Introduction

The exposure of human beings to ionizing radiation from natural sources is related to high-energy cosmic ray particles, human activities, construction materials and geological structures; and both external and internal exposures to humans arise from these sources1. Gamma-rays emitted from natural sources are largely due to 238U series and 232Th series elements and their decay products, as well as 40K which exist at trace levels in the earth’s crust2,3,4,5. The measurement of these natural sources allows determination of the background baseline of natural radioactivity6 and estimation of the radiation hazard risks to humans7,8. It is well known that construction materials of concrete, brick, sandstone, granite and other natural stones contain these natural sources. Depending on their amounts, these sources may cause changes in radiation level of the built-up environment which means the environment made by the natural environment, artificial structures (i.e., roads, buildings and bridges) and human activities; as a consequence, the radiological risks from such natural radioactivity will be changed9,10. It can be assumed that such changes may be occurring as a result of remarkable economic growth accompanied with newly constructed artificial structures.

In Vietnam, the real GDP (gross domestic product) growth rate in 2018 was 7.1%, and that was the highest growth rate in the past 10 years. Therefore, an impact can be assumed on the radiation level from natural radionuclides contained within the materials of newly constructed artificial structures, especially roads and high-rise buildings, resulting in changed radiation level and hazard risks related with them. The contents of natural radionuclides in the soil are very important because these radionuclides contribute significantly to the collective dose of the general population11. Some researchers have measured activity concentrations of natural radionuclides in soil in Vietnam12,13,14. The most recent observations13 for all of Vietnam showed that the average concentrations of 238U, 232Th and 40K were 43 ± 18 Bq kg−1, 60 ± 20 Bq kg−1 and 442 ± 230 Bq kg−1, respectively, which are 1.2, 2.0 and 1.0 times higher than the worldwide averages1. On the other hand, the total annual effective dose6, radium equivalent activity and external hazard index were estimated to be 0.54 mSv year−1, 160 Bq kg−1 and 0.43, respectively; as a consequence, it was concluded that exposure to Vietnamese soil as a building material (primarily as brick) was not an increased hazard for the human population14. A similar assessment was carried out in Sri Lanka15 and Ireland16, and similar results were reported. On the other hand, there are areas which have discouraged use of building materials from high natural background radiation areas and one example is Iran10. However, there has been no investigation to assess changes in radiation level of the built-up environment which is made by artificial structures such as roads, buildings and bridges constructed from building materials. The presence of building materials containing natural radionuclides in urban areas can be expected to cause either an increased radiation level or a decreased radiation level17. The former results from the presence of building materials which contain high concentrations of natural radioactivity sources18,19. For example, bricks contain such radioactive materials as uranium and thorium. While the levels of radiation present in bricks are low, they are higher than in some other building materials used for homes, such as wood20. The latter results from a shielding effect toward terrestrial gamma-rays and it depends on the amount and distribution of the building materials and the background level21. The investigation of radiation level in relation to changes of the built-up environment in urban areas can provide a better understanding for environmental impact assessment of urban development. A way to investigate this impact is offered by the direct measurement of absorbed dose rates6 in air 1 m above the ground surface from terrestrial gamma-rays and artificial structures, followed by a comparison of the estimated absorbed dose rates in air 1 m above the ground surface based on activity concentrations of 238U, 232Th and 40K in soil samples (i.e., from only terrestrial gamma-rays).

In this study, the measurements of absorbed dose rates in air 1 m above the ground surface were carried out for all of Vietnam. From them the detailed dose distribution map of the background related with natural radionuclides contained in soil and building materials was obtained; this map can be used for investigating external effective doses6,18. Additionally, external effective dose and the external hazard index were calculated using data obtained from direct measurements of absorbed dose rate in air and from activity concentrations in soil samples collected at 426 locations. Finally, the impact on the radiological hazard index of artificial structures utilizing low-level radiation building materials in urban areas was estimated. This study is the first large-scale attempt in Vietnam to assess changes of radiological risk on the built-up environment in relation to economic growth.

Results

Absorbed dose rates in air and activity concentrations of 238U, 232Th and 40K

The absorbed dose rates in air 1 m above the grand surface were measured by the car-borne survey technique for 58 provinces and five cities in Vietnam (Fig. 1a). This survey was carried out by positioning a dosimeter inside the car and driving on asphalt pavements (Fig. 1b). The measured dose rates inside the car were corrected by multiplying with shielding factors in order to represent unshielded external dose rates above a bare surface. The detailed dose rate distribution map for all of Vietnam is shown in Fig. 2. A heterogeneous distribution of absorbed dose rate in air was seen (n = 80,516). Dose rates of over 160 nGy h−1 were found in the north mountain area (#1 in Fig. 1a) whereas lower dose rates of under 35 nGy h−1 were found in the southeast area (#43, #45 and #64 in Fig. 1a). The average (range) of absorbed dose rate in air for all of Vietnam was estimated to be 75 ± 32 nGy h−1 (8–463 nGy h−1). Those values for northern Vietnam (#1–31 in Fig. 1a; n = 33,605) and southern Vietnam (#32–64 in Fig. 1a; n = 46,911) were 84 ± 39 nGy h−1 (19–463 nGy h−1) and 68 ± 24 nGy h−1 (8–239 nGy h−1), respectively; hence, the dose rates measured in northern Vietnam tended to be higher than those in southern Vietnam. The detailed dose rates in each municipality are shown in Fig. 3a. The maximum combined relative standard uncertainty involved in absorbed dose rate in air was 11.5%22,23. The highest dose rate was observed in Lai Chau Province (#1 in Fig. 1a), and the average (range) absorbed dose rate in air was 159 ± 42 nGy h−1 (59–306 nGy h−1), whereas the lowest value was observed in Binh Phuoc Province (#45 in Fig. 1a), and the average (range) dose rate was 36 ± 10 nGy h−1 (12–73 nGy h−1). Based on those calculated absorbed dose rates in air, the annual effective doses of all of Vietnam (n = 80,516) were 0.37 mSv year−1 (0.04–2.27 mSv year−1) for the annual indoor effective dose (IAED) and 0.09 mSv year−1 (0.01–0.57 mSv year−1) for the annual outdoor effective dose (OAED). The average of the total annual effective dose (IAED + OAED) was 0.46 mSv year−1 (0.05–2.84 mSv year−1) and that was similar to the worldwide average (0.48 mSv year−1)21. The detailed estimated annual effective dose for each province is summarized in Fig. 3b.

Figure 1
figure 1

(a) Map showing Vietnam administrative divisions consisting of 53 provinces and five cities. While Phu Quoc Island (#64) is part of Kien Giang Province (#62), it is given here as a separate listing due to its island character and size. (b) The survey routes (red lines) for measuring the count rates. Total distance traveled was 28,201 km. The red circles represent points for fixed-point measurements and collection of soil samples (n = 462). These maps drawn using the Generic Mapping Tools, version 4.5.18 (https://gmt.soest.hawaii.edu/gmt4/).

Full size image
Figure 2
figure 2

Distribution map of absorbed dose rates in air 1 m above the ground surface measured in the car-borne survey. This map was drawn using 80,516 data. This map drawn using the Generic Mapping Tools, version 4.5.18 (https://gmt.soest.hawaii.edu/gmt4/).

Full size image
Figure 3
figure 3

(a) Box-and-whisker diagram of absorbed dose rates in air in all municipalities in Vietnam based on the measurements in the car-borne survey (n = 80,515). (b) Annual indoor (IAED) and outdoor (OAED) annual effective doses in all municipalities.

Full size image

The distribution maps of 238U, 232Th and 40K in Bq kg−1 for all of Vietnam are shown in Fig. 4. The activity concentrations of each natural radionuclide were calculated from direct measurements of the gamma-ray pulse height distributions 1 m above the ground surface by fixed-point observations at 462 locations (red circles in Fig. 1b). The maximum combined relative standard uncertainty involved in activity concentration in air was 9.6%22,23. Higher activity concentrations of 238U over 170 Bq kg−1 were observed in the north mountain area (#1 in Fig. 1a) whereas lower activity concentrations under 35 Bq kg−1 were observed in Bac Giang Province (#13 in Fig. 1a) and southern Vietnam excluding the south central coast area. Contrary to this tendency, higher activity concentrations of 232Th over 70 Bq kg−1 and of 40K over 700 Bq kg−1 were observed in the south central coast area (#32, #34 and #35 in Fig. 1a). The detailed activity concentrations of 238U, 232Th and 40K in all of Vietnam and in each municipality can be found in Table 1 and Supplementary Table S1, respectively. The calculated average (range) of absorbed dose rates in air obtained from fixed-point measurements (n = 462) was 71 ± 28 nGy h−1 (20–217 nGy h−1), and the detailed dose rates for each municipality are shown in Table 2. Based on those analyzed activity concentrations, the annual effective doses of all of Vietnam (n = 462) were 0.38 mSv year−1 (0.03–1.05 mSv year−1) for the IAED and 0.09 mSv year−1 (0.01–0.26 mSv year−1) for the OAED as shown in Table 3. The average of the total annual effective dose (IAED + OAED) was 0.47 mSv year−1 (0.04–1.31 mSv year−1) and that was slightly higher value compared to calculated values from direct measurement. Additionally, that was similar to the worldwide average (0.48 mSv year−1)21.

Figure 4
figure 4

Activity concentrations of the natural radionuclides obtained from absorbed dose rates in air and soil in all municipalities in Vietnam (n = 462): (a) 238U, (b) 232Th and (c) 40K. These maps drawn using the Generic Mapping Tools, version 4.5.18 (https://gmt.soest.hawaii.edu/gmt4/).

Full size image
Table 1 Activity concentrations of natural radionuclides from direct measurements and soil concentrations.
Full size table
Table 2 Calculated absorbed dose rate in air from direct measurements and soil concentration.
Full size table
Table 3 Calculated annual effective doses from direct measurements and soil concentrations.
Full size table

Activity concentrations of 238U, 232Th, 40K and Raeq in soil samples

The activity concentrations of 238U, 232Th, 40K and Raeq in soil (n = 462) for each municipality are shown in Fig. 5.

Figure 5
figure 5

Activity concentrations for each municipality of (a) 238U, (b) 232Th, (c) 40K and (d) radium equivalent in soil.

Full size image

The soil samples were collected at the same locations as those for fixed-point measurements (Fig. 1b). The average activity concentrations of 238U, 232Th, 40K and Raeq for all of Vietnam were 35 ± 22 Bq kg−1, 57 ± 33 Bq kg−1, 551 ± 432 Bq kg−1, 159 ± 82 Bq kg−1, respectively; in other words, lower activity concentrations of 238U and higher activity concentrations of 232Th, 40K and Raeq were observed compared to data obtained from direct measurements (Table 1). The detailed activity concentrations of 238U, 232Th and 40K in soil in each municipality can be found in Supplementary Table S2. The maximum combined relative standard uncertainties involved in soil measurement were estimated to be 8.0% for 238U, 3.0% for 232Th and 2.3% for 40K22,23. The calculated average absorbed dose rates in air from soil concentrations of natural radionuclides are shown in Table 2. The average ratio of the difference (direct measurement/soil concentration in absorbed dose rate in air) for all of Vietnam was 0.9 (0.5–2.1); hence, a slightly lower dose rate was observed for the direct measurement.

External radiological hazard index

The calculated external hazard index values from direct measurements and from soil concentrations for each municipality are shown in Fig. 6a. The average Hex calculated from direct measurements and from soil concentrations for all of Vietnam was 0.4 ± 0.2 with a range from 0.1 and 1.3 (Table 1) but there were large differences between those from direct measurements and soil concentrations depending on the municipality. In particular, Long An (#51), King Giang (#62) and Phu Quoc Island (#64) in the Mekong Delta had over 1.5 times higher Hex values calculated from direct measurements compared to those from soil concentrations, whereas Son La (#16), Ha Noi (#17), Ha Nam (#22), Thai Binh (#23), Ninh Binh (#24), Gia Lai (#41), Binh Phuoc (#45), Vinh Long (#56) and Can Tho (#57) had over 1.5 times higher Hex values for the same comparison. The detailed values for all of Vietnam can be found in Table 1. The cumulative frequency distributions of the external hazard index calculated from direct measurements and soil concentrations (n = 462) are shown in Fig. 6b. The cumulative frequency distributions showed clear differences in the Hex between the measurement methods. About 65% of the Hex values calculated from direct measurements using all data were lower compared to those from soil concentrations.

Figure 6
figure 6

(a) External hazard index for each municipality in Vietnam. (b) Cumulative frequency distributions of external hazard index.

Full size image

Discussion

Determining natural radioactivity is of considerable importance for investigations of impacts on radiological risks due to artificial structures of urban development, radiation exposure in natural high background areas, and future exploitation of nuclear energy as well as investigations of the distribution of radionuclides in organs. According to the UNSCEAR report findings21, the average (range) of the outdoor absorbed dose rates in air in 25 countries was 59 nGy h−1 (18–93 nGy h−1). The results in the present study for all of Vietnam from direct measurement in the car-borne survey (75 ± 32 nGy h−1; n = 80,516) and fixed-point measurement (71 ± 28 nGy h−1; n = 462) were 1.2–1.3 times higher values compared to the worldwide average, but within the range.

The distribution map of absorbed dose rate in air had a heterogeneous distribution. It was found that 79% of all measured data exceeded the worldwide average, and 17% of all data exceeded 100 nGy h−1. The gamma radiation emitted by natural sources present in the earth’s crust, in building materials, air, water, food and the human body is largely due to primordial radionuclides, mainly 40K, 238U and 232Th series elements and their decay products. These primordial radionuclides have contributions of 35% for 40K, 25% for 238U and 40% for 232Th to the absorbed dose rate in air as gamma radiation24; consequently, the dose distribution depends on their activity concentrations25. Figure 7 presents a geographical map for Vietnam. The distribution of activity concentrations of these radionuclides mainly depended on basement geology as has been reported in the literature26. Southern Vietnam which showed a lower absorbed dose rate in air mainly has soil formed from basalt and ancient alluvial and poor nutrient sediments. Thus, the activity concentration of 40K is very dependent on dose rate because not much 238U and 232Th are contained in this rock and these sediments14. On the other hand, northern Vietnam has been formed by multiple types of rocks, including granite, tonalite, mafic rocks, granodiorite and migmatite. In those types, higher radiation levels are associated with igneous rocks such as granite which contains monazite (232Th), allanite (232Th) and zircon (238U and 232Th)27. The worldwide average activity concentrations of 40K, 238U and 232Th are 400 Bq kg−1, 35 Bq kg−1 and 30 Bq kg−1, respectively21, meaning that the average values of 40K and 232Th in the soil of all of Vietnam were 1.4 and 1.9 times higher than the worldwide average values.

Figure 7
figure 7

The geological map of Vietnam. These maps drawn using the Generic Mapping Tools, version 4.5.18 (https://gmt.soest.hawaii.edu/gmt4/).

Full size image

Direct measurements of absorbed dose rate in air generally showed reasonable agreement with the calculated absorbed dose rate in air from soil concentration results. In previous studies, considerable discrepancies were observed for some countries such as Luxembourg, Sweden, Syria and Albania depending on the activity concentrations of natural radionuclides in soil; and the UNSCEAR report noted that a discrepancy of 30% or more indicated that a single survey should not be considered representative for a country21. Thus, the obtained result (ratio of the difference: 0.9) in the present study (Table 2) could be taken as representative for Vietnam.

For 65% of all direct measurement data (n = 462), a lower value (ratio of difference < 1.0) was obtained compared to that from soil concentrations. The absorbed dose rates in air in the built-up environment are significantly affected by artificial structures such as surface and elevated roads, bridges and buildings, and the effect is different depending on the concentration of natural radionuclides contained in the construction materials22,28. If many materials containing high concentrations of natural radionuclides are utilized, it is expected that the absorbed dose rate will increase. In the present study, lower concentrations of natural radionuclides which were below the worldwide average were observed (Table 1). Thus, the newly constructed built-up environments in those areas where are represented by a metropolitan area such as Ha Noi City (#17; population density, 2013 person/km2 in 2011) might have utilized low background building materials, and they might have functioned as shielding materials toward terrestrial gamma-ray radiation. On the other hand, the ratios of the difference observed at Binh Phuoc (#45; population density, 628 person/km2 in 2011) and Dong Nai (#46; population density, 268 person/km2 in 2011) Provinces were also low (0.5), the same as the urban Ha Noi City, but this was attributed to the shielding effect by water because lakes occupy large parts of those provinces. These areas have not been affected by urbanization and it is expected that the difference in the absorbed ratio in air will not be changed in the future.

The remaining 45% of all data (n = 462) were higher (ratio of difference > 1.0) compared to the ratios from soil concentrations. In particular, higher values over 1.5 were observed in the built-up environment of An Giang (#55) and King Giang (#62) Provinces and on Phu Quoc Island (#64). That could be explained by differences of the asphalt cover construction method. In southern Vietnam, crushed granitic stones with diameters of about 10–15 cm have been used under the asphalt cover as a roadbed17, and as a result a higher absorbed dose rate in air has been measured on the asphalt compared to the dose rate measured on a bare surface, unlike in northern Vietnam (Supplementary Figure S1). Additionally, activity concentrations of natural radionuclides in soil of those areas had lower values (Fig. 5), and it was expected that those building materials had been brought from another area. In view of balancing building materials under the asphalt cover and original activity concentrations in soil, absorbed dose rates in air increased in those areas. If the survey area is limited to a small area, it might be possible to use the change of absorbed dose rates as an indicator of urbanization for that area by continuing direct measurement of absorbed dose rate.

The Hex clearly changed depending on urbanization, landform, construction methods, types of building materials and original activity concentrations in soil (Fig. 6). The Hex of the built-up environment for mainly urban areas which were densely covered by artificial structures decreased, whereas that on mainly rural areas formed by the low background basement geology had increased Hex compared to the original value which was estimated from soil concentrations. The impact of building materials on the built-up environment related to urbanization can be assessed as ranging from 0.4 to 2.3 times higher Hex with the average Hex value being 0.9 times higher based on the original value. Such future changes can be expected in other developing countries.

Changes of absorbed dose rate in air in the built-up environment in individual countries21 can be classified broadly into two groups based on absorbed dose rate in air measured from soil concentrations (i.e., activity concentrations in soil) as shown in Fig. 8. If the higher dose rate comes from the soil concentration, building materials have functioned as shielding materials. The different amount of change depends on many factors such as density of artificial structures, architectural styles and contents of natural radionuclides in building materials, Vietnam is classified into this group same as Norway and Sweden, etc. Conversely, if the lower dose rate comes from soil concentrations, the building materials themselves have functioned as sources, and changes of absorbed dose rate in air in the built-up environment will be large. Even when evaluated locally for Ha Noi City (#17) and Phu Quoc Island (#64), this finding is consistent with world trends. Hence, the changes of absorbed dose rate in air in the built-up environment after urban development can be partly predicted by measuring activity concentrations in soil.

Figure 8
figure 8

Measured absorbed dose rates in air from direct measurements and from soil concentration, and ratios of differences (direct measurements/soil concentrations) for individual countries.

Full size image

The UNSCEAR report21 estimated that the average worldwide exposure to natural radiation sources is 2.4 mSv year−1. External terrestrial radiation accounts for 0.48 mSv year−1. Exposure from natural radiation sources can be categorized into the existing exposure situations presented by ICRP18 and IAEA29. Additionally, IAEA considered the situation in which the activity concentration of no radionuclide in either the uranium decay chain or the thorium decay chain exceeds 1 kBq kg−1 and the activity concentration of 40K does not exceed 10 kBq kg−1. The reference levels for this exposure situation was set as 1–20 mSv year−118,29. In Vietnam the average annual doses calculated from the direct measurements and from the soil concentrations are below this reference level. Therefore, it can be considered that there is no new radiological risk from the built-up environment for Vietnam.

Methods

Car-borne survey

The measurements of the count rates from natural radionuclides such as 40K, 238U series and 232Th series elements were carried out using a car-borne survey technique during August and December 2015, July 2016, February and August 2017, March and August 2018 and March 2019 in all of Vietnam (331,212 km2). A car-borne survey technique is very useful to make a fast assessment of the dose rate in a large area30,31. The survey routes encompassed 58 provinces and five cities (Fig. 1a). Main roads excluding express ways were used to the extent possible, primarily centered on residential areas (Fig. 1b). The total distance of the survey routes was 28,201 km long. The routes travelled over asphalt pavements and measurements used a 3-in. × 3-in. NaI(Tl) scintillation spectrometer with a global positioning system (EMF-211, EMF Japan Co., Osaka, Japan). This spectrometer had been calibrated by an accreditation body according to ISO/IEC 17,025. The NaI(Tl) scintillation spectrometer was positioned 1 m above the ground surface at the center of the car. There were two researchers and a driver in each car during the measurements. Measurement of count rate inside the car was performed every 30 s while the car was moving with a speed around 40 km h−1. Latitude and longitude at each measurement point were measured at the same time as the count rates, and the count rates within gamma-ray energies of 50 keV–3.2 MeV were recorded. The contribution of cosmic rays to the gamma-ray pulse height distribution was subtracted using the energy stored in the range from 3.0 to 3.2 MeV32. The photon peaks of 40K (Eγ = 1.464 MeV) and 208Tl (Eγ = 2.615 MeV) were used for gamma-ray energy calibration from the channel number and gamma-ray energy before the measurements. The peak position was determined accurately by smoothing the gamma-ray pulse height distribution. Because count rates were measured inside the car, shielding by the car body was also estimated by making measurements inside and outside the care at 30-s intervals during 2 min at 462 locations (red circles in Fig. 1b). Those measurements were done above asphalt surfaces. The shielding factor related to the car body (SFcarbody) was calculated from the correlation between count rates inside and outside the car. In this study, SFcarbody values were found of 2.01 and 1.71. The correlation between count rates inside and outside the car is shown in Supplementary Figure S2. Additionally, those measurements were done above bare surfaces for 2 min to estimate shielding effect by asphalt pavement (SFaspahlt) at 319 locations and SFaspahlt values were found of 1.37 for southern Vietnam and 0.96 for northern Vietnam. The correlation between dose rates in air and count rates outside the car is given in Supplementary Figure S1. The gamma-ray pulse height distributions were also measured outside the car above bare surfaces for 10 min, at 462 locations (red circles in Fig. 1b) for estimating the dose rate conversion factor (DCF) (nGy h−1/cps). The gamma-ray pulse height distributions were then unfolded using the 22 × 22 response matrix method33 and absorbed dose rates in air1 were calculated. These calculated dose rates were used to estimate DCF as the correlation between dose rates and count rates outside the car because it is difficult to obtain the photon peaks in the 30-s measurement of the car-borne survey. In this study, DCF of 0.14 nGy h−1/cps was found. The correlation between count rates on the asphalt surface and absorbed dose rate in air is given in Supplementary Figure S3. The above three factors were multiplied by the corrected count rates outside the car, and the absorbed dose rates (nGy h−1) were calculated. Thus, the absorbed dose rate in air outside the car at 1 m above the ground surface (Dair) can be calculated using Eq. (1):

$${D}_{air}={C}_{in}\times {SF}_{carbody}\times {SF}_{asphalt}\times DCF$$
(1)

where Cin is the count rate (cps) inside the car obtained by the measurements for every 30-s interval. All obtained data from the car-borne surveys were plotted on a distribution of absorbed dose rates in air in Vietnam using a minimum curvature algorithm of GMT34. This is the method for interpolating data by presuming a smooth curved surface from the data of individual points.

Annual effective dose

The conversion coefficient from absorbed dose rate in air to effective dose6,18 (0.7 Sv G year−1) for an adult, and the outdoor (0.2) and indoor (0.8) occupancy factors21 were used for estimating the annual outdoor (OAED) and indoor (IAED) annual effective doses based on the absorbed dose rate in air (Dair) measured from the car-borne survey; OAED and IAED (mSv year−1) were calculated using Eqs. (2) and (3)25.

$$OAED={D}_{air}\times 0.7\times 8760\times 0.2\times {10}^{-6}$$
(2)
$$IAED={D}_{air}\times 0.7\times 8760\times 0.8\times {10}^{-6}$$
(3)

Calculation of activity concentrations of 238U, 232Th and 40K from measured absorbed dose rate in air

The gamma-ray pulse height distributions obtained 1 m above bare surfaces at 462 locations (Fig. 1b) were converted to the energy bin spectrum of incident gamma-rays which is a distribution of gamma-ray flux density to each energy bin to estimate the distribution of activity concentrations of 40K, 238U series and 232Th series for all of Vietnam using the 22 × 22 response matrix method33. In this method, energy bins which are unequal intervals in the gamma-ray energy range from 0 to 3.2 MeV were set: 1.464 MeV for 40K, 1.765 MeV and 2.205 MeV for 214Bi (238U series) and 2.615 MeV for 208Tl (232Th series). The energy intervals for the bins were given from the literature33. The calculation for the 22 × 22 response matrix for the 3-in. × 3-in. NaI(Tl) scintillation spectrometer was done using the Monte Carlo code, SPHERIX35. The gamma-ray flux density and dose rate per unit solid angle were assumed to be almost isotropic in a natural environment. After unfolding the gamma-ray pulse height distribution, clear peaks from 40K (energy range 1.39–1.54 MeV), 214Bi (energy range 1.69–1.84 MeV and 2.10–2.31 MeV) and 208Tl (energy range 2.51–2.27 MeV) were observed in the spectrum. Additionally, the gamma-ray flux densities per unit activity concentrations of the 40K, 238U series and 232Th series were calculated to evaluate each activity concentration (Bq kg−1) of the natural radionuclides using the one-dimensional Monte Carlo gamma transport code36.

Measurement of activity concentration in soils

The soil samples were collected from a layer extending to 15 cm below the ground surface at 462 locations (red circles in Fig. 1b). Samples were dried for 24 h at 110 °C and then sieved. Particles less than 2 mm in size were retained for the activity concentration (Bq kg−1) measurement. After homogenizing the sieved samples, the particle samples were packed in a U-8 polypropylene container. Activity concentrations of 40K (Eγ = 1.461 MeV), 214Pb (Eγ = 0.351 MeV) and 214Bi (Eγ = 0.609 MeV) for 238U series and 228Ac (Eγ = 911 MeV) for 232Th series in each sample were measured for 30,000 s with a high-purity germanium semiconductor detector (GMX10P, ORTEC, Oak Ridge, TN). The standard volume source (U8 type; 5, 10, 20, 30, or 50 mm height) which had been calibrated by an accreditation body according to ISO/IEC 17025 was for estimating counting efficiency determination. This source contained nine radionuclides (Mn-54, Cr-51, Co-57, Co-60, Sr-85, Y-88, Cd-109, Cs-137, Ce-139). The relative efficiency of this detector was 17.50%, and activity concentrations were calculated by the absolute method. The activity concentration of 238U series was calculated using a weighted average of the activity concentrations of 214Pb and 214Bi37.

Absorbed dose rate estimated from activity concentrations in soil

There is a direct connection between activity concentrations in soil and terrestrial gamma-ray radiation. The absorbed dose rate in air 1 m above the ground surface can be calculated based on the measured activity concentrations of 238U, 232Th and 40K using Eq. (4)38:

$${D}_{soil}=0.43{A}_{U}+0.666{A}_{Th}+0.042{A}_{K}$$
(4)

where AU, ATh and AK are the concentrations of 238U, 232Th and 40K in Bq kg−1, respectively.

Radium equivalent activity (Ra eq)

Radium equivalent index in Bq kg−1 is a widely utilized radiological hazard index. It is a convenient index to compare the specific activities of samples containing different amounts of 238U, 232Th and 40K39.

$${Ra}_{eq}={A}_{U}+1.43{A}_{Th}+0.077{A}_{K}$$
(5)

It has been assumed that 370 Bq kg−1 of 238U or 259 Bq kg−1 of 232Th or 4,810 Bq kg−1 of 40K produce the same gamma dose rate and that gamma dose rate is related to the external gamma dose and internal dose due to radon and its daughters such as thoron.

Radiological hazard index (H ex)

The external hazard index due to gamma radiation was calculated using Eq. (6)9,25:

$${H}_{ex}=\frac{{A}_{U}}{370}+\frac{{A}_{Th}}{259}+\frac{{A}_{K}}{4810}\le 1$$
(6)

where AU, ATh and AK are the concentrations in Bq kg−1 of 238U, 232Th and 40K, respectively. When the external hazard index values did not exceed the acceptable limit which is less than unity (Hex < 1), radiation hazards could be assumed to be negligible in the study area.