Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2026 May 22;16:23509. doi: 10.1038/s41598-026-52671-9

External radiation exposure via car-borne survey among returnees in former restricted regions in Fukushima a decade after the nuclear accident

Yasutaka Omori 1, Yuki Oda 2, Masahiro Hosoda 1,2,✉, Yasuyuki Taira 3, Mayumi Shimizu 4, Ryohei Yamada 1, Hirofumi Tazoe 1, Masaru Yamaguchi 2, Minoru Osanai 2, Naofumi Akata 1, Shinji Tokonami 1
PMCID: PMC13416151  PMID: 42173999

Abstract

Prolonged radiation exposure continues in Fukushima Prefecture, Japan, due to the presence of long-lived artificial radionuclides (mostly 134Cs and 137Cs) discharged from the Fukushima Daiichi Nuclear Power Plant in 2011. Extensive decontamination has reduced the ambient gamma dose rates, allowing the government to lift evacuation orders. However, the external effective doses additionally received from the discharged radionuclides should be evaluated by considering the heterogeneous spatial distribution of ambient gamma dose rates from primordial radionuclides (238U- and 232Th-series elements and 40K) in the affected areas. The present study aimed to comparatively evaluate the doses of artificial and primordial radionuclides in Kawauchi Village and Tomioka and Okuma Towns by discriminatively measuring the dose rates associated with these radionuclides using a car-borne survey technique, thus elucidating the prolonged radiological impact over a decade after the nuclear accident. The results showed that the annual external effective doses of 134Cs and 137Cs ranged from 0 to 1.7 mSv (median range: 0.04–0.55 mSv), whereas those of primordial radionuclides ranged from 0.11 to 0.35 mSv. At more than 90% of the 938 measurement points, the annual doses of 134Cs and 137Cs were well below 1 mSv, which represent the lowest value of the reference level band (1–20 mSv) used for existing exposure situations conceptualised by the International Commission on Radiological Protection. The additional external dose remained below 1 mSv and was of the same order of magnitude as the external dose from the primordial radionuclides because the evacuation order was lifted. Government bodies can incorporate such data to promote public understanding of radiation in the affected areas to reduce concern surrounding radiation exposure on returnees, migrants, and evacuees.

Keywords: Absorbed dose rate in air, Evacuation, External radiation exposure, Fukushima Daiichi Nuclear Power Plant, Primordial radionuclides, Radioactive cesium, Returnees

Subject terms: Environmental sciences, Natural hazards

Introduction

The Fukushima Daiichi Nuclear Power Plant (FDNPP) accident, which occurred in March 2011, caused heavy contamination of gamma-emitting radionuclides (e.g., 132Te, 131I, 134Cs, and 137Cs with half-lives of 3.2 d, 8.0 d, 2.1 y, and 30 y, respectively) and the elevation of ambient dose equivalent rates (ranging from 1 to 100 µSv h-1) in the living environment of Fukushima Prefecture, northeastern Japan1,2. The Japanese Government issued evacuation orders to 12 municipalities located close to the FDNPP. The evacuation areas were redesignated in April 2012 into three categories, depending on the estimated annual doses (< 20, 20–50, and > 50 mSv) resulting from external exposure to the deposited artificial radionuclides3. Consequently, the number of evacuees exceeded 160,000, which included voluntary evacuees in May 20124.

The elevated ambient dose equivalent rates decreased substantially because of the decay of the released short-lived radionuclides (e.g. 132Te and 131I) as well as the removal of the deposited long-lived radionuclides (e.g. 134Cs and 137Cs) via migration into the deeper soil layers, weathering effects, and extensive decontamination activities throughout Fukushima Prefecture5,6. Based on the dose rate reduction and reconstruction of infrastructure facilities, the evacuation order had been lifted by 2020 for all of the areas except “Difficult-to-Return Zones” where initial annual effective doses exceeded 50 mSv. However, only 10–20% of evacuees returned to their hometowns in most municipalities where evacuation orders were lifted4. These municipalities include those within difficult-to-return zones. One reason why evacuees hesitate to return is their concerns about prolonged exposure to artificial radiation and its health effect7–10. Similar concerns are pronounced in local municipal staff who continue to work in branch offices outside the municipality, showing a higher risk perception of external radiation exposure than those working at main offices inside the municipality11.

Several investigations have been conducted to evaluate the external radiation exposure received by returnees in areas where evacuation orders were lifted. The external radiation doses were evaluated as additional doses that excluded the contribution of the primordial radionuclides (238U- and 232Th-series elements and 40K) present in the ground. Additional external doses were evaluated in the following ways: one study used 134Cs and 137Cs activity concentrations of the soil samples on the ground surface for Kawauchi Village12, and another used ambient dose equivalent rates subtracted by the background dose rate from the primordial radionuclides for Tomioka, Okuma, and Futaba Towns13,14. In the latter case, the representative value (0.04 µSv h-1 of background dose rates obtained before the FDNPP accident was applied to the estimation with the assumption of their geographically homogeneous distribution in the areas. However, when the decontamination measures were completed and more than 10 years after the FDNPP accident, the fractions of the doses received from the primordial radionuclides increased relative to those from 134Cs and 137Cs. In such cases, the assumption of a homogeneous distribution of the background dose rates increases the uncertainty of the estimated additional external doses. Site-specific values of background dose rates should be applied for estimation as in previous studies15–17. Sato et al.17 performed an additional dose evaluation in 2020 for Okuma and Tomioka Towns, using site-specific background dose rates. However, they focused mainly on the validation of their constructed model and did not report the variability in the additional doses administered in these two towns.

The present study surveyed the spatial distribution of outdoor ambient gamma dose rates from the primordial radionuclides together with those from artificial radionuclides in Kawauchi Village and Tomioka and Okuma Towns (Fig. 1). The survey identified 134Cs and 137Cs among the artificial radionuclides discharged from the FDNPP. The doses from radioactive cesium and primordial radionuclides were comparatively evaluated using outdoor dose rate data by introducing an occupancy factor and a building dose reduction factor to clarify the prolonged radiological impact of the nuclear accident in these areas. The present survey focused only on the exposure to external radiation. Internal radiation exposure from ingestion and inhalation of natural and artificial radionuclides was reported by Anderson et al.18 and Tamakuma et al.19 and is not considered in this study.

Fig. 1.

Fig. 1

Open in a new tab

Locality map of survey area. Solid circle represents the site of Fukushima Daiichi Nuclear Power Plant.

Survey design

Survey area

Radiation surveys were conducted in Kawauchi Village, Tomioka Town, and Okuma Town, all of which are within 30 km of the FDNPP. Evacuation orders were issued to the three municipalities and 30,300 residents were evacuated4. In Kawauchi Village, evacuation orders were lifted for the entire region by 2016, and 83% (= 1862/2238) of the registered population lived there as of December 202420. In contrast, the evacuation orders for Tomioka and Okuma Towns were lifted for 93% and 49% of their regions by 2023, respectively, and difficult-to-return zones remained. As of December 2024, 23% (= 2565/11338) and 9% (= 874/9947) of the registered population lived in Tomioka and Okuma, respectively21,22. Dose rate measurements were performed in both living areas and difficult-to-return zones, whereas dose evaluations were limited to living areas.

Measurement of ambient gamma dose rates from primordial and artificial radionuclides

Ambient gamma dose rates determined in the present study were absorbed dose rates in air (unit: nGy h-1). The dose rates from primordial and artificial radionuclides 134Cs and 137Cs were evaluated based on a combination of in-situ measurements of gamma-ray pulse-height distributions and car-borne survey measurements. The in-situ and car-borne survey measurements were carried out from June 2021 to September 2023. Sunny or cloudy conditions were selected for the measurements to exclude the effects of shielding due to snow cover and the deposition of 214Pb and 214Bi during rainfall. To consider the removal of 134Cs and 137Cs from the living environment, the dose rates were decay-corrected to a reference date of 30 September 2023. The reduction rates (half-lives) of the dose rates were evaluated following Ogura et al.16 using dose-rate monitoring data at several monitoring stations installed by the national and prefectural governments. The ambient dose equivalent rates measured on 1 June 2021 and 30 September 2023 at 34 of the 77 monitoring stations in the survey areas were analysed using a simple exponential function, whereas the data at the other monitoring stations were excluded from the analysis because missing data and/or a remarkable decrease in dose rate due to decontamination measures were observed. The results showed a median dose rate half-life as 10.5 y (half-life range: 5.0–23.4 y) (Fig. 2). The half-life value was larger than 1 y or less23, causing a rapid decrease in dose rates during the first few years after the accident, but it was closer to 5–9 y, as evaluated from dose-rate and 137Cs deposition measurements 5 years or more after the accident24,25. Although the half-life distribution may have bimodality with dominant peaks at 10 and 20 y (Fig. 2), the relative difference in decay correction between these two half-lives for survey period (i.e., 2 y) is only 7%. Consequently, decay correction was applied to the dose rates using a simple exponential function with the half-life of 10.5 y, and the measured dose rates were corrected by 10%.

Fig. 2.

Fig. 2

Open in a new tab

Frequency distribution of half-lives of ambient dose equivalent rates measured at monitoring stations in the survey area.

The absorbed dose rates in air from the primordial radionuclides were evaluated based on in-situ measurements of the pulse-height distributions of gamma rays with and energy range of 50–3200 keV. The measurement points were selected from 85 accessible subareas, into which the survey area was subdivided using a 2 × 2 km mesh from north to south and east to west. The in-situ measurements were performed for 900 s at 1 m above the ground surface or pavement within each of the 85 meshed areas using a 3 in. × 3 in. NaI(Tl) scintillation spectrometer (EMF211, EMF Japan Co. Ltd., Japan). The gamma-ray pulse-height distributions were unfolded based on a 22 × 22 response matrix method26 to obtain the incident gamma ray flux density energy spectra (energy bin widths: 100–280 keV). The response matrix method assumes an isotropic radiation field, and primordial radionuclides in the ground are treated as approximately semi-infinite homogeneous volume sources. The absorbed dose rates in air from the primordial radionuclides were determined by focusing on the observed flux density at the photon peaks of 40K (1464 keV), 214Bi (1765 and 2205 keV) for the 238U-series elements, and 208Tl (2615 keV) for the 232Th-series elements. The concentrations of 40K, 238U, and 232Th were calculated by comparing the observed flux densities with those simulated using the unit concentrations of these elements distributed homogeneously in the ground. The absorbed dose rates in air from the primordial radionuclides were calculated by applying dose conversion factors to account for the concentrations of the respective elements. The pile up of gamma-ray counts for energies higher than 1400 keV due to 134Cs contamination was small enough to avoid overestimating the dose rates from primordial radionuclides. Further details regarding the spectral analysis can be seen in Minato26 and Omori et al.27

The absorbed dose rates in air from 134Cs and 137Cs were determined using car-borne surveys combined with in-situ measurements. The EMF211 detector was installed in a sedan-type car equipped with a global position system sensor. Gamma-ray counts inside the car and locality information (longitude and latitude) were continuously measured and recorded per 30 s with the car running at a speed of approximately 40 km h-1. This allowed the accumulation of gamma-ray counts over approximately 350 m along the surveyed roads at the measurement interval. In-situ measurements using the EMF211 were performed to evaluate the gamma-ray shielding capability of the car body and conversion factor from the counts to dose rates. Twelve locations were selected, which included flat plain, semi-mountainous (foot of mountain) and mountainous sites in the residential areas. The car-body shielding was examined by comparing the inside and outside gamma-ray counts per 30 s for 2 min at 10 of the 12 points. The comparison revealed strong linear correlation between inside and outside gamma-ray counts (Fig. 3a); the car-body shielding, expressed by a ratio of gamma-ray counts outside to inside, was determined as 1.54 (standard error: 0.04). The in-situ measurements of gamma-ray pulse-height distributions at 12 points were used to evaluate gamma-ray counts and absorbed dose rates in air from primordial and artificial radionuclides. The gamma-ray counts outside were linearly correlated with the absorbed dose rates in air, and the conversion factor from the counts to dose rates was determined as 0.0013 nGy h-1(cpm)-1 (standard error: 0.0001 nGy h-1 (cpm)-1) (Fig. 3b). Although the number of measurement locations were limited, the factor values for car-body shielding and dose conversion agreed well with those reported in previous studies. Consequently, the absorbed dose rate in air (D) was calculated as follows:

graphic file with name d33e521.gif 1

Fig. 3.

Fig. 3

Open in a new tab

Comparison of gamma-ray counts inside and outside the car (a) and that of gamma-ray counts outside the car and absorbed dose rate in air (b). Regression lines are drawn with determination coefficient values (R2).

where N represents gamma-ray counts per 30 s inside the car, and is multiplied by a factor of 2 to convert the gamma-ray counts per minute. The dose rates were subtracted by the dose rates of the primordial radionuclides obtained by in-situ measurements in the same meshed subareas to evaluate the dose rates of 134Cs and 137Cs. The calculated dose rates showing negative values were truncated to 0 nGy h-1, indicating no contribution of 134Cs and 137Cs to the dose rates. Further details regarding the car-borne survey technique can be found in Hosoda et al.28

The measurement uncertainty of dose rates in car-borne surveys was estimated at up to 5% considering uncertainties in gamma-ray counts, car-body shielding factor, and dose conversion factor. The measurement uncertainty of dose rates from primordial radionuclides using the in-situ measurements was approximately 3%. Details of the calculation procedure can be found in Hosoda et al.29

Evaluation of exposure to radiation from primordial and artificial radionuclides

Annual external exposure to radiation from the primordial radionuclides and 134Cs and 137Cs was evaluated separately. The present study did not measure the absorbed dose rates in air in the dwellings; therefore, they were evaluated from the outdoor dose rates. The annual external doses (Ep) from the primordial radionuclides were calculated using the following equation:

graphic file with name d33e590.gif 2

where Dp is the absorbed dose rate in air from the primordial radionuclides; OF is the indoor occupancy factor; and Rp and DCF are the indoor-to-outdoor dose rate ratio for natural gamma radiation and the conversion factor from the absorbed dose in air to the effective dose, respectively. The annual external doses (Ea) of 134Cs and 137Cs were calculated as follows:

graphic file with name d33e619.gif 3

where Da is the absorbed dose rate in air from 134Cs and 137Cs, and Ra is the indoor-to-outdoor dose rate ratio for artificial gamma radiation (i.e. a dose reduction factor for dwellings). In the evaluation of exposure to artificial radiation, the contributions of 134Cs and 137Cs deposited on the ground were only taken into account and those of the indoor radioactive contamination were not considered.

Dose evaluations used the primary measured data on the absorbed dose rates in air from the primordial and artificial radionuclides along the roads obtained via car-borne surveys. The data were collected over a 30 s interval and, consequently, 499, 271 and 168 points were evaluated for Kawauchi Village and Tomioka and Okuma Towns, respectively. These points were limited to areas where evacuation orders had been lifted and evacuees could live normally. The indoor occupancy factor was taken as 0.8; this represents an intermediate value between the indoor and outdoor worker’ cases, according to the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) 2020/2021 Report30. The indoor-to-outdoor dose rate ratio for natural gamma radiation was taken as unity based on results of several regional studies compiled by Omori et al.31, whereas the ratio for artificial gamma radiation was taken as 0.4 for wooden dwellings reported by Yoshida-Ohuchi et al.32, Furuta and Takahashi33 and Matsuda et al.34 According to the UNSCEAR report, DCF value of 0.70 Sv Gy-1 was used for evaluation30.

Results

Figure 4 shows a spatial distribution map of the absorbed dose rates in air from primordial radionuclides in Kawauchi Village and Tomioka and Okuma Towns. Meshes without data were considered inaccessible or non-residential areas. The dose rates ranged from 18 to 56 nGy h-1 with an average (± standard deviation) of 39 ± 8 nGy h-1 and median of 38 nGy h-1 for Kawauchi Village; from 22 to 46 nGy h-1 with an average of 33 ± 6 nGy h-1 and median of 33 nGy h-1 for Tomioka Town; and 22 to 43 nGy h-1 with an average of 31 ± 5 nGy h-1 and median of 30 nGy h-1 for Okuma Town. Dose rates were higher in Kawauchi Village than those for Tomioka and Okuma Towns.

Fig. 4.

Fig. 4

Open in a new tab

Spatial distributions of absorbed dose rates in air from primordial radionuclides for Kawauchi Village and Tomioka and Okuma Towns. The solid circle represents the site of the Fukushima Daiichi Nuclear Power Plant (FDNPP). The mesh size is 2 km × 2 km across north to south and east to west directions.

Figure 5 shows a spatial distribution map of the absorbed dose rates in air from artificial radionuclides. Only 134Cs and 137Cs were identified as artificial radionuclides discharged from the FDNPP in the in-situ measurements of gamma-ray pulse-height distributions. The dose rates ranged from 0 to 196 nGy h-1 with a median of 13 nGy h-1 for Kawauchi Village, 0 to 1380 nGy h-1 with a median of 172 nGy h-1 for Tomioka Town, and 15 to 2860 nGy h-1 with a median of 590 nGy h-1 for Okuma Town. Kawauchi Village showed lower dose rates because of the relatively small amount of 134Cs and 137Cs deposition35, decontamination measures were performed earlier, and consequently, evacuation orders were lifted by 2016. At 88% (= 441/499) of the qualified points, the dose rates of the artificial radionuclides were lower than those of the primordial radionuclides. In contrast, the areas with the dose rates higher than 500 nGy h-1 were located close to the FDNPP. These areas were categorised as those with an annual additional dose of 50 mSv or higher at the initial assessment immediately after the FDNPP accident. In the evacuation-order-lifted areas of Tomioka and Okuma Towns, the dose rates remained low, ranging from 4 to 532 nGy h-1 with a median of 130 nGy h-1. At 13% (= 55/438) of the qualified points, the dose rates from the artificial radionuclides were lower than those from the primordial radionuclides.

Fig. 5.

Fig. 5

Open in a new tab

Spatial distributions of absorbed dose rates in air from artificial radionuclides for Kawauchi Village and Tomioka and Okuma Towns as of September 2023. The FDNPP is located in the rectangular area.

Table 1 shows the estimated annual effective doses from the external exposure to gamma radiation from primordial radionuclides. The annual external dose ranged from 0.11 to 0.35 mSv across three municipalities. The doses in all residential areas were below the nation-wide (0.33 mSv) and prefectural (0.35 mSv) averages of external doses from terrestrial radiation31.

Table 1.

Annual external effective doses from primordial and artificial radionuclides on returnees in Kawauchi Village and Tomioka and Okuma Towns as of September 2023.

Location
(n)a 		Annual effective dose (mSv)
Primordial radionuclides Artificial radionuclides
Minimum 1st quartile Median 3rd quartile Maximum Minimum 1st quartile Median 3rd quartile Maximum

Kawauchi

n = 499

 0.11 0.22 0.22 0.27 0.35 0 0.01 0.04 0.09 0.62

Tomioka

n = 271

 0.15 0.21 0.21 0.21 0.28 0.01 0.27 0.55 0.76 1.7

Okuma

n = 168

 0.16 0.19 0.21 0.24 0.26 0.05 0.20 0.36 0.49 1.3
Open in a new tab

a n: Number of measurement sites for evaluating contribution from artificial radionuclides.

Table 1 shows the additional external doses of 134Cs and 137Cs. Unlike the exposure from the primordial radionuclides, these doses were widely distributed depending on the ambient dose rates, which ranged from 0 to 0.62 mSv for Kawauchi Village, 0.01 to 1.7 mSv for Tomioka Town, and 0.05 to 1.3 mSv for Okuma Town. The highest dose rate was obtained in the evacuation areas of Okuma Town, whereas the highest effective dose was obtained in Tomioka Town because effective dose evaluations were performed only for living areas. However, even for Tomioka and Okuma Towns, at more than 90% of the qualified points, the additional doses were well below 1 mSv; this represents the lowest value of the reference level band (1–20 mSv) used for existing exposure situations conceptualised by the International Commission on Radiological Protection36,37 and the Japanese Government’s policy target level for dose reduction38. The points showing effective doses of more than 1 mSv were limited to the former difficult-to-return zones, where evacuation orders were lifted during the survey period.

Discussion

The absorbed dose rates in air from the primordial radionuclides tended to be higher in Kawauchi Village than those in Tomioka and Okuma Towns. This characteristic spatial distribution has also been observed in previous studies39,40. Kawauchi Village overlies Cretaceous granitic basement, whereas Tomioka and Okuma Towns are located on Neogenic/Quaternary sedimentary rocks and deposits41. The granitic rock area has higher uranium and thorium contents than the sedimentary rock area, as revealed by river sediment analysis42. The geological setting of this survey area constrained the characteristic distribution of the absorbed dose rates in air from the primordial radionuclides.

The use of representative background dose rate value (0.04 µSv h-1) is not expected to largely affect dose assessment for residents in the areas a decade after the accident. Based on the quantitative relationship between the absorbed dose in air and ambient dose equivalent in the range of environmental gamma radiation, that is the ratio of ambient dose equivalent to absorbed dose in air as approximately 1.2343,44, the average values of the ambient dose equivalent rates for the three municipalities ranged from 0.04 to 0.05 µSv h-1. Additionally, the ratios of the standard deviations of the background dose rates to their averages were < 0.20. This implies that most in-situ measurement locations have a small dose rate variability of approximately 0.01 µSv h-1 compared to their average range mentioned above. Generally, the gamma radiation detectors frequently used in Japan are 1 in. × 1 in. NaI(Tl) scintillation survey meters that have a relative uncertainty of approximately 15%. By taking background dose rate ranges, its average values, and measurement uncertainty of gamma radiation detectors into consideration, the representative background dose rate value (0.04 µSv h-1) is applicable to dose assessment in these study areas.

Table 2 summarises the time variations in estimated external effective doses of artificial radionuclides received by residents across three municipalities. Ishikawa et al.45 evaluated the received doses for the first four months (March to July 2011) in the framework of the Fukushima Health Management Survey. The evaluated doses were mainly caused by exposure to short-lived artificial radionuclides (e.g. 132Te and 131I) during evacuation. After the evacuation orders were gradually lifted, Taira et al.12 and Matsuo et al.13 investigated the ambient dose equivalent rates and activity concentrations of 134Cs and 137Cs to evaluate annual additional doses for returnees. To make a better comparison to the present study, the assumptions in dose calculation in these studies were adjusted by the following: the use of ambient dose equivalent rates, the background dose rates as 0.04 µSv h-1, indoor and outdoor occupancy factors as 0.8 and 0.2, respectively, dose reduction factor for wooden dwellings as 0.4, and conversion factor from ambient dose equivalent to effective dose as 0.60. The conversion factor was calculated from the quantitative relationships among the ambient dose equivalent, absorbed dose in air, and effective dose mentioned above. The time-series variations in the estimated effective dose revealed that the doses remained consistently below 1 mSv and in the same order of magnitude as the external dose from the primordial radionuclides because the evacuation order was lifted (Table 2). Table 2 also highlights that the external doses accumulated for the first four months after the accident were equivalent to a few-year-long chronic radiation exposure from primordial radionuclides compared with the median values.

Table 2.

Estimated time variations in external effective doses affecting residents in Kawauchi Village and Tomioka and Okuma Towns after the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident.

Time of evaluation Effective dosea (mSv; median value) Reference
Kawauchi Tomioka Okuma
Dose from artificial radionuclides
 March–July 2011 0.6 (1.5)b 0.3 (1.5)b 0.7 (1.7)b Ishikawa et al.45
 December 2012 0.6 Modified from Taira et al.12 c
 October 2017 0.9 Modified from Matsuo et al.13 d
 September 2023 0.04 0.55 0.36 The present study
Dose from primordial radionuclides 0.22 0.21 0.21 The present study
Open in a new tab

a Effective doses are expressed as annual doses at the time of evaluation except for those reported by Ishikawa et al.’s study45, who evaluated the doses during the first four months after the FDNPP accident.

b 95 percentile values are shown in brackets.

c Ambient dose equivalent rates (excluding the background dose rate of 0.04 µSv h-1) were used to evaluate effective doses with the indoor and outdoor occupancy factors as 0.8 and 0.2, respectively, dose reduction factor for wooden dwellings as 0.4, and a conversion factor from ambient dose equivalent to effective dose as 0.60.

d Conversion factor (0.60) from the ambient dose equivalent to the effective dose was applied with indoor and outdoor occupancy factors of 0.8 and 0.2, respectively, to the results of Matsuo et al.13

This study focused only on external radiation exposure from artificial and primordial radionuclides. For internal radiation exposure, Anderson et al.18 reported the doses due to intake of drinking water as 0.02–1 mSv for 222Rn (one of the 238U-series elements) and 6 × 10− 4 mSv for 137Cs; however, Tamakuma et al.19 reported the doses due to inhalation as 0.8 mSv for 222Rn and 2 × 10− 6 mSv for 137Cs. Although the annual doses from intake of foodstuffs have not been investigated, those are expected to be < 0.1 mSv from the artificial radionuclides based on results of duplicate diet samples46 and food monitoring47,48. Combining the present study’s evaluations with these reported results, external exposure to radioactive cesium dominates the additional radiation exposure due to the nuclear accident; however, it is a minor component compared to exposure from natural radiation sources.

Some evacuees remain concerned about prolonged exposure to artificial radiation and its health effects, resulting in a low return rate to their hometown7–10. Interviews49,50 revealed that a reduction in ambient dose equivalent rate/radiation exposure, with progress and follow-up in decontamination are critical for their return. Studies have also revealed that the decision depends on infrastructure recovery including medical facilities. According to Kobayashi et al.51, the number of internal migrants from outside Fukushima Prefecture to evacuation-order-lifted areas increased after the nuclear accident as well as the returnees. Reportedly, the mental health of migrants was affected by their radiation risk perception because people outside Fukushima had limited access to information about radiation. The present study confirmed that external exposure to artificial radiation remained below 1 mSv, although the Japanese government began discussing whether the evacuation order should be lifted when decontamination was completed, and the estimated annual external dose from artificial radiation was reduced to < 20 mSv. However, the returnees and migrants actually received doses below 1 mSv annually because it takes more time to lift evacuation orders due to infrastructure reconstruction. Nakayama et al.52 and Hirota et al.53 demonstrated that health-related anxiety about radiation exposure was substantially lower for Fukushima residents who trusted government ministries and local governments. Hence, Japanese and local governments should incorporate such data and actively promote public understanding regarding the radiation situation in the affected areas to reduce the psychological impact of radiation exposure on returnees, migrants, and evacuees.

The present study has certain limitations. Although we evaluated additional external doses using the dose rates on roads obtained via the car-borne surveys, the values of dose conversion and occupancy factors were assumed and personal dosimetry were not performed. These may contribute to uncertainty in dose evaluation and result in differences from realistic personal doses. Direct measurements of personal doses showed that they were 0.1–0.2 of the effective doses evaluated from the dose rate measurements via air-borne monitoring54,55. According to Wainwright et al.56, dose rates measured via air-borne monitoring were higher than those obtained using car-borne monitoring by a factor of approximately 1.7, which may be supported by the difference in the dose rates between paved roads and unpaved surfaces (a factor of 2) reported by Kinase et al.57 Based on these reports, the effective doses shown in the present study may be overestimated by a factor of 2–5. In contrast, the assumptions of dose conversion and occupancy factors minutely contribute to the uncertainty when compared with the dose rate measurements. In the present study, DCF value of 0.7 Sv Gy-1, equivalent to the factor value of 0.6 from ambient dose equivalent to effective dose was applied, whereas Yajima et al.58 evaluated the conversion factor from the ambient dose equivalent to personal dose as 0.7, based on direct personal dose measurements. Questionnaires in previous studies15,55 revealed the indoor occupancy factor as 0.9 for residents in evacuation-order-lifted areas, which is 10% higher than the value applied in the present study. The in-situ measurements were performed in both ground and paved road in this study. Practically, dose rates above bare ground are nearly equal to those on roads. This implies that constituent ratio of uranium, thorium and potassium is apparently similar each other. This is supported from the previous studies indicating no apparent difference in dose rates between the grounds and paved road observed in four prefectures of Japan29,59.

Conclusion

This study evaluated the annual external radiation exposure among returnees in formerly restricted regions of Fukushima, a decade after the FDNPP accident. The absorbed dose rate in air from the primordial radionuclides ranged from 18 to 56 nGy h-1. By taking the detectors’ measurement uncertainty into consideration, the representative background dose rate value (0.04 µSv h-1) was applicable to dose assessment in the study regions. Because of the decade since the nuclear accident, 134Cs and 137Cs primarily contribute to the annual external effective doses. However, at more than 90% of the qualified points, the effective doses were well below 1 mSv, which represented the lowest value of the reference level band (1–20 mSv) used for existing exposure situations as conceptualised by the International Commission on Radiological Protection and the Japanese Government’s policy target level for dose reduction. Japanese and local governments should incorporate such data and actively promote public understanding regarding the radiation situation in areas affected by the nuclear accident to reduce anxiety and psychological impact of radiation exposure on returnees, migrants, and evacuees.

Acknowledgements

The authors would like to thank the staff of Kawauchi Village and Tomioka and Okuma Town Offices for arranging the surveys.

Author contributions

Yasutaka Omori: Formal analysis, Investigation, Visualization, Writing—original draft, Writing—review & editing. Yuki Oda: Date curation, Formal analysis, Investigation, Visualization, Writing—original draft, Writing—review & editing. Masahiro Hosoda: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Validation, Writing—review & editing. Yasuyuki Taira: Writing—review & editing, Investigation, Methodology, Validation. Mayumi Shimizu: Investigation, Project administration, Writing—review & editing. Ryohei Yamada: Investigation, Writing—review & editing. Hirofumi Tazoe: Investigation, Writing—review & editing. Masaru Yamaguchi: Investigation, Writing—review & editing. Minoru Osanai: Investigation, Writing—review & editing. Naofumi Akata: Investigation, Writing—review & editing. Shinji Tokonami: Conceptualization, Resources, Writing—review & editing.

Funding

This work was financially supported by the Research Project on the Health Effects of Radiation organised by the Ministry of the Environment, Japan, and a Grant for Scientific Research in Master Course at Hirosaki University Graduate School of Health Sciences.

Data availability

Data will be made available on request.

Declarations

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  • 1.Hosoda, M. et al. The time variation of dose rate artificially increased by the Fukushima nuclear crisis. Sci. Rep.1, 87. 10.1038/srep00087 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Lyons, C. & Colton, D. Aerial measuring system in Japan. Health Phys.102 (1), 509–515. 10.1097/HP.0b013e31824d0056 (2012). [DOI] [PubMed] [Google Scholar]
  • 3.Nuclear Emergency Response Headquarters of Japan. Review of Restricted Areas and Areas under Evacuation Orders. (Accessed 1 August 2025). https://warp.ndl.go.jp/web/20250902181203/http://www.kantei.go.jp/saigai/pdf/20120330gensaihonbu.pdf (2012).
  • 4.Croüail, P. et al. Analysis of the modalities of return of populations to the contaminated territories following the accident at the Fukushima power plant. Radioprotection55 (2), 79–93. 10.1051/radiopro/2020049 (2020). [Google Scholar]
  • 5.Ando, M., Sasaki, M. & Saito, K. Evaluation of decreasing trend in air dose rate and ecological half-life within an 80 km range from Fukushima Dai-ichi Nuclear Power Plant, using car-borne survey data measured by KURAMA systems up to 2018. J. Nucl. Sci. Technol.57 (12), 1319–1330. 10.1080/00223131.2020.1789008 (2020).
  • 6.Yoshimura, K. Air dose rates and cesium-137 in urban areas—deposition, migration, and time dependencies after nuclear power plant accidents. J. Nucl. Sci. Technol.59 (1), 25–33. 10.1080/00223131.2021.1973608 (2022). [Google Scholar]
  • 7.Matsunaga, H., Orita, M., Oishi, K., Taira, Y. & Takamura, N. Intention to return in residents of Okuma and its characteristics: the evacuation order was lifted eight years after the Fukushima Daiichi Nuclear Power Station accident. J. Radiat. Res.62 (5), 868–870. 10.1093/jrr/rrab058 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Hidaka, T. et al. Concerns related to returning home to a difficult-to-return zone after a long-term evacuation due to Fukushima Nuclear Power Plant accident: a qualitative study. PLoS ONE 17 (8), e0273684. 10.1371/journal.pone.0273684 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Matsunaga, H. et al. Latest topics of interest on radiation risk communication: a decade after Japan’s Fukushima Daiichi Nuclear Power Plant accident. Disaster Med. Public Health Prep. 17, e531. 10.1017/dmp.2023.194 (2023). [DOI] [PubMed] [Google Scholar]
  • 10.Ito, N. et al. Non-returning factors from an interview survey of 16 residents of Katsurao village 12 years after the Fukushima nuclear power plant accident. Radiat. Prot. Dosim. 200 (16–18), 1528–1531. 10.1093/rpd/ncae107 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Matsunaga, H., Kashiwazaki, Y., Orita, M., Taira, Y. & Takamura, N. Risk perception of internal and external radiation exposure among administration staff affected by the Fukushima Daiichi Nuclear Power Plant accident. J. Environ. Radioact. 248, 106869. 10.1016/j.jenvrad.2022.106869 (2022). [DOI] [PubMed] [Google Scholar]
  • 12.Taira, Y. et al. Evaluation of environmental contamination and estimated exposure doses after residents return home in Kawauchi Village, Fukushima Prefecture. Environ. Sci. Technol.48 (8), 4556–4563. 10.1021/es404534y (2014). [DOI] [PubMed] [Google Scholar]
  • 13.Matsuo, M. et al. Evaluation of environmental contamination and estimated radiation exposure dose rates among residents immediately after returning home to Tomioka Town, Fukushima Prefecture. Int. J. Environ. Res. Public Health. 16 (9), 1481. 10.3390/ijerph16091481 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Liu, M. et al. Temporal variation in environmental radioactivity and radiation exposure doses in the restricted areas around the Fukushima Daiichi Nuclear Power Plant. Sci. Rep.13, 22459. 10.1038/s41598-023-49821-8 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Sato, T., Andoh, M., Sato, M. & Saito, K. External dose evaluation based on detailed air dose rate measurements in living environments. J. Environ. Radioact. 210, 105973. 10.1016/j.jenvrad.2019.05.005 (2019). [DOI] [PubMed] [Google Scholar]
  • 16.Ogura, K. et al. Discriminative measurement of absorbed dose rates in air from natural and artificial radionuclides in Namie Town, Fukushima Prefecture. Int. J. Environ. Res. Public Health. 18 (3), 978. 10.3390/ijerph18030978 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Sato, R. et al. Assessment of individual external exposure doses based on environmental radiation in areas affected by the Fukushima Daiichi Nuclear Power Station accident. Environ. Int.194, 109148. 10.1016/j.envint.2024.109148 (2024). [DOI] [PubMed] [Google Scholar]
  • 18.Anderson, D. et al. Fukushima’s tap and groundwater a decade after the nuclear accident with radiocesium, tritium, and radon. Environ. Sci. Technol.59 (10), 4906–4914. 10.1021/acs.est.4c14601 (2025). [DOI] [PubMed] [Google Scholar]
  • 19.Tamakuma, Y. et al. Indoor and outdoor exposure to atmospheric radionuclides in areas affected by the Fukushima nuclear accident. Sci. Total Environ.1008, 181027. 10.1016/j.scitotenv.2025.181027 (2025). [DOI] [PubMed] [Google Scholar]
  • 20.Kawauchi Village Office. Population and Household. (Accessed 1 August 2025). https://www.kawauchimura.jp/population/ (2024).
  • 21.Tomioka Town Office. Housing and Evacuation Status. (Accessed 1 August 2025). https://www.tomioka-town.jp/life/7/38/ (2024).
  • 22.Okuma Town Office. Housing and Evacuation Status. (Accessed 1 August 2025). https://www.town.okuma.fukushima.jp/life/1/9/ (2024).
  • 23.Saito, K. et al. Summary of temporal changes in air dose rates and radionuclide deposition densities in the 80 km zone over five years after the Fukushima Nuclear Power Plant accident. J. Environ. Radioact. 210, 105878. 10.1016/j.jenvrad.2018.12.020 (2019). [DOI] [PubMed] [Google Scholar]
  • 24.Shikaze, Y. et al. Development of a prediction model for ambient dose equivalent rate distribution based on ecological half-life profiles using LASSO regression and KURAMA data. Radiat. Prot. Dosim. 201(15), 1025–1042. 10.1093/rpd/ncaf086 (2025). [DOI] [PubMed]
  • 25.Kim, M. et al. Effects of non-decontamination human activities on the reduction of ambient dose equivalent rates in residential areas near the Fukushima Daiichi Nuclear Power Plant. J. Environ. Radioact. 294, 107931. 10.1016/j.jenvrad.2026.107931 (2026). [DOI] [PubMed] [Google Scholar]
  • 26.Minato, S. Diagonal elements fitting technique to improve response matrixes for environmental gamma ray spectrum unfolding. Radioisotopes50, 463–471. 10.3769/radioisotopes.50.10_463 (2001). [Google Scholar]
  • 27.Omori, Y. et al. Improvement of spectral analysis using a NaI(Tl) scintillation spectrometer to evaluate ambient gamma dose rates from primordial radionuclides. Radiat. Prot. Dosim. 199 (18), 2233–2237. 10.1093/rpd/ncad258 (2023). [DOI] [PubMed] [Google Scholar]
  • 28.Hosoda, M. et al. Estimation of external dose by car-borne survey in Kerala, India. PLoS ONE 10 (4), e0124433. 10.1371/journal.pone.0124433 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Hosoda, M. et al. Environmental radiation monitoring and external dose estimation in Aomori Prefecture after the Fukushima Daiichi Nuclear Power Plant accident. Jpn J. Health Phys.51 (1), 41–50. 10.5453/jhps.51.41 (2016). [Google Scholar]
  • 30.United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR). Sources, Effects and Risks of Ionizing Radiation. UNSCEAR 2020/2021 Report to the General Assembly with Scientific Annexes. Volume II. Scientific Annex B (United Nations, 2022).
  • 31.Omori, Y. et al. Japanese population dose from natural radiation. J. Radiol. Prot.40 (3), R99–R17. 10.1088/1361-6498/ab73b1 (2020). [DOI] [PubMed] [Google Scholar]
  • 32.Yoshida-Ohuchi, H., Hosoda, M., Kanagami, T., Uegaki, M. & Tashima, H. Reduction factors for wooden houses due to external γ–radiation based on in situ measurements after the Fukushima nuclear accident. Sci. Rep.4, 7541. 10.1038/srep07541 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Furuta, T. & Takahashi, F. Study of radiation dose reduction of buildings of different sizes and materials. J. Nucl. Sci. Technol.52 (6), 897–904. 10.1080/00223131.2014.990939 (2015). [Google Scholar]
  • 34.Matsuda, N., Mikami, S., Sato, T. & Saito, K. Measurements of air dose rates in and around houses in the Fukushima Prefecture in Japan after the Fukushima accident. J. Environ. Radioact. 166, 427–435. 10.1016/j.jenvrad.2016.03.012 (2017). [DOI] [PubMed] [Google Scholar]
  • 35.Saito, K. et al. 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, 308–319. 10.1016/j.jenvrad.2014.02.014 (2015). [DOI] [PubMed] [Google Scholar]
  • 36.International Commission on Radiological Protection (ICRP). The 2007 Recommendations of the International Commission on Radiological Protection. ICRP Publication 103. Ann. ICRP 37 (2–4). https://journals.sagepub.com/doi/pdf/10.1177/ANIB_37_2-4 (Elsevier, 2007). [DOI] [PubMed]
  • 37.International Commission on Radiological Protection (ICRP). Radiological Protection of People and the Environment in the Event of a Large Nuclear Accident: Update of ICRP Publications 109 and 111. ICRP Publication 146. Ann. ICRP 49 (4). https://journals.sagepub.com/doi/pdf/10.1177/ANIB_49_4 (SAGE Publications, 2020). [DOI] [PubMed]
  • 38.Cabinet Secretariat of Japan. Basic Concepts for Promoting Decontamination. (Accessed 1 August 2025). https://www.cas.go.jp/jp/genpatsujiko/pdf/20110826001.pdf (2011).
  • 39.Furukawa, M. & Shingaki, R. Terrestrial gamma radiation dose rate in Japan estimated before the 2011 Great East Japan Earthquake. Radiat. Emer. Med.1 (1–2), 11–16 (2012). [Google Scholar]
  • 40.Sanada, Y., Yoshimura, K., Urabe, Y., Iwai, T. & Katengeza, E. W. Distribution map of natural gamma-ray dose rates for studies of the additional exposure dose after the Fukushima Dai-ichi Nuclear Power Station accident. J. Environ. Radioact 223–224. 10.1016/j.jenvrad.2020.106397 (2020). [DOI] [PubMed]
  • 41.Kubo, K., Yanagisawa, Y., Yoshioka, T. & Takahashi, Y. Geology of the Namie and Iwaki-Tomioka District with Geological Sheet Map at 1:50000. Geological Survey of Japan, Tsukuba. https://www.gsj.jp/Map/EN/geology4-7.html#07058 (1994).
  • 42.Imai, N. & Okai, T. Distribution of the natural radiation dose of Japan. Supplement to Geochemical Map of Japan. Geological Survey of Japan, National Institute of Advanced Industrial Science and Technology, Tsukuba. https://gbank.gsj.jp/geochemmap/data/pdf/Nradiation.pdf (2014).
  • 43.International Commission on Radiological Protection (ICRP). Conversion Coefficients for Radiological Protection Quantities for External Radiation Exposures. ICRP Publication 116. Ann. ICRP 40 (2–5). https://journals.sagepub.com/doi/pdf/10.1177/ANIB_40_2-5 (Elsevier, 2010). [DOI] [PubMed]
  • 44.Saito, K. & Petoussi-Henss, N. Ambient dose equivalent conversion coefficients for radionuclides exponentially distributed in the ground. J. Nucl. Sci. Technol.51 (10), 1274–1287. 10.1080/00223131.2014.919885 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Ishikawa, T. et al. External doses available for epidemiological studies related to the Fukushima Health Management Survey: first 4-month individual doses and municipality-average doses for the first year. J. Epidemiol.32 (Suppl. 12), S11–S22. 10.2188/jea.JE20210166 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Sato, O., Nonaka, S. & Tada, J. Intake of radioactive materials as assessed by the duplicate diet method in Fukushima. J. Radiol. Prot.33 (4), 823–838. 10.1088/0952-4746/33/4/823 (2013). [DOI] [PubMed] [Google Scholar]
  • 47.Yamaguchi, I. & Takahashi, H. Estimating of internal radiation doses due to food consumption and its reduction applying the food regulation after the Fukushima nuclear accident using national food-monitoring data. J. Natl. Inst. Public Health. 70 (1), 61–75 (2021). [Google Scholar]
  • 48.Osanai, M. et al. Long-term analysis of internal exposure dose-reduction effects by food regulation and food item contribution to dose after the Fukushima Daiichi Nuclear Power Plant accident. Foods12 (6), 1305. 10.3390/foods12061305 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Orita, M. et al. Determinants of the return to hometowns after the accident at Fukushima Dai-ichi nuclear power plant: a case study for the village of Kawauchi. Radiat. Prot. Dosim. 156 (3), 383–385. 10.1093/rpd/nct082 (2013). [DOI] [PubMed] [Google Scholar]
  • 50.Reconstruction Agency of Japan. Survey of Residents’ Intentions among Those Affected by the Nuclear Disaster. (Accessed 1 August 2025). https://www.reconstruction.go.jp/topics/main-cat1/sub-cat1-4/ikoucyousa/ (2025).
  • 51.Kobayashi, T. et al. Internal migration patterns to Fukushima following the nuclear accident based on vital statistics and the Fukushima health management survey. BMC Public Health25, 1985. 10.1186/s12889-025-23199-w (2025). [DOI] [PMC free article] [PubMed]
  • 52.Nakayama, C. et al. Lingering health-related anxiety about radiation among Fukushima residents as correlated with media information following the accident at Fukushima Daiichi Nuclear Power Plant. PLoS ONE 14 (5), e0217285. 10.1371/journal.pone.0217285 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Hirota, S., Nakayama, C., Yoshinaga, S., Moriyama, N. & Yasumura, S. Relationship between radiation knowledge and radiation effect anxiety on the next generation: an analysis of a questionnaire survey disseminated to residents in Fukushima Prefecture. Jpn J. Public Health. 70 (7), 415–424. 10.11236/jph.22-053 (2023). (in Japanese with English abstact). [DOI] [PubMed] [Google Scholar]
  • 54.Naito, W., Uesaka, M., Kurosawa, T. & Kuroda, Y. Measuring and assessing individual external doses during the rehabilitation phase in Iitate village after the Fukushima Daiichi nuclear power plant accident. J. Radiol. Prot.37 (3), 606–622. 10.1088/1361-6498/aa7359 (2017). [DOI] [PubMed] [Google Scholar]
  • 55.Nomura, S. et al. Low dose of external exposure among returnees to former evacuation areas: a cross-sectional all-municipality joint study following the 2011 Fukushima Daiichi nuclear power plant incident. J. Radiol. Prot.40 (1), 1–18. 10.1088/1361-6498/ab49ba (2020). [DOI] [PubMed] [Google Scholar]
  • 56.Wainwright, H. M., Seki, A., Chen, J. & Saito, K. A multiscale Bayesian data integration approach for mapping air dose rates around the Fukushima Daiichi Nuclear Power Plant. J. Environ. Radioact. 167, 62–69. 10.1016/j.jenvrad.2016.11.033 (2017). [DOI] [PubMed] [Google Scholar]
  • 57.Kinase, S., Sato, S., Sakamoto, R., Yamamoto, H. & Saito, K. Changes in ambient dose equivalent rates around roads at Kawamata town after the Fukushima accident. Radiat. Prot. Dosim. 167 (1–3), 340–343. 10.1093/rpd/ncv275 (2015). [DOI] [PubMed] [Google Scholar]
  • 58.Yajima, K. et al. Estimating annual individual doses for evacuees returning home to areas affected by the Fukushima nuclear accident. Health Phys.109 (2), 122–133. 10.1097/HP.0000000000000308 (2015). [DOI] [PubMed] [Google Scholar]
  • 59.Shimo, M., Minato, S. & Sugino, M. A survey of environmental radiation Aichi, Gifu and Mie prefectures. Trans. Atom. Energy Soc. Jpn. 41 (9), 44–54. 10.3327/jaesj.41.954 (1999). [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

  1. Ando, M., Sasaki, M. & Saito, K. Evaluation of decreasing trend in air dose rate and ecological half-life within an 80 km range from Fukushima Dai-ichi Nuclear Power Plant, using car-borne survey data measured by KURAMA systems up to 2018. J. Nucl. Sci. Technol.57 (12), 1319–1330. 10.1080/00223131.2020.1789008 (2020).

Data Availability Statement

Data will be made available on request.

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES