. Our results show that
P this relationship prevailed in eastern Fukushima rivers during the
k2 Ai Ii
Cm ¼ P ; (6) two years after the accident. The slope of the regression line be-
k1 Ai
tween concentrations and inventories, which reflects the mobility
of dissolved 137Cs from the catchment, was 2.68  10
7 m2 L
1 in
where SAiIi/SAi is equivalent to the mean inventory across a
2012 and slightly less at 2.14  10
7 m2 L
1 in 2013 (Fig. 5a). A
catchment. This equation indicates that the 137Cs concentration is
parallelism test between the two regression lines indicated that
proportional to the catchment-mean inventory if water and 137Cs
parallelism could not be rejected (P > 0.05), i.e., differences
are conserved throughout the river (Eqs. (3) and (4)), i.e., dissolved
90 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95

Fig. 2. Temporal changes in water levels of the Abukuma River at the Kuroiwa Observatory (thin line) and the Niida River at the Haramachi Observatory (bold line). Hourly
precipitation data at Fukushima City (bar graph) for 10 days before each sampling date are also shown. Open triangles indicate the sampling dates.

Table 3
134 137
Activity concentrations of Cs and Cs of river water in 2012.

Sampling point Activity concentrationa 134

Cs/137Csb Particulate to total 137
Cs (%)

1
1
1
Dissolved (Bq L ) Particulate (Bq L ) Particulate (Bq g -SS)
134 137 134 137 134 137
Cs Cs Cs Cs Cs Cs Dissolved Particulate

Ukedo River
Ukedo-U e e e e e e e e e
Ukedo-D 0.350 ± 0.0066 0.590 ± 0.0079 0.300 ± 0.0140 0.480 ± 0.0240 65.0 ± 3.1 110.0 ± 5.2 0.96 ± 0.02 1.01 ± 0.07 44
Odaka River 0.093 ± 0.0018 0.140 ± 0.0020 0.066 ± 0.0075 0.100 ± 0.0091 9.1 ± 1.0 14.2 ± 1.3 1.00 ± 0.02 0.99 ± 0.14 40
Ohta River 0.220 ± 0.0064 0.340 ± 0.0110 0.400 ± 0.0110 0.600 ± 0.0160 18.0 ± 0.5 27.1 ± 0.7 0.97 ± 0.04 1.00 ± 0.04 64
Mano River 0.041 ± 0.0024 0.062 ± 0.0023 0.017 ± 0.0037 0.022 ± 0.0039 3.2 ± 0.7 4.1 ± 0.8 0.99 ± 0.07 1.16 ± 0.33 27
Uta River 0.020 ± 0.0012 0.030 ± 0.0015 0.012 ± 0.0024 0.017 ± 0.0028 2.7 ± 0.6 3.8 ± 0.7 1.00 ± 0.08 1.06 ± 0.27 36

Niida River
Ganbe Dam 0.075 ± 0.0022 0.120 ± 0.0037 0.081 ± 0.0053 0.100 ± 0.0068 25.6 ± 1.7 32.1 ± 2.1 0.94 ± 0.04 1.22 ± 0.11 45
Iitoi River 0.082 ± 0.0028 0.120 ± 0.0033 0.110 ± 0.0073 0.200 ± 0.0086 24.7 ± 1.6 44.9 ± 1.9 1.03 ± 0.04 0.83 ± 0.07 63
Niida-U 0.097 ± 0.0048 0.150 ± 0.0056 0.150 ± 0.0060 0.230 ± 0.0071 52.5 ± 2.1 80.4 ± 2.5 0.97 ± 0.06 0.98 ± 0.05 61
Hiso (Nagadoro) 0.150 ± 0.0044 0.240 ± 0.0051 0.250 ± 0.0120 0.390 ± 0.0150 121.5 ± 5.8 186.9 ± 7.2 0.94 ± 0.03 0.96 ± 0.06 63
Mizunashi-U 0.160 ± 0.0061 0.230 ± 0.0069 0.110 ± 0.0110 0.180 ± 0.0130 50.0 ± 5.0 79.1 ± 5.6 1.04 ± 0.05 0.92 ± 0.11 44
Mizunashi-D 0.110 ± 0.0039 0.170 ± 0.0044 0.240 ± 0.0081 0.370 ± 0.0097 17.6 ± 0.6 27.0 ± 0.7 0.97 ± 0.04 0.97 ± 0.04 69
Niida-D 0.086 ± 0.0034 0.130 ± 0.0043 0.050 ± 0.0056 0.070 ± 0.0069 15.3 ± 1.7 21.5 ± 2.1 0.99 ± 0.05 1.07 ± 0.16 35
Niida River Mouth 0.075 ± 0.0031 0.140 ± 0.0055 0.098 ± 0.0075 0.130 ± 0.0100 12.2 ± 0.9 16.5 ± 1.2 0.80 ± 0.05 1.13 ± 0.12 48

Abukuma River
Abukuma River 0.015 ± 0.0009 0.025 ± 0.0010 0.029 ± 0.0025 0.046 ± 0.0030 3.6 ± 0.3 5.7 ± 0.4 0.90 ± 0.06 0.95 ± 0.10 65
Hirose River 0.004 ± 0.0005 0.008 ± 0.0006 0.010 ± 0.0019 0.014 ± 0.0023 6.6 ± 1.2 8.6 ± 1.5 0.84 ± 0.12 1.22 ± 0.31 64
Ishida River 0.004 ± 0.0006 0.008 ± 0.0008 0.013 ± 0.0014 0.022 ± 0.0022 9.2 ± 1.1 16.0 ± 1.6 0.99 ± 0.16 1.01 ± 0.15 73

e, Not measured.
a
Corrected to the value at sampling date.
b
Corrected to the value on 11 March 2011.
 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95 91

Table 4
134 137
Activity concentrations of Cs and Cs of river water in 2013.

Sampling point Activity concentrationa 134

Cs/137Csb Particulate to total 137
Cs (%)

Dissolved (Bq L
1) Particulate (Bq L
1) Particulate (Bq g
1-SS)

134 137 134 137 134 137
Cs Cs Cs Cs Cs Cs Dissolved Particulate

Ukedo River
Ukedo-U 0.030 ± 0.0016 0.063 ± 0.0026 0.023 ± 0.0023 0.047 ± 0.0040 12.0 ± 1.2 24.0 ± 2.0 0.96 ± 0.06 0.98 ± 0.13 43
Ukedo-D 0.190 ± 0.0010 0.390 ± 0.0020 0.099 ± 0.0019 0.210 ± 0.0038 29.0 ± 0.6 63.0 ± 1.1 0.98 ± 0.01 0.95 ± 0.03 35
Odaka River 0.035 ± 0.0012 0.071 ± 0.0022 0.068 ± 0.0038 0.150 ± 0.0073 2.8 ± 0.2 6.1 ± 0.3 0.99 ± 0.05 0.91 ± 0.07 68
Ohta River 0.086 ± 0.0008 0.180 ± 0.0016 0.039 ± 0.0023 0.083 ± 0.0045 4.6 ± 0.3 9.7 ± 0.5 0.96 ± 0.01 0.95 ± 0.08 32
Mano River 0.025 ± 0.0013 0.053 ± 0.0022 0.021 ± 0.0024 0.040 ± 0.0038 29.0 ± 3.3 55.0 ± 5.3 1.02 ± 0.07 1.14 ± 0.17 43
Uta River 0.007 ± 0.0007 0.016 ± 0.0008 0.004 ± 0.0010 0.006 ± 0.0014 2.2 ± 0.6 3.7 ± 0.9 1.00 ± 0.10 1.30 ± 0.47 27

Niida River
Ganbe Dam 0.056 ± 0.0017 0.130 ± 0.0033 0.017 ± 0.0031 0.030 ± 0.0062 5.7 ± 1.0 9.9 ± 2.0 0.90 ± 0.04 1.19 ± 0.33 19
Iitoi River e e e e e e e e e
Niida-U 0.036 ± 0.0008 0.081 ± 0.0016 0.037 ± 0.0036 0.080 ± 0.0069 28.0 ± 2.7 59.0 ± 5.1 0.96 ± 0.03 1.00 ± 0.13 50
Hiso (Nagadoro) 0.076 ± 0.0017 0.160 ± 0.0032 0.054 ± 0.0052 0.110 ± 0.0098 67.0 ± 6.5 130.0 ± 12.0 1.03 ± 0.03 1.07 ± 0.14 41
Mizunashi-U e e e e e e e e e
Mizunashi-D 0.055 ± 0.0015 0.130 ± 0.0028 0.007 ± 0.0025 0.013 ± 0.0049 34.0 ± 11.0 59.0 ± 23.0 0.92 ± 0.03 1.22 ± 0.62 9
Niida-D 0.049 ± 0.0014 0.110 ± 0.0026 0.090 ± 0.0049 0.200 ± 0.0094 63.0 ± 3.4 140.0 ± 6.6 0.97 ± 0.04 0.98 ± 0.07 65
Niida River Mouth e e e e e e e e e

Abukuma River
Abukuma River 0.006 ± 0.0005 0.015 ± 0.0008 0.016 ± 0.0022 0.028 ± 0.0035 6.9 ± 0.9 12.0 ± 1.5 0.87 ± 0.08 1.24 ± 0.23 65
Hirose River 0.005 ± 0.0005 0.010 ± 0.0007 0.021 ± 0.0021 0.052 ± 0.0039 10.0 ± 1.0 25.0 ± 1.9 1.05 ± 0.12 0.88 ± 0.11 84
Ishida River e e e e e e e e e

e, Not measured.
a
Corrected to the value at sampling date.
b
Corrected to the value on March 11, 2011.

Fig. 3. Distribution of the 137Cs inventory on 28 June 2012 (colors; from MEXT, 2012) and dissolved and particulate 137
Cs concentrations (white and black bars, respectively)
measured in 2012. Open circles indicate sampling points. Black lines outline the catchments for the sampling points.
92 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95

Fig. 4. Distribution of the 137Cs inventory on 28 September 2013 (colors; from MEXT, 2013; Nuclear Regulation Authority, 2013) and dissolved and particulate 137Cs concentrations
(white and black bars, respectively) measured in 2013. Open circles indicate sampling points. Black lines outline the catchments for the sampling points.

between the two slopes were not significant. However, longer ob-
servations will be required to confirm whether this tendency
changes in Fukushima rivers in the future.
The geology, land use, and vegetation cover in the catchment
may affect the permeability of rock to groundwater, which may be
related to the constants k1 and k2 here. The geology of the Abukuma
Mountains (Fig. 1) mainly consists of granite, and sedimentary
rocks are distributed in the plain along the coast, indicating that
these conditions may be different within a given catchment.
Although the significant relationship between the dissolved 137Cs
concentration and mean 137Cs inventory implies that influences of
these conditions on dissolved 137Cs transportation are relatively
small, detailed investigations will be needed in the future to clarify
this issue.
Concentrations of particulate 137Cs in river water and in sus-
pended solids (SS) were found to have weak correlations with the
catchment-mean 137Cs inventory in 2012, and no significant cor-
relations in 2013 (Fig. 5b and c). Those poor correlations may be
attributed to the fact that our samples were collected during base-
flow conditions, when SS mobility was low. In periods of low
mobility, base flow does not transport particulate 137Cs from distant
parts of a catchment; thus, the assumption in Eq. (4) that particu-
late 137Cs is equally originated from the every part of a catchment is
not valid. Therefore, 137Cs concentrations in SS were not correlated
with the catchment-mean 137Cs inventory (Fig. 5c). The low
mobility of particulate 137Cs was reflected in its relatively small
contribution to the total 137Cs concentration, which averaged 49%
with a range of 9e84% (Tables 3 and 4). In previous studies of
Fukushima rivers, in which data during high-flow conditions were
included, the mean percentage of particulate 137Cs in 2011 was
reported as 74% in the Hiso River and 82% in the Wariki River, both
of which are tributaries of the Niida River (Ueda et al., 2013).
The relationship between 137Cs concentrations dissolved in river
water and in SS (Fig. 5d) reflects the degree of interaction (sorption/
desorption) between dissolved and particulate 137Cs in river water.
As no clear relationship was found (r2 ¼ 0.35 and 0.10 for 2012 and
2013, respectively), it appears that dissolved and particulate 137Cs
in river water interact quite weakly under base-flow conditions.
3.3. Contribution of 137
Cs inventory in portions of catchments
Our previous discussion suggests that dissolved 137Cs can be
equally contributed from every part of a catchment, while partic-
ulate 137Cs cannot originate from every part under base-flow con-
ditions. To test these results statistically, we tried to evaluate the
contributions of different portions of catchments to 137Cs concen-
trations. Because a certain point in the catchment contributes to the
sampling point via water flow, the distance between them should
be measured along the flow path instead of by the linear distance.
Therefore, we introduced flow path lengths into the analysis, and
 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95
Fig. 5. Relationships between the catchment-mean 137Cs inventory and (a) dissolved 137Cs concentration, (b) particulate 137Cs concentration, and (c) 137Cs concentration in SS. Open
and closed circles indicate the values in 2012 and 2013, respectively. (d) Relationship between the dissolved 137Cs concentration and 137Cs concentration in SS. Regression lines are
dashed for 2012 and solid for 2013.
these lengths were traced along not only the river course but also
the course on the slope (Fig. 6). As mentioned earlier, the catch-
ments were divided into meshes to determine their catchment
boundaries. The flow direction (slope aspect) at each mesh was
calculated using a DEM. The length of the flow path from each of
these meshes to the sampling point was then calculated based on
the flow direction in the catchment (Mita sov
a and Hofierka, 1993),
and these values were normalized against the maximum flow path
length (the flow path to the most upstream end) in the catchment.
Fig. 6 shows the normalized flow path lengths for some points and
contour lines (white line).
We derived mean 137Cs inventories over the area within the
certain normalized flow path lengths. The correlations between the
137
Cs inventory and dissolved 137Cs concentration (Fig. 7a) had
lower coefficients of determination (r2) when the flow path was
shorter, suggesting that dissolved 137Cs concentration was not
come from only catchment area near sampling point. The value of r2
increases with the flow path length and has the maximum value at
1.0 of flow path length: i.e., the whole catchment area of the
sampling point. Thus, the mean inventory for the whole catchment
showed the best correlation with the dissolved 137Cs concentration.
On the other hand, the analogous exercise for 137Cs concentra-
tion in SS (Fig. 7b) resulted in higher r2 values for shorter flow path
lengths; r2 decreased with path length in 2012 and showed no clear
relationship with flow path length in 2013. Thus, particulate 137Cs
93
was more closely related to inventory near the sampling point than
to inventory in the whole catchment in 2012, although this trend
disappeared in 2013.
3.4. Transport processes of 137
Cs under base-flow conditions
Based on the above results, transport processes of dissolved and
particulate 137Cs under base-flow conditions can be estimated as
follows. During high-flow events, particulate 137Cs supplied from
slope erosion is transported downstream for a short time as well as
dissolved 137Cs. During such conditions, the concentrations are
expected to be correlated to the catchment inventory. Actually,
Yoshimura et al. (2015) reported that particulate 137Cs concentra-
tions collected using a time-integrated SS sampler were correlated
with the inventories of Fukushima river catchments; the sampler
collects combined material during base- and high-flow conditions
(Phillips et al., 2000).
After water discharge decreases, a part of the suspended sedi-
ment is deposited and temporally stored in the riverbed. Under
base-flow conditions, fine particles in the riverbed are the main
sources of suspended sediment (Grimshaw and Lewin, 1980), and
sediment supply from slope erosion to the channel is negligible.
Because of the low transport force of river flow during such con-
ditions, particulate 137Cs concentrations originate from areas
within close sections of the river (as shown in Fig. 7b). At such
94 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95
Fig. 6. Conceptual illustration of the flow path length for an example catchment. The map indicates the catchment area above the sampling point. The flow paths are traced along
the slope (dashed line) and river (solid line) from each point (open circle) to the sampling point. Small numbers beside the each point indicate the normalized flow path lengths for
each point. Flow path lengths for the sampling point and the most upstream end are 0.0 and 1.0, respectively. White lines indicate contours of the normalized flow path length.
times, the correlation with the catchment inventory became weak.
As time passes after the last discharge event, stored sediment
sources decrease from the riverbed. Accordingly, the weaker cor-
relations of particulate 137Cs in 2013 may have resulted from the
Fig. 7. Coefficients of determination r2 between 137Cs inventory averaged over the area
within segments of the normalized flow path above the sampling point and (a) dis-
solved 137Cs concentration and (b) 137Cs concentration in SS for 2012 and 2013.
longer base-flow before sampling during this year (Fig. 2b). Dis-
solved 137Cs can originate from a larger area of the catchment (as
shown in Fig. 7a), and hence, the correlation with the catchment
inventory was high even under base-flow conditions. These data
show that dissolved and particulate 137Cs behaved independently
under base-flow conditions.
Our results regarding the transport processes of 137Cs under
base-flow conditions are based on observations and a simple model
explaining the relationship between the 137Cs concentration and
catchment inventory. In the future, investigations of the spatial
distribution of 137Cs concentrations and their relation with riverbed
sediments, soil, and geological features along the river may provide
more clear evidence of the transport processes taking place in this
region.
4. Conclusions
We investigated the distribution of radiocesium in river water in
eastern Fukushima and the factors controlling concentrations un-
der base-flow conditions. Observations of dissolved and particulate
137
Cs concentrations at 16 sampling points in seven river systems
during 2012 and 2013 lead to the following conclusions.
Dissolved 137Cs concentrations had a significant correlation with
the mean 137Cs inventory in the catchment above each sampling
point during 2012 and 2013. On the other hand, the particulate
137
Cs concentrations were not correlated with either the 137Cs in-
ventory or the dissolved 137Cs concentration. Those results show
that 1) the concentration of dissolved 137Cs under base-flow con-
ditions is dominated by the 137Cs inventory of the catchment, which
is conserved during transport by the river, and 2) there is little or no
interaction between particulate and dissolved forms of
radiocesium.
We examined relations between the 137Cs concentration in river
water and the mean 137Cs inventory over the area within the certain
flow path lengths above the sampling point to evaluate the con-
tributions of different portions of catchments to 137Cs
 S. Ochiai et al. / Journal of Environmental Radioactivity 144 (2015) 86e95
concentrations. Coefficients of determination for dissolved 137Cs
concentrations were highest for the longest flow path, i.e., the
whole catchment area, and weaker with shorter flow paths. Co-
efficients of determination for particulate 137Cs concentrations
were only moderately high for the shortest flow path in 2012 and
quite low throughout the flow path in 2013. These results suggest
that particulate 137Cs under base-flow conditions is mainly derived
from riverbed sediments temporally stored near the sampling point
after the last high-flow event. Dissolved 137Cs can originate from a
larger area of the catchment even under base-flow conditions.
Additionally, the results show that dissolved and particulate 137Cs
concentrations under base-flow conditions behave independently
during transport in catchments.
Acknowledgments
We are grateful to H. Kakimoto and S. Takaya (Zax Co. Ltd.) for
their technical help in sample pretreatment. This work was per-
formed under a contract with the government of Aomori Prefec-
ture, Japan.
References
Garcia-Sanchez, L., 2008. Watershed wash-off of atmospherically deposited radio-
nuclides: review of the fluxes and their evolution with time. J. Environ. Radi-
oact. 99, 563e573.
Geospatial Information Authority of Japan, 2014. Fundamental Geospatial Data.
Available at: http://www.gsi.go.jp/kiban/index.html (in Japanese).
Grimshaw, D.L., Lewin, J., 1980. Source identification for suspended sediments.
J. Hydrol. 47, 151e162.
Helton, J., Muller, A., Bayer, A., 1985. Contamination of surface-water bodies after
reactor accidents by the erosion of atmospherically deposited radionuclides.
Health Phys. 48, 757e771.
Kimura, T., 1980. A simple and rapid method of collecting radionuclides from rain
water. Anal. Chim. Acta 120, 419e422.
Mikami, S., Maeyama, T., Hoshide, Y., Sakamoto, R., Sato, S., Okuda, N.,
Demongeot, S., Gurriaran, R., Uwamino, Y., Kato, H., Fujiwara, M., Sato, T.,
Takemiya, H., Saito, K., 2015. Spatial distributions of radionuclides deposited
onto ground soil around the Fukushima Dai-ichi Nuclear Power Plant and their
temporal change until December 2012. J. Environ. Radioact. 139, 320e343.
Ministry of Education, Culture, Sports, Science and Technology (MEXT), 2012. Re-
sults of the (i) Fifth Airborne Monitoring Survey and (ii) Airborne Monitoring
Survey Outside 80km from the Fukushima Dai-ichi NPP. Available at: http://
radioactivity.nsr.go.jp/en/contents/6000/5790/24/203_0928_14e.pdf.
Ministry of Education, Culture, Sports, Science and Technology (MEXT), 2013. (i)
Results of the Sixth Airborne Monitoring and (ii) Airborne Monitoring out of the
80km Zone of Fukushima Dai-ichi NPP. Available at: http://radioactivity.nsr.go.
jp/en/contents/7000/6099/24/203_e_0301_18.pdf.
Ministry of Land, Infrastructure, Transport and Tourism, Japan (MLIT), 2014a. Water
Information System. Available at: http://www1.river.go.jp/ (in Japanese).
95
Ministry of Land, Infrastructure, Transport and Tourism, Japan (MLIT), 2014b. Na-
tional Land Numerical Information. Available at: http://nlftp.mlit.go.jp/ksj-e/
index.html.
Mitasova
, H., Hofierka, J., 1993. Interpolation by regularized spline with tension: II.
Application to terrain modeling and surface geometry analysis. Math. Geol. 25,
657e669.
Monte, L., 1995. Evaluation of radionuclide transfer functions from drainage basins
of fresh water systems. J. Environ. Radioact. 26, 71e82.
Nagao, S., Kanamori, M., Ochiai, S., Tomihara, S., Fukushi, K., Yamamoto, M., 2013.
Export of 134Cs and 137Cs in the Fukushima river systems at heavy rains by
Typhoon Roke in September 2011. Biogeosci. Dis. 10, 2767e2790.
Nuclear Emergency Response Headquarters, Government of Japan, 2011. Report of
the Japanese Government to the IAEA Ministerial Conference on Nuclear Safety
- The Accident at TEPCO's Fukushima Nuclear Power Stations. Available at:
http://japan.kantei.go.jp/kan/topics/201106/iaea_houkokusho_e.html.
Nuclear Regulation Authority, Japan, 2013. Monitoring air dose rates from a series of
aircraft surveys 30 months after the Fukushima Daiichi NPS accident. Available
at: http://www.nsr.go.jp/english/library/data/special-report_0114.pdf.
Phillips, J.M., Russell, M.A., Walling, D.E., 2000. Time-integrated sampling of fluvial
suspended sediment: a simple methodology for small catchments. Hydrol.
Process 14, 2589e2602.
Saito, K., Tanihata, I., Fujiwara, M., Saito, T., Shimoura, S., Otsuka, T., Onda, Y.,
Hoshi, M., Ikeuchi, Y., Takahashi, F., Kinouchi, N., Saegusa, J., Seki, A.,
Takemiya, H., Shibata, T., 2015. Detailed deposition density maps constructed by
large-scale soil sampling for gamma-ray emitting radioactive nuclides from the
Fukushima Dai-ichi Nuclear Power Plant accident. J. Environ. Radioact. 139,
308e319.
Sakaguchi, A., Tanaka, K., Iwatani, H., Chiga, H., Fan, Q., Onda, Y., Takahashi, Y., 2015.
Size distribution studies of 137Cs in river water in the Abukuma Riverine system
following the Fukushma Dai-ichi Nuclear Power Plant accident. J. Environ.
Radioact. 139, 379e389.
Small, H., Stevens, T.S., Bauman, W.C., 1975. Novel ion exchange chromatographic
method using conductometric detection. Anal. Chem. 47, 1801e1809.
Smith, J.T., Wright, S.M., Cross, M.A., Monte, L., Kudelsky, A.V., Saxe n, R.,
Vakulovsky, S.M., Timms, D.N., 2004. Global analysis of the riverine transport of
90 137
Sr and Cs. Environ. Sci. Technol. 38, 850e857.
Tsuji, H., Yasutaka, T., Kawabe, Y., Onishi, T., Komai, T., 2014. Distribution of dissolved
and particulate radiocesium concentrations along rivers and the relations be-
tween radiocesium concentration and deposition after the nuclear power plant
accident in Fukushima. Water Res. 60, 15e27.
Ueda, S., Hasegawa, H., Kakiuchi, H., Akata, N., Ohtsuka, Y., Hisamatsu, S., 2013.
Fluvial discharges of radiocaesium from watersheds contaminated by the
Fukushima Dai-ichi Nuclear Power Plant accident, Japan. J. Environ. Radioact.
118, 96e104.
Yamamoto, M., Kofuji, H., Shiraishi, K., Igarashi, Y., 1998. An attempt to evaluate dry
deposition velocity of airborne 210Pb in a forest ecosystem. J. Radioanal. Nucl.
Chem. 227, 81e87.
Yamashiki, Y., Onda, Y., Smith, H.G., Blake, W.H., Wakahara, T., Igarashi, Y.,
Matsuura, Y., Yoshimura, K., 2014. Initial flux of sediment-associated radio-
cesium to the ocean from the largest river impacted by Fukushima Daiichi
Nuclear Power Plant. Sci. Rep. 4, 3714,. http://dx.doi.org/10.1038/srep03714.
Yoshimura, K., Onda, Y., Sakaguchi, A., Yamamoto, M., Matsuura, Y., 2015. An
extensive study of the concentrations of particulate/dissolved radiocaesium
derived from the Fukushima Dai-ichi Nuclear Power Plant accident in various
river systems and their relationship with catchment inventory. J. Environ.
Radioact. 139, 370e378.