External effective dose from natural radiation for the Umbria region (Italy)

ABSTRACT

This study presents the map of the external effective annual dose rate (1:200,000 scale) due to terrestrial and cosmic radiation. The terrestrial dose is assessed via gamma ray spectroscopy combining radiometric data from airborne surveys and laboratory measurements. The geostatistical method Collocated CoKriging is used for the spatial interpolation of the sparse gamma ray data, adopting a high-resolution geological map as ancillary information. The obtained numerical map is integrated with the cosmic radiation effective dose rate calculated using the CARI-7 software tool that considers the effects of altitude, latitude, and the solar magnetic activity cycle. The absorbed dose rate due to radioactivity of the main lithological groups is studied and, for the most populated municipalities, the population-weighted average effective dose is also calculated. For future generations, this map will be a reference tool for evaluating radiological effects in case of accidental events like radioactive fallout or environmental contaminations.

KEYWORDS:

• Terrestrial radioactivity
• cosmic rays
• gamma rays
• collocated CoKriging
• public exposure
• effective dose rate

1. Introduction

For each individual living on the Earth, the radiation exposure arisen from artificial sources, including medical uses of radiation and nuclear accidents, generally represents a minor contribution. It follows that the assessment of natural radiation background is the most significant step towards a comprehensive understanding of the population exposure and an optimization of good radiation safety practices. Moreover, the knowledge of the natural background permits to discern and detect a possible increase of radiation caused by accidental events linked to nuclear power plant functioning (Fuma et al., 2017; Strand et al., 2014), to the release of radioactive pollutants in the environment (Devi & Chauhan, 2020; Salmani-Ghabeshi et al., 2016) or to the enhancement of naturally occurring radioactive materials (Ali et al., 2019; Xhixha et al., 2015).

In the health physics, the effective dose is an approximate measure of the stochastic risk applied to a reference person (Fisher & Fahey, 2017) introduced with the purpose of setting limits for radiation protection (ICRP, 1977). From the physics point of view, the effective dose is the energy deposited in the matter by ionizing radiation weighted for the type of radiation and radiosensitivity of organs and tissues. Since these weighting factors are affected by considerable uncertainties (i.e., age and gender of individual), the effective dose should not be used to estimate cancer incidence, in particular for low dose rates.

The annual committed dose from natural radiation includes the sum of internal dose (E_INT), due to the inhalation or ingestion of naturally occurring radionuclides in food and gases, and the external dose (E_EXT) generated by cosmic rays and terrestrial radionuclides and absorbed by an individual living both outdoor and indoor. In the last decades, diverse attempts were made to investigate the outdoor dose rate through maps of the cosmic and terrestrial doses at regional (Szabo et al., 2017; Yesilkanat et al., 2015), national (Chen et al., 2009; Quindós Poncela et al., 2004) or continental (Cinelli et al., 2017; Szegvary, Conen, et al., 2007) scale. While the cosmic contribution can be computed by utilizing analytical models implemented in dosimetry software tools (FAA, 2021; Sato, 2015), the terrestrial component can be retrieved by the measurement of radionuclides abundances through gamma ray measurements (Cinelli et al., 2020; Verdoya et al., 2009).

The cosmic dose (E_C), on average, contributes for about 50% of E_EXT and it can considerably vary in time and space, being mainly dependent on the altitude and, secondly, on the geomagnetic latitude and the solar activity (Omori et al., 2020). The terrestrial dose (E_T) depends on the amount of the terrestrial radionuclides (²³⁸U, ²³²Th and ⁴⁰K) in the environment, whose spatial distribution is mainly driven by local geological features (Aközcan, 2014; Torres et al., 2018). The investigation of E_T has also indirect implications for predicting the concentration of geogenic radon, a noble gas produced by the uranium decay series, which plays a starring role in the origin of inhalation dose (Kropat et al., 2017; Szegvary, Leuenberger, et al., 2007). The presence of natural radioactivity in rocks employed as building materials is also the main source of the indoor exposure (Puccini et al., 2014).

In this study, we present the map of the external effective dose from natural radiation in the Umbria region (Italy). Thanks to the great variety of lithotypes and to the distribution of the population in a relatively wide range of altitudes, this region represents an ideal case for testing the effective dose mapping. The map includes the contributions from cosmic radiations, calculated using data from the CARI-7 software tool, and from terrestrial sources, evaluated on the basis of gamma ray spectroscopy measurements on rock and soil samples performed in laboratory and through Airborne Gamma Ray Spectroscopy (AGRS) surveys.

2. Material and methods

2.1. Study area

Umbria is a region of Central Italy not bordering with any geographical or political boundary and is called, for this reason and for the color of its lush meadows and hills, the ‘green heart of Italy’. Its territory hosts a total of 865,013 inhabitants (ISTAT, 2021) distributed in 2 provinces (Perugia and Terni) and 92 municipalities. The region extends for 8474 km², offering a wide variety of geomorphological features through the succession of valleys, mountain ranges, highlands, and plains. Partly hilly and mountainous, and partly flat and fertile, its topography includes a portion of the central Apennines, with the highest point at Mt. Vettore (2476 m a.s.l.) on the northern border with the Marche region, then gently descending towards its western part, reaching its lowest point at Attigliano village (96 m a.s.l.), on the southern border with Lazio. Umbria hosts sedimentary rocks (limestone, sandstone), igneous rocks (lava, tuff) and deposits of unconsolidated quaternary sediments (less than 2.6 My old). Accurate information on surfacing lithologies and on the morphology of the territory are of fundamental importance in view of planning and performing an effective sampling campaign aimed at a meticulous dose assessment.

Adopting the geological information contained in the 1:10,000 scale map of the Umbrian Regional Geological Database (Motti & Natali, 2016; Regione Umbria – Servizio Geologico, 2014), the 215 cartographic units of the region were grouped in two distinct categories, (i) Rocky Formations and (ii) Quaternary Deposits, investigated following different approaches (Table 1 and Figure 1).

Figure 1. Map of the Rocky Formations (RF, in violet) and Quaternary Deposits (QD, in pink) of Umbria, together with the flight track followed by the AGRS aircraft (in grey) and the points in which samples were collected on rocks (in red) and soils (in green). Cartographic reference system WGS 84, UTM Zone 32N.

Map of the Umbria region colored to identify Rocky Formations and Quaternary Deposits with points showing the position where rocks and soils samples were collected and lines showing the flight tracks of the AGRS aircraft.

Table 1. Surface area (in km² and in percentage with respect to the total area of the region) and number of cartographic units of Quaternary Deposits (QD) and Rocky Formations (RF) together with the information about the rock and soil sampling campaign and the airborne data taking.

	Quaternary Deposits (QD)	Rocky Formations (RF)	Global
Area (km^2)	4222	4252	8474
Area (%)	50	50	100
Cartographic units	116	99	215
Rock samples	11	272	283
Soil samples	0	14	14
Airborne samples	5614	1528	7142
Area/samples (km^2)	0.8	2.3	1.1

Following the sampling strategy of Callegari et al. (2013) and Strati et al. (2015) the cartographic units of Rocky Formations (RF), relatively homogeneous in composition and geological history, were characterized via an extensive sampling campaign during which were collected 272 rock samples from fresh outcrops (Figure 2(a)) belonging to 99 different cartographic units. The samples were analysed through high-resolution gamma ray spectroscopic techniques via High Purity Germanium (HPGe) detectors (Figure 2(b)). In addition, 14 soil samples were collected in areas undisturbed and morphologically suitable for pedogenesis and similarly analysed (Table 1).

Figure 2. (a) Outcrop of a Rocky Formation (RF) where a rock sample was collected and then measured in laboratory with the (b) MCA_Rad system comprised of two HPGe detectors, an automatic sample changer, and a copper and lead shielding. (c) Aerial photo of a Quaternary Deposit (QD), investigated through airborne gamma measurements performed by (d) the Radgyro aircraft, equipped with NaI(Tl) detectors.

Four panels with pictures of a rocky outcrop, the MCA_Rad detector system, a field with some trees photographed from above, and the Radgyro on a meadow.

The Quaternary Deposits (QD) comprise (i) the sediments resulting from landslides and groundwater debris deposits, (ii) alluvial and terraced deposits and (iii) the successions of the smaller basins characterizing the valley areas (Figure 2(c)). The cartographic units QD are usually heterogeneous and possibly subject of high spatial variability not imputable to their classification according to the cartographic units. For this reason, these deposits were investigated via AGRS surveys with a custom-equipped aircraft (Figure 2(d)), collecting a total of 5614 spectra during the overflights. Additionally, 11 QD rock samples were collected to compare the two different analyses employed for RF and QD measurements. Complementary, 1528 AGRS measurements were performed on RF units (Table 1).

2.2. Gamma ray measurements

The measurements of the specific activity of U, Th and K of the rock samples were performed using the MCA_Rad (Figure 2(b)), an equipment comprised of two coaxial p-type HPGe detectors specifically designed and calibrated for natural radioactivity measurements (Xhixha et al., 2016). A copper and lead shielding enables the suppression of laboratory background by approximately two orders of magnitude, permitting to reach an overall (statistical and systematic) uncertainty of U, Th and K activity less than 5% (Xhixha et al., 2013).

The AGRS surveys were performed via the Radgyro (Baldoncini et al., 2017; Baldoncini et al., 2018) (Figure 2(d)), an aircraft expressly engineered to be a flying multisensorial platform devoted to gamma ray spectroscopic measurements. The Radgyro features three u-blox EVK-6T GPS antennas for the logging of the geographic latitude and longitude and seven altimetric sensors for the estimation of the flight altitude (Alberi et al., 2017). The AGRS measurements are performed with the AGRS_16L, a modular NaI(Tl) scintillation detector composed of four 4L–crystals and placed in the middle of the Radgyro hull. The acquired gamma spectra are then processed offline to obtain information on U, Th and K specific activities with a time resolution of 10 s, comparable to a measurement every 300 m of flight at typical speed (100 km/h) and altitude (100 m) for AGRS surveys (IAEA, 2003).

2.3. Mapping radiometric data

Combining AGRS, rock and soil samples, a total of 7142 gamma measurements were collected during the data-taking campaign. The Collocated CoKriging, a geostatistical interpolation algorithm (Guastaldi et al., 2013), was used to map the collected sparse and heterogeneous specific activity data over the entire Umbria region. This algorithm predicts the spatial distributions by characterizing the coregionalization between primary under-sampled variables wherein a secondary correlated variable is available at all prediction locations. In this case, the secondary variable, built by assigning a numeric value to each cartographic unit of the 1:10,000 geological map of Umbria (Motti & Natali, 2016; Regione Umbria – Servizio Geologico, 2014), steers the spatial estimation of the specific activities (primary variable) of U, Th and K in the entire territory (Figure 3). The result was a 50m–resolution specific activity map of the Umbria region for each of the three radioelements.

Figure 3. Schematic sketch explaining the rationale of the Collocated CoKriging interpolation algorithm. The experimental semi-variograms (ESV) describing the spatial correlation of the under-sampled primary variables (specific activities of U, Th and K) and the ESV of the secondary variable (the geological information) are calculated and modelled. The two ESVs are combined to calculate and model the cross-semi-variogram (XESV), permitting the prediction of the value of the primary variable in all points of the output grid in which the secondary variable is available.

A six-section figure with three experimental semi-variograms and corresponding models for primary and secondary variables and cross experimental semi-variogram, geological map with the position of the gamma measurements, colored geological units and radioactivity map.

2.4. Mapping dosimetric data

The spatial information on the specific activities was used to derive the terrestrial absorbed dose rate $D_{\mathrm{T}}\left[{\displaystyle \frac{\mathrm{nGy}}{\mathrm{h}}}\right]$ according to: (1) $D_{\mathrm{T}}=C_{\mathrm{K}}\cdot A_{\mathrm{K}}+C_{\mathrm{U}}\cdot A_{\mathrm{U}}+C_{\mathrm{Th}}\cdot A_{\mathrm{Th}}$(1) where A_K, A_U and A_Th are the specific activities of the three radioelements, in Bq/kg, in the outdoor soil, $C_{\mathrm{K}}=0.0417$, $C_{\mathrm{U}}=0.462$, $C_{\mathrm{Th}}=0.604$ are the dose conversion factors in (nGy/h)/(Bq/kg) for K, U and Th, respectively (UNSCEAR, 2008). The terrestrial effective dose $E_{\mathrm{T}}^{\mathrm{\prime }}\left[{\displaystyle \frac{\mu \mathrm{Sv}}{\mathrm{yr}}}\right]$ received by an adult spending the whole year living outdoor is calculated as: (2) $E_{\mathrm{T}}^{{}^{\mathrm{\prime }}}=D_{\mathrm{T}}\cdot \mathrm{CF}\cdot T\cdot {10}^{-3}$(2) where $\mathrm{CF}=0.7\frac{\mathrm{Sv}}{\mathrm{Gy}}$ is the conversion factor from absorbed dose in air to the effective dose received by adults and $T=8760\mathrm{h}$ are the hours in a year.

The cosmic flux depends mainly on the 11-year solar modulation, the altitude, and the geomagnetic coordinates. The cosmic effective dose $E_{\mathrm{C}}^{{}^{\mathrm{\prime }}}$ in a specific location and datetime can be calculated using software dedicated to its modeling. For this purpose, we use CARI-7, a dosimetry program developed by the USA Federal Aviation Administration’s (FAA, 2021).

Neglecting the seasonal variations due to atmospheric conditions producing a $E_{\mathrm{C}}^{{}^{\mathrm{\prime }}}$ fluctuation of ∼ ± 2.5%, the main factor affecting the time variability of $E_{\mathrm{C}}^{{}^{\mathrm{\prime }}}$ is the solar cycle causing a full-range variation of ∼10% between its maximum and its minimum. Since solar activity is inversely correlated to cosmic radiation intensity at ground, the mean value of the cosmic effective dose was calculated by averaging the estimates obtained in correspondence of the last maximum (April 2014) and last minimum (December 2019) of solar activity.

An assessment of the latitudinal variability of the $E_{\mathrm{C}}^{{}^{\mathrm{\prime }}}$ showed that, for the Umbria region, the discrepancy between the northernmost and the southernmost point at the same altitude is lower than 3%. It was therefore chosen to set in the software the geographical coordinates to Perugia (43.1167 N, 12.3833 E), situated in the middle of the region.

The dose in the 0–3000 m altitude range was modeled averaging $E_{\mathrm{C}}^{{}^{\mathrm{\prime }}}$ calculated with CARI-7 in the above-mentioned maximum and minimum solar conditions and then fitting these values with an exponential curve. The obtained best fit function was used to parametrize the effective cosmic dose received by an adult spending the whole year living outdoor $E_{\mathrm{C}}^{\mathrm{\prime }}\left[{\displaystyle \frac{\mu \mathrm{Sv}}{\mathrm{yr}}}\right]$according to the altitude h [m] a.s.l.: (3) $E_{\mathrm{C}}^{\mathrm{\prime }}=259.9\cdot e^{(5.786\cdot {10}^{-4})\cdot \mathrm{h}}$(3) The $E_{\mathrm{C}}^{\mathrm{\prime }}$ map of the Umbria region was therefore reconstructed by employing the altitude obtained from a 10m-resolution Digital Elevation Model (DEM) of Umbria (Tarquini et al., 2007).

2.5. Building the external effective dose map

The annual external effective dose $E_{\mathrm{EXT}}$ is the sum of the effective terrestrial dose ($E_{\mathrm{T}}$) and the effective cosmic dose ($E_{\mathrm{C}}$), each including an outdoor contribution ($E_{\mathrm{T}}^{\mathrm{OUT}}$and $E_{\mathrm{C}}^{\mathrm{OUT}}$) and an indoor contribution ($E_{\mathrm{C}}^{\mathrm{IN}}$and $E_{\mathrm{T}}^{\mathrm{IN}}$): (4) $E_{\mathrm{EXT}}=E_{\mathrm{T}}+E_{\mathrm{C}}=E_{\mathrm{T}}^{\mathrm{OUT}}+E_{\mathrm{T}}^{\mathrm{IN}}+E_{\mathrm{C}}^{\mathrm{OUT}}+E_{\mathrm{C}}^{\mathrm{IN}}$(4)

The outdoor terrestrial effective dose $E_{\mathrm{T}}^{\mathrm{OUT}}\left[{\displaystyle \frac{\mu \mathrm{Sv}}{\mathrm{yr}}}\right]$ received by an adult in a year is calculated as: (5) $E_{\mathrm{T}}^{\mathrm{OUT}}=E_{\mathrm{T}}^{\mathrm{\prime }}\cdot (1-\mathrm{OF})$(5) where $\mathrm{OF}=0.8$ (UNSCEAR, 2000) is the annual indoor Occupancy Factor and $E_{\mathrm{T}}^{\mathrm{\prime }}\left[{\displaystyle \frac{\mu \mathrm{Sv}}{\mathrm{yr}}}\right]$is the terrestrial effective dose calculated according to Equation (2).

The indoor terrestrial effective dose ($E_{\mathrm{T}}^{\mathrm{IN}}$) is originated by the radionuclides in building materials as well as the radiation produced by the outdoor soil penetrating the dwellings. It can be evaluated on the basis of the average value ratio of indoor to outdoor absorbed dose rate $R=1.4$(UNSCEAR, 2000). It follows that $E_{\mathrm{T}}^{\mathrm{IN}}\left[{\displaystyle \frac{\mu \mathrm{Sv}}{\mathrm{yr}}}\right]$ is calculated as: (6) $E_{\mathrm{T}}^{\mathrm{IN}}=E_{\mathrm{T}}^{\mathrm{\prime }}\cdot R\cdot \mathrm{OF}$(6) The outdoor ($E_{\mathrm{C}}^{\mathrm{OUT}}$) and indoor ($E_{\mathrm{C}}^{\mathrm{IN}}$) contributions to the annual cosmic effective dose ($E_{\mathrm{C}}$) are calculated as: (7) $E_{\mathrm{C}}^{\mathrm{OUT}}=E_{\mathrm{C}}^{\mathrm{\prime }}\cdot (1-\mathrm{OF})$(7) (8) $E_{\mathrm{C}}^{\mathrm{IN}}=E_{\mathrm{C}}^{\mathrm{\prime }}\cdot \mathrm{OF}\cdot \mathrm{SF}$(8) where $E_{\mathrm{C}}^{\mathrm{\prime }}\left[{\displaystyle \frac{\mu \mathrm{Sv}}{\mathrm{yr}}}\right]$ is the cosmic effective dose calculated according to Equation (3) and $\mathrm{SF}=0.8$, the Shielding Factor, is the fraction of cosmic dose not shielded by dwellings (UNSCEAR, 2000).

Combining Equation (4) with Equations (5–8) and adopting $\mathrm{OF}=0.8$, $R=1.4$ and $\mathrm{SF}=0.8$, the external effective dose is calculated as: (9) $E_{\mathrm{EXT}}=E_T^{\mathrm{\prime }}\cdot 1.32+E_C^{\mathrm{\prime }}\cdot 0.84$(9)

3. Results and discussion

In quantitative mapping, the huge effort put in data collection and in spatial interpolation methods has to be accompanied by equal care in the map production. The representation of results has to be achieved avoiding loss of information and featureless maps (Reimann, 2005). In this study, the chromatic variations in the Main Map of the legends of E_EXT, E_T and E_C were assigned to specific values of dose rate corresponding to specific percentiles of the data distribution (Figure 4). Since the use of order statistics does not assume any underlying data distribution, this strategy appears particularly suitable for the observed positive skewed distributions. The frequency histogram of E_T (Figure 4(b)) suggests a bimodal distribution with a tail on high values which can be explained by the strong relationship between terrestrial dose and the feature of the three main lithologies (carbonates, sandstones, and volcanic rocks) of the geological map.

Figure 4. Relative frequency distributions of the data obtained for (a) the total outdoor effective dose (E_EXT), (b) terrestrial effective dose (E_T) and (c) cosmic effective dose (E_C) together with the percentiles chosen for the chromatic variations of the legend colour bars in the Main Map. For graphical needs, the tails of the histograms corresponding to the 1.6% of data for E_EXT, to the 1.8% for E_T and to the 0.1% for E_C are not reported.

Three histograms showing the frequency distributions of the external, terrestrial, and cosmic effective dose rate. Vertical dashed lines identify the percentile values used for the legend construction.

The lognormal tendency of the frequency histogram of E_C values (Figure 4(c)) reflects the exponential dependency of cosmic dose on the altimetric values of the investigated area. As a consequence of these peculiarities, the percentiles used for the three maps are different, aiming to show the spatial distribution of the data structures. Moreover, the percentiles used as lower limits for the highest classes of E_EXT, E_T and E_C correspond to the global averaged values reported in (UNSCEAR, 2008) for the external (870 µSv/yr), terrestrial (480 µSv/yr) and cosmic (390 µSv/yr) effective dose rates.

The terrestrial absorbed dose rate (D_T, see Equation (1)) in the RF of the Umbria region can be studied in a propaedeutic way for the evaluation of the radiological implications coming from the use of these rocks as building materials. About 90% of the Umbrian production in the mining sector concerns the extraction of inert materials from limestones, clays, and sandstones for civil and industrial uses. The remaining market sector is represented by the extraction of tuff employed as ornamental stone for civil use. In this perspective, four different groups of rocks are identified on the basis of lithological and tectonic arguments: (i) carbonates (Coccioni et al., 2013; Di Naccio et al., 2005) and (ii) turbiditic deposits of the Umbro-Marchean series (Barsella et al., 2009; Tinterri & Muzzi Magalhaes, 2011), (iii) sandstones of Tuscan units (Amendola et al., 2016; Barsella et al., 2009) and (iv) magmatic rocks of the Vulsini volcanic districts (Palladino & Simei, 2002).

The boxplots of D_T (Figure 5), obtained starting from the activity concentrations measured from rock samples, show a clear distinction between sedimentary and magmatic rocks. The carbonates (limestone and marls, Triassic – Paleogene) of the Umbro-Marchean series, which occupies 45% of the RF surface area, present the lowest values of D_T with a median of 7 nGy/h; the high outliers represent samples having a siliciclastic component (marly mudstones). Despite the similar deposition environment, the sandstones of the Tuscan units (Macigno formation, Paleogene – Miocene) show a slightly increasing trend with respect to the turbiditic deposits of the Umbro-Marchean series (sandstone and marly mudstones, Miocene); for both cases, the low outliers refer to the most carbonate portions of the sequences. The magmatic rocks (tuffs and porphyritic basalts, Pleistocene), corresponding to 3% of the RF surface area, are characterized by a wide range of values (86–389 nGy/h) with two high outliers from tephritic rocks.

Figure 5. Boxplots of the absorbed dose rate D_T (in nGy/h) calculated according to Equation (1) starting from the activity concentrations measured in the rock samples of the (i) carbonates (127 samples, 45% of the RF area) and (ii) turbiditic deposits of the Umbro-Marchean series (95 samples, 33% of the RF area), (iii) sandstones of Tuscan units (31 samples, 13% of the RF area) and (iv) magmatic rocks of the Vulsini volcanic districts (14 samples, 3% of the RF area).

A boxplot graph showing the terrestrial effective dose rate of the four categories of rock formations of the Umbria region. The magmatic rocks category is significantly higher than the other three.

The relative contributions of the terrestrial (E_T) and cosmic (E_C) components to the external effective doses (Figure 6) reproduce the effect of two concurrent causes: from one side the most elevated areas (high cosmic doses) in the eastern region are dominated by the presence of limestones and marls (low terrestrial doses), on the other side the outcrops of magmatic rocks (high terrestrial doses) are in relatively low altitudes (low cosmic doses) in the area south-west of Orvieto. In the alluvial plains and the hills in the central zone of the region, where the main rocks are the turbiditic deposits of the Umbro-Marchean series and quaternary deposits, the terrestrial and cosmic contributions are approximately equivalent.

Figure 6. Annual (a) terrestrial (E_T) and (b) cosmic (E_C) effective dose rate of the Umbria region expressed as a relative contribution in percentage of the external effective dose rate (E_EXT). Cartographic reference system WGS 84, UTM Zone 32N.

Two maps of the Umbria region are shaded with different tones of a color representing the relative contribution to the external effective dose rate of terrestrial and cosmic dose. The higher contribution of the terrestrial dose is in the south-west area of Umbria, while the cosmic dose contribution is higher in the south-east area.

In Figure 7, the population-weighted average of the external effective dose (<E_EXT>), together with its terrestrial (<E_T>) and cosmic (<E_C>) components, is reported for the most populated (>10,000 inhabitants) municipalities of the Umbria region hosting globally 80% of its population. E_EXT normally ranges between 457 and 671 µSv/yr with approximately equal terrestrial and cosmic components. A separate mention has to be dedicated to Orvieto, where the terrestrial component is dominant (83%) since this city is located in the south-west of the region and on the flat summit of a large butte of volcanic tuff at a relatively low altitude (325 m a.s.l.).

Figure 7. Population-weighted average external effective dose rate (<E_EXT>) for the municipalities of the Umbria region with more than 10,000 inhabitants; the terrestrial (<E_T>) and the cosmic (<E_C>) contribution are reported respectively with green and blue filling. The number in red for each bar indicates the percentage of the population of the municipality with respect to the total regional population.

Bar graph representing the weighted average of the outdoor effective dose rate for the most populated municipalities of the Umbria region. Orvieto has a substantially higher average than all the other municipalities.

4. Final remarks

In this study, the map of the annual external effective dose rate (E_EXT) of the Umbria region (Italy), including the terrestrial and cosmic components, is presented. For the terrestrial contribution, the starting points are the specific activities of U, Th and K measured by means of gamma ray spectroscopy (airborne and in laboratory) and spatialized in the entire territory by applying the Collocated CoKriging. The terrestrial effective dose (E_T) is calculated assuming that an adult spends 80% of their time inside dwellings (OF = 0.8) and a ratio R = 1.4 between the indoor and outdoor terrestrial dose. The cosmic effective dose (E_C) is calculated using the CARI-7 software tool and considering that 80% of the cosmic radiation is not shielded indoor (SF = 0.8).

Comparing the estimates obtained for the Umbria region with the global estimates reported in UNSCEAR (2008), it is possible to infer that only for the 2% of its territory the external effective dose is higher than the global average (870 µSv/yr), while the population-weighted effective dose of the region is 566 µSv/yr. In the investigated municipalities (Figure 7), E_EXT is always lower than the global average value except for Orvieto, where the average value of E_EXT= 1375 µSv/yr is dominated by the high terrestrial dose of the surrounding volcanic rocks.

As it is easy to guess, the values of OF, SF and R depend on multiple features, e.g., people life style from one side and materials of the buildings on the other. Omori et al., (2020) measured a value of R ranging between 0.72 (light-gauge steel houses) and 2.8 (concrete houses). According to UNSCEAR, (2000), SF can be 1 for wooden houses and 0.4 for concrete buildings leading to the universal representative value of 0.8. Sato (2016) calculates the SF as a function of the wall thickness and proposes a slightly higher representative value (SF = 0.91). The variability of these coefficients is an additional source of uncertainty associated with effective dose estimates.

In conclusion, with the caveat that the external effective dose is calculated using reference coefficients affected by intrinsic variabilities, the presented map represents an instrument for understanding the environment hosting approximately 865,000 people and for optimizing radiation protection procedures. Because it is an imperfect tool, this map should not be used for epidemiological studies since more biological and statistical factors must be considered for a more appropriate evaluation of health effects on the population. This cartographic result informs the population that ionizing radiation from its own territory is under control since it was properly measured. Moreover, the public source radiometric data can be used for further studies (e.g. building materials characterization). For the future generations, the map of external effective dose from natural radiation can be used as a reference by the Umbria region community in case of accidental events like radioactive fallout or environmental contaminations.

Software

The geostatistical analysis was performed using Geovariances ISATIS®. The cosmic dose was calculated using the CARI-7 software tool. The map was produced using Esri ArcGIS 10.7. The radiometric data can be downloaded here: https://www.regione.umbria.it/geologia/carta-della-radioattivita-naturale-dell-umbria

Acknowledgments

The authors thank Maurino Antongiovanni, Paola Fabbri, Giovanni Fiorentini, Vincenzo Guidi, Claudio Pagotto and Emanuele Tufarolo for their valuable support.

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The data that support the findings of this study are available from the corresponding author, K.G.C.R., upon reasonable request.

Additional information

Funding

This work was partially funded by the National Institute of Nuclear Physics (INFN) through the ITALian RADioactivity project (ITALRAD). The authors would like to acknowledge the support of (i) Project ‘Protocolli Operativi Scalabili per l’agricoltura di precisione – POSITIVE’ – CUP: D41F18000080009, (ii) Project ‘Monitoraggio degli sversamenti illegali attraverso l’impiego sinergico di tecnologie avanzate – C4E’ – CUP: B46C18000750005, (iii) FAR 2019–2020 of the University of Ferrara, and (iv) GEOexplorer Impresa Sociale S.r.l.