Episode Analysis of Deposition of Radiocesium from the FukushimaDaiichi Nuclear Power Plant AccidentYu Morino,* Toshimasa Ohara, Mirai Watanabe, Seiji Hayashi, and Masato Nishizawa

Center for Regional Environment Research, National Institute for Environmental Studies, 16-2, Onogawa, Tsukuba, Ibaraki,305-8506, Japan

*S Supporting Information

ABSTRACT: Chemical transport models played key roles in under-standing the atmospheric behaviors and deposition patterns ofradioactive materials emitted from the Fukushima Daiichi nuclearpower plant after the nuclear accident that accompanied the greatTohoku earthquake and tsunami on 11 March 2011. However, modelresults could not be sufficiently evaluated because of limitedobservational data. We assess the model performance to simulate thedeposition patterns of radiocesium (137Cs) by making use of airbornemonitoring survey data for the first time. We conducted ten sensitivitysimulations to evaluate the atmospheric model uncertainties associatedwith key model settings including emission data and wet depositionmodules. We found that simulation using emissions estimated with aregional-scale (∼500 km) model better reproduced the observed 137Csdeposition pattern in eastern Japan than simulation using emissionsestimated with local-scale (∼50 km) or global-scale models. In addition, simulation using a process-based wet deposition modulereproduced the observations well, whereas simulation using scavenging coefficients showed large uncertainties associated withempirical parameters. The best-available simulation reproduced the observed 137Cs deposition rates in high-deposition areas(≥10 kBq m−2) within 1 order of magnitude and showed that deposition of radiocesium over land occurred predominantlyduring 15−16, 20−23, and 30−31 March 2011.

■ INTRODUCTION

Enormous quantities of radionuclides were released into theatmosphere after the nuclear accident at the Fukushima Daiichinuclear power plant (FDNPP) on 11 March 2011.1−3 Toestimate the atmospheric behavior of the radionuclides,particularly iodine-131 (131I), and cesium-137 (137Cs), atmos-pheric modeling studies were conducted over local,4 region-al,5−7 and global2,8 scales. Observational data played a criticalrole in the assessment of model performance. Previous modelsimulations used data primarily from surface monitoring ofatmospheric concentrations and deposition. However, monitor-ing of atmospheric deposition did not start until 18 March2011, and thus model performance could not be evaluated forthe period before 18 March, the period when the largestradiocesium emissions presumably occurred.1

Recently, data from an airborne monitoring surveyconducted by the Ministry of Education, Culture, Sports,Science, and Technology became available.9 In this study, wemade the first use of these data to assess the performance ofmodels of radiocesium deposition patterns. The airborne surveydata have two advantages for the evaluation of chemicaltransport models. One is the high spatial resolution. Combinedanalysis of this highly resolved observational data withmodeling results helps to reveal the detailed mechanism ofradiocesium deposition. The other advantage is the wide data

coverage over eastern Japan. Total radiocesium deposition overland in Japan can be estimated from the airborne monitoringdata, and the fact that the simulated total deposition ofradiocesium over land can be validated adds credibility to the137Cs budget analysis.In addition, by using the high-resolution observational data,

we can assess the suitability of different model settings forcritical modules. Because there was some uncertainty inselecting model settings (e.g., emission and wet deposition),we conducted nine sensitivity simulations and assessed themodel performance. These simulations gave us insight intouncertainties and plausible model settings for simulation ofradiocesium deposition, although a comprehensive uncertaintyanalysis is beyond the scope of this study.We then used the validated model to assess the radiocesium

budget and deposition mechanisms. Budget analysis andepisode analysis had previously been reported.5−7 However,the simulated deposition patterns of these analyses were notverified, owing to lack of observational data. Our analysisexplains the mechanism by which the observed high-deposition

Received: November 12, 2012Revised: February 1, 2013Accepted: February 7, 2013Published: February 7, 2013

Article

pubs.acs.org/est

© 2013 American Chemical Society 2314 dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−2322

areas formed, and thus it may be the basis for future studies onradiocesium deposition.

■ METHODOLOGYSimulation Models. We simulated distributions of 137Cs by

using the Weather Forecast and Research Model version 3.110

and a 3D chemical transport model, Models-3 CommunityMultiscale Air Quality (CMAQ),11 for the period from 10March to 20 April 2011. We conducted one standard-casesimulation (STD) and nine sensitivity simulations assummarized in Table 1. The model domain is shown in Figure

1. The prefectures are indicated by numbers in Figure 1 (e.g.,P1 for Iwate Prefecture and P5 for Fukushima Prefecture). Thebasic model settings are described in our previous article,5

although the horizontal resolution of the model used in thisstudy (3 km) was finer than that in our previous study (6 km),and emission data were updated to the latest estimate from theJapan Atomic Energy Agency (JAEA).7

We conducted simulations with three sets of emission data:the estimates from JAEA,7 the Norwegian Institute for AirResearch (NILU),2 and Tokyo Electric Power Company(TEPCO)3 (Figure S2 of the Supporting Information). Allthree emission estimates are based on inversion methods usingsimulation models and observational data. The JAEA analysiscombined local- and regional-scale models, whereas NILU useda global-scale model, and TEPCO used a local-scale model. Themodel used for the JAEA estimate used a mesh size of 3 km,and that of TEPCO was 1 km. NILU used an objectiveinversion, whereas JAEA and TEPCO estimated release ratesfrom a combination of observational data and atmosphericsimulations under the assumption of a unit release rate. JAEAfirst used observed concentrations of radiocesium in air at 10measurement sites1 and then modified their estimate by usingthe air dose rate at three monitoring sites in FukushimaPrefecture4 and surface deposition rates at 19 monitoring sitesover eastern Japan.7 NILU used air concentrations at 45monitoring stations (6 in Japan, 5 in the northern PacificOcean, 12 on the North American continent, and 12 inWestern Europe) and surface deposition rates at 46 monitoringstations over Japan7 and in Tokai-mura in Ibaraki Prefecture.TEPCO used the air dose rate measured from a monitoring carthat moved around the FDNPP, and estimated emission ratesof 131I, 134Cs, 137Cs, and noble gases by assuming emissionratios for the respective nuclides. TEPCO estimated emissionrates during 12−31 March 2011, and thus the simulation periodfor the EM3 simulation is 10−31 March.We also compared three wet deposition settings. The wet

deposition modules are described in detail in section S1 of theSupporting Information and briefly described below. In CMAQv4.6, wet deposition rates of accumulation-mode aerosols arecalculated by considering washout time, which is calculatedfrom the ratio of the water content of precipitation and that ofclouds.11 The wet deposition module is process-based, and wetdeposition amounts of aerosols calculated with CMAQ havebeen validated in several previous studies.12,13 We alsoconducted a simulation with the wet deposition module ofthe JAEA model (WD2 case).7 In that model, wet depositionrates are calculated using a scavenging coefficient (Λ), which isa function of the precipitation rate. This wet deposition moduleis an empirical module with fitting parameters included.Simulation with Λ of the JAEA model multiplied by a factorof 10 (WD3 case) was also conducted as shown later.Because the observed diameter of particulate radiocesium

differs among studies,14 in our previous study we set the meandiameter and standard deviation to 1 μm and 1.1, respectively.5

Recently, observed size distributions of radiocesium havebecome available.15 Activity size distributions of 134Cs and137Cs in aerosols were measured at Tsukuba, a city 170 kmsouthwest of the FDNPP, during 28 April−12 May and 12−26May 2011. Means and standard deviations of 137Cs aerosoldiameters during the two periods were derived after the datawere fit to log-normal distributions. We used the derived meandiameter (0.65 μm) and standard deviation (1.35) in thesensitivity simulation (DD2 case). Aerosol size distributionschange during transport because deposition rates differ byparticle size. However, we did not consider this change in thissimulation.

Table 1. Setup Parameters Used for Ten Model Simulations

simulation emissionsa wet depositionb particle diameter

STD JAEA7 CMAQ11 1 μmEM2 NILU2 CMAQ 1 μmEM3 TEPCO3 CMAQ 1 μmWD2 JAEA Scav. coeff.7 1 μmE2W2 NILU Scav. coeff. 1 μmE3W2 TEPCO Scav. coeff. 1 μmWD3 JAEA Scav. coeff. × 10 1 μmE2W3 NILU Scav. coeff. × 10 1 μmE3W3 TEPCO Scav. coeff. × 10 1 μmDD2 JAEA CMAQ Kaneyasu et al.15

aJAEA, Japan Atomic Energy Agency; NILU, the Norwegian Institutefor Air Research; TEPCO, Tokyo Electric Power Company. bCMAQ,Community Multiscale Air Quality; Scav. coeff., scavenging coefficient.

Figure 1. Model domain used in the Community Multiscale AirQuality (CMAQ) simulation. Numbered prefectures: 1, Iwate; 2,Akita; 3, Yamagata; 4, Miyagi; 5, Fukushima; 6, Ibaraki; 7, Tochigi; 8,Gunma; 9, Chiba; 10, Saitama; 11, Tokyo; 12, Kanagawa; 13,Shizuoka; 14, Yamanashi; 15, Nagano; 16, Niigata. The white squareindicates the site of the Fukushima Daiichi nuclear power plant(FDNPP), and the white cross indicates Mount Tsukuba.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−23222315

Observational Data. Daily deposition rates of 137Cs weremonitored with bulk samplers over 46 Japanese prefecturesstarting on 18 March 2011.16 We assumed that the depositionrates measured with bulk samplers were between the wetdeposition rates and the total deposition rates (i.e., dry pluswet) as in our previous study.5 As radiocesium deposition wasdominated by wet processes, we believe that uncertainties dueto this assumption are small.In addition, we used data from the airborne monitoring

survey to evaluate model performance; details of the surveymethodology are available elsewhere,9 and a brief description isgiven in section S2 of the Supporting Information. Note that

the airborne measurements over eastern Japan were conductedfrom June to November 2011, and thus these data cannot bedirectly compared with simulated deposition during March−April 2011. However, the measured amount of depositedradiocesium decreased by about 1.8% for reasons other thanphysical attenuation between 31 May−2 July and 22 October−5 November 2011.9 In addition, radiocesium discharge wasestimated to be small (0.3% of deposited 137Cs) in a forestedcatchment on Mount Tsukuba (Figure 1) over the year afterthe accident, as detailed elsewhere.17 Although no such budgetstudies have been conducted in other areas, the radiocesiumdischarge from a forest is not expected to be large. These

Figure 2. Observed (Obs) and simulated (Model) 137Cs deposition rates. Upper panels show deposition rates averaged over 18 March−20 April(18−31 March for EM3) at 15 surface monitoring sites (Figure 1). Both total and wet deposition rates are indicated because the deposition ratesmeasured with bulk samplers were assumed to fall between the wet deposition rates and the total deposition rates, as noted in the text. Middle panelsshow daily deposition rates over the 15 surface monitoring sites. Lower panels show the comparison between 137Cs total deposition on model gridsas determined from airborne monitoring and model simulation.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−23222316

results suggest that the decrease in radiocesium due tosediment discharge or resuspension from surfaces was smallduring this period.

■ RESULTS AND DISCUSSION

Model Validation with Uncertainty Analysis. Wecompared the observed and simulated deposition data at 15surface monitoring stations (Figure 2 and Figure S3 of theSupporting Information; measurements were not conducted inMiyagi Prefecture, owing to damage to an instrument), and theobserved and simulated precipitation rates at meteorologicalmonitoring stations near surface monitoring sites (Figure S4 ofthe Supporting Information). Temporal variations of precip-itation rates were reproduced by the model with a correlationcoefficient averaged over the 16 monitoring stations (Figure S4of the Supporting Information) of 0.50. Simulations STD,EM2, and EM3 reproduced the total deposition of 137Cs during18 March−20 April within 1 order of magnitude for most of themonitoring stations, although cases EM2 and EM3 generallyunderestimated and overestimated the observations, respec-tively. Daily deposition rates were also reproduced within 1order of magnitude for the high-deposition cases (≥1 kBq m−2

day−1) in the STD and EM3 cases, whereas the EM2 caseunderestimated the observation at some sites in the Kantoregion (P6−P12 in Figure 1) during 21−23 March. Thisunderestimation was mostly associated with the low emissionestimate for 21−22 March. Overall, simulations STD, EM2, and

EM3 reproduced the observations (≥1 kBq m−2 day−1) withina factor of 10 for 75%, 38%, and 75%, respectively (part a ofTable 2). By contrast, for observed depositions higher than 0.1kBq m−2 day−1, the proportion of observations reproduced bysimulations within a factor of 2 (FA2) or 10 (FA10) werehigher in the EM2 case than in the STD case (part b of Table2), because the EM2 case reproduced the observations better inApril suggesting that the JAEA analysis underestimated the137Cs emission rates in April. For low-deposition cases (<0.1kBq m−2 day−1), all simulations generally underestimated theobservations (Figure S3 of the Supporting Information). Forthese cases, it is possible that the observed values were affectedby contamination of instruments or by particle resuspensionfrom surfaces, which were not included in the current model.These possibilities were also suggested in a previous study.2

Although it is difficult to evaluate the effect of these processes,the low-deposition cases do not have a large effect on totaldeposition, and we did not examine this problem further.We also evaluated the model performance by a comparison

with the airborne monitoring data (Figures 2 and 3). Theresults in the STD case were most consistent with observations.In high-deposition areas (≥10 kBq m−2), the STD simulationreproduced the observations within 1 order of magnitude inmost cases (96%), whereas the EM2 case overestimated 12% ofthe observations by more than 1 order of magnitude, and theEM3 case underestimated 11% of the observations by morethan 1 order of magnitude (Figure 2 and Table 2). The STD

Table 2. Comparison between Base-Case (STD) and Sensitivity Simulations; FA2 and FA10 Are the Proportions of SimulatedData That Reproduce the Observations within a Factor of 2 or 10, respectively; r and n Are Correlation Coefficient and Numberof Data Points, Respectively

(a) Comparison with surface monitoring (cutoffa: 1 kBq m−2 day−1)

STD EM2 EM3 WD2 WD3 DD2

FA2 (%) 0.0 0.0 25.0 37.5 37.5 0.0FA10 (%) 75.0 37.5 75.0 62.5 62.5 75.0n 8 8 8 8 8 8

(b) Comparison with surface monitoring (cutoffa: 0.1 kBq m−2 day−1)

STD EM2 EM3 WD2 WD3 DD2FA2 (%) 5.6 7.4 5.6 13.0 9.3 5.6FA10 (%) 25.9 37.0 33.3 33.3 27.8 25.9r 0.843 0.441 0.280 0.659 0.777 0.844n 54 54 36 54 54 54

(c) Comparison with airborne monitoring (cutoffb: 10 kBq m−2, n = 2448)

STD EM2 EM3 WD2 WD3 DD2FA2 (%) 57.0 34.6 40.0 44.9 54.9 56.8FA10 (%) 95.6 87.8 88.9 98.7 99.6 95.5r 0.663 0.526 0.308 0.639 0.720 0.663

(d) Comparison with airborne monitoring (cutoffb: 1 kBq m−2, n = 8314)c

STD EM2 EM3 WD2 WD3 DD2FA2 (%) 28.6 23.8 24.4 31.5 30.4 28.3FA10 (%) 65.6 64.7 60.7 70.4 69.6 65.5

(e) Budget analysis (PBq)

STD EM2 EM3 WD2 WD3 DD2 Obsd

emission 8.79 36.63 10.04 8.79 8.79 8.79total deposition over land 2.21 4.98 0.95 2.03 3.19 2.19 2.40wet deposition over land 2.16 4.87 0.91 1.97 3.14 2.17total deposition over ocean 1.81 3.48 1.63 2.22 3.00 1.80wet deposition over ocean 1.75 3.24 1.52 2.12 2.94 1.75outflow from the domain 4.75 28.09 7.46 4.50 2.58 4.78

aMinimum cutoff of observed daily deposition rates. bMinimum cutoff of model grid with observed deposition rates. cCorrelation coefficients are notshown as they are almost identical with those in part c of Table 2. dData from airborne monitoring survey.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−23222317

case reproduced the observed high-deposition areas well(Figure 3); specifically to the northwest of the FDNPP inFukushima Prefecture (P5), the central part of FukushimaPrefecture (Naka-dori region), Gunma and Tochigi prefectures(P7 and P8), the southern and northern parts of MiyagiPrefecture (P4), and the northeastern and southern parts ofIbaraki Prefecture (P6).Note that the performances of the EM2 and EM3

simulations using surface deposition data were the oppositeof the performances of the same simulations using airbornemonitoring data. This difference was caused by the differencesin the analytical periods. For the model validation using surfacedeposition data, the analytical period was March 18−April 20,and deposition over March 20−23 and March 30−31 mademajor contributions, as detailed in the next section. By contrast,for the model validation using airborne monitoring data, theanalytical period was March 11−April 20, and deposition onMarch 15−16 made the largest contribution. Thus, these results

suggest that EM2 (EM3) overestimated (underestimated) the137Cs release rate during March 15−16, and underestimated(overestimated) the 137Cs release rate during March 20−23 andMarch 30−31.The observed and simulated (STD case) total deposition

over the model domain were 2.4 PBq and 2.2 PBq, respectively(Table 2). This agreement between observed and simulatedresults indicates that the JAEA emission estimates were accuratefor the periods when air masses moved over the land. The EM2(EM3) simulation overestimated (underestimated) totaldeposition of 137Cs by a factor of 2.1 (2.5). These differenceswere caused primarily by the differences in the emissionestimate for 15 March 2011.We also evaluated the sensitivity of models to the wet

deposition modules by comparing the STD case with the WD2and WD3 cases (Table 2, Figures S5 and S6 of the SupportingInformation). In the WD2 simulation, high-deposition areasextended farther from the FDNPP as compared to observations

Figure 3. Maps showing observed and simulated 137Cs deposition.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−23222318

and the STD simulation (Figure S6 of the SupportingInformation), suggesting an underestimation of the scavengingcoefficient in the JAEA model. Simply multiplying thescavenging coefficient of the JAEA model by a factor of 10(i.e., the WD3 case) improved the model’s reproduction of theobserved deposition pattern (Figure S6 of the SupportingInformation).In the DD2 simulation, dry deposited amount over the land

and ocean decreased by 51% and 34% respectively compared tothe STD case. However, both the differences in total depositionover both land and ocean and the differences in the amount ofoutflow from the model domain are less than 1%, suggestingthat the choice of particle diameters in the model settings has asmall effect on the radiocesium budget (Table 2).To examine the possibility that simulations other than the

STD simulation could better explain the observed radiocesiumdeposition patterns, we also conducted E2W2, E3W2, E2W3,and E3W3 simulations, the settings of which are given in Table1 (also Table S2 and Figures S5 and S6 of the SupportingInformation). The E3W2 and E2W3 simulations clearlyoverestimated and underestimated respectively the observed

total radiocesium deposition over land (part c of Table S2 ofthe Supporting Information). The E2W2 and E3W3 simu-lations better reproduced the observed total radiocesiumdeposition than did the EM2 and EM3 simulations,respectively. However, spatial distributions of radiocesiumdeposition in the E2W2 and E3W3 simulations differedsubstantially from the observed deposition (Figure S6 of theSupporting Information), and correlation coefficients betweenobserved and simulated radiocesium depositions over eachmodel grid were much lower than in the STD case (part a ofTable S2 of the Supporting Information).Overall, the simulations using JAEA emission estimates best

reproduced the observed deposition patterns over easternJapan. This result suggests that to simulate the depositionpatterns of fine particles on a regional scale, emission estimatesshould also be conducted with a regional-scale model ratherthan a local- or global-scale model.The wet deposition modules of Terada et al.7 seem to

underestimate Λ, and simulation with Λ multiplied by a factorof 10 better reproduced observations. As summarized in FigureS1 of the Supporting Information, Λ varies greatly among

Figure 4. (a) Simulated 137Cs budget, (b) 137Cs deposition rates over land, (c) 137Cs deposition rates over the ocean, and (d) simulated andobserved21 wind direction at the FDNPP.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−23222319

studies suggesting that the choice of Λ is a source of much

uncertainty. In an atmospheric simulation of radionuclides after

the Chernobyl accident, a wet deposition scheme based on

relative humidity better reproduced observed radiocesium

deposition than a parametrization based on precipitation

rates.18 These results indicate that wet deposition modules

based only on precipitation rates include large uncertainties,

and thus, we recommend the process-based wet depositionmodule, such as a module for CMAQ.Overall, the STD simulation reproduced the temporal and

spatial variations of 137Cs deposition over eastern Japan well,and thus episode analysis using this model would be helpful inunderstanding the deposition mechanism of radiocesium.

Episode Analysis of Radiocesium Deposition. Weevaluated the 137Cs budget in the model domain (part a of

Figure 5. Simulated 137Cs deposition during the eight episodes (a−h) and during the periods other than these episodes (i).

Environmental Science & Technology Article

dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−23222320

Figure 4) by quantifying the contributions of individualprocesses, such as advection, emissions, and dry/wetdepositions, to the atmospheric concentrations using ProcessAnalysis, an analytical tool employed in CMAQ,11 and thespatiotemporal variations of 137Cs deposition during eachepisode (Figure 5). In part a of Figure 4, the positive advectionterm refers to outflow. After the large 137Cs release on 12March 2011, most 137Cs emitted from the FDNPP wastransported outside of the model domain. During this period,when the high-137Cs air mass was transported to the northeast(part d of Figure 4), the precipitation rate was low, and thus the137Cs deposition rate was low as well. By contrast, after theintense 137Cs releases during 15−16, 20−21, and 30 March,large amounts of 137Cs were deposited over the model domainthrough wet processes (parts b and c of Figure 4), as detailedlater. Overall, 137Cs deposition was dominated by wetdeposition: 98% of total deposition over land and 96% overthe ocean occurred through wet processes.In the STD case, most deposition over land occurred during

15−16 March (72.0% of the total deposition during the modelperiod), 20−24 March (15.7%), and 30−31 March (11.4%).These proportions changed to 93.7%, 4.9%, and 0.4%respectively in the EM2 case, and 62.7%, 28.9%, and 0.1%respectively in the EM3 case. These results again demonstratethat the emission patterns have a large effect on the temporalpatterns of radiocesium deposition over land. Deposition of137Cs over land was high when the wind at the FDNPP blewinland and transient cyclones passed over or south of Japan.5

The patterns of 137Cs deposition and precipitation during theperiod of high 137Cs deposition are shown in Figure 5 andFigure S7 of the Supporting Information, respectively. On themorning of 15 March, when northerly winds were predominantover eastern Japan, 137Cs was blown into the Kanto region.However, during this period precipitation was low, and thuswet deposition was low. In the afternoon, a southeasterly windtransported a radiocesium plume inland from the FDNPP, andprecipitation started over the northern and central parts (Naka-dori region) of Fukushima Prefecture (P5) in the evening.Radiocesium was effectively deposited to the northwest of theFDNPP and over the Naka-dori region (part b of Figure 5).These mechanisms resulted in the widely recognized Λ-shapeddeposition pattern in and around Fukushima Prefecture. Inaddition, the high-137Cs plume transported to the Kanto regionin the morning was transported to Tochigi and Gunmaprefectures (P7 and P8) by southeasterly winds in theafternoon. Precipitation started in Tochigi and Gunmaprefectures in the afternoon of 15 March, and large amountsof 137Cs were deposited over these areas as well. In the morningof 16 March, precipitation occurred over wide areas of easternJapan, and wet processes deposited 137Cs over a wide areaaround the FDNPP.On 20 March, when southerly winds were predominant

around the FDNPP, the radiocesium plume was transportednorthward, and precipitation occurred in the evening over thenorthern part of Miyagi Prefecture (P4), as well as over a widearea including Iwate (P1), Akita (P2), and northern Yamagata(P3) prefectures. As a result, there were high-deposition areasin the northern parts of Miyagi Prefecture.During 21−23 March, northerly winds were predominant

around the FDNPP, and the 137Cs plume extended to theKanto region. During this period, precipitation covered a largepart of eastern Japan (particularly the Kanto region), and

radiocesium was deposited over the eastern part of the Kantoarea (e.g., the southern part of Ibaraki Prefecture (P6)).On 30 March, the 137Cs plume was transported to the west

around noon, and precipitation covered the eastern part ofFukushima Prefecture in the afternoon. Radiocesium depositionoccurred around the FDNPP. During the periods other thanthese specific episodes, deposition was much lower (1.0% of thetotal deposition over the model period, part I of Figure 5).As this discussion shows, the mechanisms of radiocesium

deposition become clearer as more observational data becomeavailable and model analysis advances. However, whereas thesimulated radiocesium deposition patterns over land werevalidated in detail, validation of the deposition over the ocean isstill very limited and should be addressed in future studies. Theobservational data of atmospheric concentration and depositionover the ocean are limited, and thus combined analysis usingboth atmospheric and oceanic models would effectivelyimprove our understanding of the behaviors of radiocesiumwhen air masses were transported to the east.A comprehensive model sensitivity analysis is beyond the

scope of this study. Future studies should include acomprehensive evaluation of the model settings. In particular,the JAEA emission estimates were conducted using a regional-scale model, which seems to underestimate wet depositionrates. Emission estimates using a regional-scale model with aprocess-based wet deposition module would be necessary infuture studies. In addition, inversion modeling using thisairborne monitoring data would be also useful to further refineemission estimates when air masses were transported to thewest.Model intercomparison would also provide insight into

model uncertainties and reasonable model settings. The resultsof atmospheric simulation models are critical input data forocean19 and terrestrial water simulation models. Thus, furtherrefinement of atmospheric models is necessary for theevaluation of the future migration of radiocesium in theenvironment.20

■ ASSOCIATED CONTENT*S Supporting InformationAdditional details of our analysis. This material is available freeof charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*Tel: +81-29-850-2544, fax: +81-29-850-2480, e-mail: morino.yu@nies.go.jp.NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe thank the staff of the Ministry of Education, Culture,Sports, Science, and Technology for carrying out measurementsand providing observation data sets. We thank H. Nagai, G.Katata, and H. Terada for useful discussions on modelsimulations, and N. Suzuki and Y. Imaizumi for helpfulcomments on the behavior of radiocesium after atmosphericdeposition.

■ REFERENCES(1) Chino, M.; Nakayama, H.; Nagai, H.; Terada, H.; Katata, G.;Yamazawa, H. Preliminary estimation of release amounts of I-131 and

Environmental Science & Technology Article

dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−23222321

Cs-137 accidentally discharged from the Fukushima Daiichi nuclearpower plant into the atmosphere. J. Nucl. Sci. Technol. 2011, 48 (7),1129−1134.(2) Stohl, A.; Seibert, P.; Wotawa, G.; Arnold, D.; Burkhart, J. F.;Eckhardt, S.; Tapia, C.; Vargas, A.; Yasunari, T. J. Xenon-133 andcaesium-137 releases into the atmosphere from the Fukushima Dai-ichi nuclear power plant: Determination of the source term,atmospheric dispersion, and deposition. Atmos. Chem. Phys. 2012, 12(5), 2313−2343.(3) Estimation of the released amount of radioactive materials intothe atmosphere as a result of the accident in the Fukushima DaiichiNuclear Power; Tokyo Electric Power Company: Tokyo, Japan, 2012;http://www.tepco.co.jp/en/press/corp-com/release/betu12_e/images/120524e0205.pdf.(4) Katata, G.; Terada, H.; Nagai, H.; Chino, M. Numericalreconstruction of high dose rate zones due to the Fukushima Dai-ichi Nuclear Power Plant accident. J. Environ. Radioactiv. 2012, 111, 2−12.(5) Morino, Y.; Ohara, T.; Nishizawa, M., Atmospheric behavior,deposition, and budget of radioactive materials from the FukushimaDaiichi nuclear power plant in March 2011. Geophys. Res. Lett. 2011,38.(6) Yasunari, T. J.; Stohl, A.; Hayano, R. S.; Burkhart, J. F.; Eckhardt,S.; Yasunari, T. Cesium-137 deposition and contamination of Japanesesoils due to the Fukushima nuclear accident. Proc. Natl. Acad. Sci.,U.S.A. 2011, 108 (49), 19530−19534.(7) Terada, H.; Katata, G.; Chino, M.; Nagai, H. Atmosphericdischarge and dispersion of radionuclides during the Fukushima Dai-ichi Nuclear Power Plant accident. Part II: verification of the sourceterm and analysis of regional-scale atmospheric dispersion. J. Environ.Radioactiv. 2012, 112, 141−54.(8) Takemura, T.; Nakamura, H.; Takigawa, M.; Kondo, H.;Satomura, T.; Miyasaka, T.; Nakajima, T. A numerical simulation ofglobal transport of atmospheric particles emitted from the FukushimaDaiichi nuclear power plant. Sola 2011, 7, 101−104.(9) Results of the Fourth Airborne Monitoring Survey by MEXT;Ministry of Education, Culture, Sports, Science and Technology:Tokyo, Japan, 2011; http://radioactivity.mext.go.jp/en/contents/4000/3179/24/1270_1216.pdf.(10) Skamarock, W. C.; Klemp, J. B.; Dudhia, J.; Gill, D. O.; Barker,D. M.; Duda, M. G.; Huang, X. Y.; Wang, W.; Powers, J. G. Adescription of the Advanced Research WRF Version 3; NCAR/TN-475+STR; Natl. Cent. for Atmos. Res.,: Boulder, Colo., 2008.(11) Byun, D.; Schere, K. L. Review of the governing equations,computational algorithms, and other components of the models-3Community Multiscale Air Quality (CMAQ) modeling system. Appl.Mech. Rev. 2006, 59 (1−6), 51−77.(12) Appel, K. W.; Foley, K. M.; Bash, J. O.; Pinder, R. W.; Dennis,R. L.; Allen, D. J.; Pickering, K. A multi-resolution assessment of theCommunity Multiscale Air Quality (CMAQ) model v4.7 wetdeposition estimates for 2002−2006. Geosci. Model Dev. 2011, 4 (2),357−371.(13) Morino, Y.; Ohara, T.; Kurokawa, J.; Kuribayashi, M.; Uno, I.;Hara, H. Temporal variations of nitrogen wet deposition across Japanfrom 1989 to 2008. J. Geophys. Res.−Atmos. 2011, 116.(14) Sportisse, B. A review of parameterizations for modelling drydeposition and scavenging of radionuclides. Atmos. Environ. 2007, 41(13), 2683−2698.(15) Kaneyasu, N.; Ohashi, H.; Suzuki, F.; Okuda, T.; Ikemori, F.Sulfate Aerosol as a Potential Transport Medium of Radiocesium fromthe Fukushima Nuclear Accident. Environ. Sci. Technol. 2012, 46 (11),5720−5726.(16) Ministry of Education, Culture, Sports, Science and Technology,Reading of Radioactivity Level in Fallout by Prefecture. http://www.mext.go.jp/english/incident/1307872.htm.(17) Radioactive Materials and Environmental Disaster Research inNIES Annual Report 2012, National Institute for EnvironementalStudies: Tsukuba, Japan, 2012; http://www.nies.go.jp/kanko/annual/ae-18/ae18.pdf.

(18) Brandt, J.; Christensen, J. H.; Frohn, L. M. Modelling transportand deposition of caesium and iodine from the Chernobyl accidentusing the DREAM model. Atmos. Chem. Phys. 2002, 2, 397−417.(19) Kawamura, H.; Kobayashi, T.; Furuno, A.; In, T.; Ishikawa, Y.;Nakayama, T.; Shima, S.; Awaji, T. Preliminary numerical experimentson oceanic dispersion of I-131 and Cs-137 discharged into the oceanbecause of the Fukushima Daiichi nuclear power plant disaster. J. Nucl.Sci. Technol. 2011, 48 (11), 1349−1356.(20) Yoshida, N.; Kanda, J. Tracking the Fukushima radionuclides.Science 2012, 336 (6085), 1115−1116.(21) Tokyo Electric Power Company, Radiation dose measured inthe Fukushima Daiichi Nuclear Power Station 2011, http://www.tepco.co.jp/en/nu/fukushima-np/f1/data/2011/index-e.html.

Environmental Science & Technology Article

dx.doi.org/10.1021/es304620x | Environ. Sci. Technol. 2013, 47, 2314−23222322