Introduction

The main cause of human exposure to ionizing radiation is natural radioactivity due to the fact that naturally occurring radionuclides are generally found in the environment. Radon, a decay product of 238U, can accumulate in any building mainly through its daughter isotopes (218Po, 214Pb, 214Bi and 214Po) and as a result considered as the major contributor of about 50% of the annual effective dose received from both natural and anthropogenic sources. Researches have proven that radon is the second leading cause of lung cancer in the world particularly on the continents of America, Asia and Europe [1,2,3]. Furthermore, according to the World Health Organization (WHO), radon is responsible for 3–14% of all recorded lung cancer cases in the world [4].

Since ionizing radiation can be detrimental to human health international bodies has set reference levels for indoor radon concentrations to regulate the concentration of radon in buildings and to reduce the risk to lung cancer. For example, WHO has a reference level of 100 Bq m−3 [4], the Basic Safety Standards Directive of European Union has it at 300 Bq m−3 [5] and the International Commission of Radiological Protection also has a reference value of 300 Bq m−3 [6]. The International Atomic Energy Agency Basic Safety Standards has set this at 200 Bq m−3 for dwellings and 1000 Bq m−3 for workplaces [7]. Radon moves freely in air and can easily enter the human respiratory system when its progenies are attached to dust particles. Also, the entry of radon into an indoor environment can lead to its accumulation to high concentrations. The concentration of radon in a building is dependent on factors such as soil composition, types of building materials used, ventilation and the age of the building. The entry routes of radon into indoor environments may include fissures in connecting parts of buildings, concrete floor and walls, heating and ventilation, drainage pipes and air conditioning ducts [2, 8,9,10,11,12,13,14,15,16,17,18,19,20].

In recent years, a number of indoor radon concentration measurements has been carried out in the African region to assess the radon levels in dwellings and sometimes in workplaces. For example, Aladeniyi et al. [21] measured indoor radon levels in 158 buildings in southwestern Nigeria using CR-39 detectors. The mean indoor radon level recorded in that study was 64.8 Bq m−3 [21]. Afolabi et al. [22] also estimated radon concentrations in 87 selected offices in a Nigerian university using a Pro 3-series radon detector for 48 h and recorded a mean concentration of 33.3 Bq m−3. Mlay and Makundi [23] studied radon concentrations in the vicinity of Manyoni Uranium Deposit in Singida in Tanzania. A total of 41 houses were surveyed using an Alpha-Guard for a period of two months and obtained a mean indoor radon level of 166 ± 12 Bq m−3 [23]. Hussein [24] determined indoor radon levels in 45 dwellings in 15 cities in Egypt using LR-115 detector and observed a mean concentration of 30.9 ± 7.3 Bq m−3. Le Roux et al. [25] estimated indoor radon concentrations in about 60 homes in Vredenburg and Saldanha, South Africa. The study employed the use of electret ion chambers as detectors and obtained mean concentrations of 108 Bq m−3 and 104 Bq m−3 for Saldanha and Vredenburg, respectively [25]. Determining the indoor radon concentrations and subsequently radon mapping for the Greater Accra region of Ghana is of utmost importance as it has capital of the country and with so many infrastructural developments. Some surveys have been conducted to raise the awareness of environmental radioactivity and to project areas that are at risk to radon in the region. Nsiah-Akoto et al. [26] studied the indoor radon levels and its related effective dose rates within Dome in Greater Accra by using Solid State Nuclear Track Detectors (SSNTD) with LR-115 type II detectors. The number of studied dwellings and workplaces were 40 and the corresponding mean indoor radon concentration were within 466.9 ± 1.2 Bq m−3 [26]. Otoo et al. [27] investigated the seasonal indoor radon levels in 228 buildings within the Accra metropolis. The passive radon CR-39 SSNTD was used for his study and they reported that the indoor radon fell between 13.6 and 533.7 Bq m−3 [27]. Kpordzro [28] assessed the factors that influence indoor radon concentrations in Dome and its surrounding areas. The LR-115 type II detector was used to measure indoor radon and soil-gas radon concentration in different seasons. The mean indoor radon concentration was determined to be 165 Bq m−3 and 73 Bq m−3 for rainy and dry seasons, respectively [28].

In spite of these and other studies conducted so far, no radon map has yet been developed for Greater Accra which would aid authorities and experts with regard to the development of an appropriate policy to decrease public exposure to radon in the region. When this is done it would result in the improvement of the quality of life and the general health of the inhabitants of the region. Since there is not enough data on the indoor radon concentrations in some parts of the region, an effort has been made to estimate the indoor radon concentrations in 95 dwellings to evaluate the annual effective dose due to the inhalation of radon. The indoor radon levels in homes were measured by means of National Radiation Protection Board (NRPB) dosimeters with CR-39 detectors. The measurements were taken between December 2021 and March 2022 for a period of 90 days. The annual mean radon concentrations were predicted through the measured concentrations during the period. Successively, a radon map was created with an ArcGIS software and three diverse interpolation methods, within a grid with 5 × 5 km2 dimensions.

Methods and measurements

Study area

The Greater Accra region which is surrounded by the Eastern, Central and Volta regions on the north, west and east, respectively, is the second most populated region in Ghana. It has the smallest area and occupying a total land size of 3,245 km2 representative of 1.4 percent of the total land area of the country. It is the most urbanized region with approximately 87.4% of its total population residing in urban centers. It is located between longitudes 0° 20′ 0″ W and 0° 5′ 0″ W and latitudes 5° 40′ 0″ N and 5° 45′ 0″ N. The topography of the study area is made up of diverse categories of rocks and soils. The major categories of rocks are the quartzites, gneiss, granites, dahomeyan schists and the precambrian Togo series. The soil types comprise of lateritic sandy clay soils, residual clays and gravels and the alluvial and marine motted clay soils [29].

Measuring techniques

The type of residences surveyed in this study were villas, apartments and houses (old and new). The selected residences were in general built with concrete, sandcrete, bricks from clay, cement, gravels and floor tiles. Most of the buildings in this survey are with concrete walls, tiled floors and cemented. They had few windows, small window sizes and some have air conditioners installed in the rooms. Several of the dwellings in the study area were constructed with building materials such as various types of cement, gravels, gneiss rock, sandstone, beach sand, sand, quartzite, clay bricks and tiles. A number of buildings which were more than 50 years old were initially built with sand and clay bricks and were later cemented and floor tiled due to modernization.

A total of 95 dwellings as indicated by the red points in Fig. 1 were studied. Indoor radon concentrations were measured mainly in either living room or bedroom on ground floor dependent on where residents spend most of their time in the house. The measurements were done using NRPB radon dosimeter with CR-39 detector (Baryotrak type) of size 1 cm × 1 cm × 0.1 cm which has a background of 0.5 tracks mm−2. The NRPB type has a Lower limit of detection (LLD) value of 44.5 kBq h m−3 with a minimum measurable radon activity concentration of 20.6 Bq m−3 for an exposure period of 3 months. The CR-39 detector has a sensitivity of 1.5–1.7 (track cm−2) (Bq d m−3) dependent on the duration of storage prior to measurement. The track density of the CR-39 detectors used is 0.00003346 track mm−2 (Bq h mm−3)−1. The etching system has a bulk etch rate of 7 µm h−1 at 90 °C [8, 49] (Fig. 2).

Fig. 1
figure 1

Map of the study area

Full size image
Fig. 2
figure 2

NRPB dosimeter with a CR-39 detector

Full size image

In each home, a detector was placed at a height between 1 and 1.5 m above the ground, the distance of placement was greater than 0.5 m from every wall, and at least 15 cm from every other objects. Concerning the recruitment of participants, homes were selected randomly subject to the extent of the inhabited area. Most of the homes surveyed were constructed between 7 and 80 years ago with bricks made of cement and sand along with tiled floorings. In order to understand the radon concentrations, differences between the building locations, the type of building structures and their ages were taken note of all through the analysis of the results. After the detectors were exposed to radon, they were covered in aluminum foil and conveyed to the environmental laboratory of the Institute of Radiochemistry and Radioecology of the University of Pannonia, Veszprem, Hungary for analysis. In the laboratory, the detectors were washed with deionized water, dried and afterwards chemically etched. Then the CR-39 was etched with the following conditions: 6.25 M NaOH solution, temperature of 90 °C and at a duration of 8 h. The track densities were counted with the aid of a high-resolution image scanner and analysis software. The concentrations of indoor radon were determined with Eq. (1) [19, 20]:

$$ {\text{C}}_{{{\text{Rn}}}} = \left( {{\text{N}}_{{\text{t}}} - {\text{N}}_{{\text{b}}} } \right) \times {\text{E}}/\left( {{\text{T}} \times {\text{A}}} \right) $$
(1)

where CRn stands for the mean indoor radon level in Bq m−3, Nt represents the total number of tracks, Nb is the number of background tracks, E presents the calibration factor in Bq m−3 h tracks−1 mm2, T stands for the period of exposure in hours and A is the reading area of the tracks in mm2. The determination of the calibration factor and its process have been performed with certified radon source in the calibration chamber of the Institute of Radiochemistry and Radioecology, University of Pannonia [20, 30].

Estimation of the annual radiation dose, ELCR and LCC related to indoor radon exposure

The annual effective dose caused by the inhalation of indoor radon were determined with Eq. (2) [19]:

$$ {\text{E}}_{{{\text{Rn}}}} = {\text{C}}_{{{\text{Rn}}}} \times {\text{F}}_{{{\text{Rn}}}} \times {\text{t}} \times {\text{K}}_{{{\text{Rn}}}} $$
(2)

where ERn is the annual effective dose attributable to the indoor radon exposure in mSv year−1, CRn represents the annual mean radon levels in homes in Bqm−3, FRn denotes the indoor equilibrium factor for radon and its daughter isotopes, t is the estimated sum of hours spent indoor per annum and KRn denotes the endorsed dose conversion factor in Bq h m−3. The International Commission on Radiological Protection (ICRP) in 2010 and the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) in 2000 recommended the following values: F = 0.40; t = 7,000 h; KRn = 9 nSv per unit of integrated radon concentrations [4, 31].

The Excess Lifetime Cancer Risk (ELCR) shows the probability of the populace developing excess cancer due to exposure to radionuclides (radon) over a period at a given exposure level. ELCR per 100,000 people was determined with the equation [19, 29]:

$$ {\text{ELCR}} = {\text{E}}_{{{\text{Rn}}}} \times {\text{DL}} \times {\text{RF}} $$
(3)

where ERn is the annual effective dose, DL is the mean lifetime duration projected to be 70 years; and RF represents terminal cancer risk per Sievert given by the ICRP as 5.5 × 10−2 Sv−1.

The Lung Cancer Cases (LCC) for each year per 1,000,000 people was calculated with respect to the lung cancer induction value of 18 × 10−6 mSv−1, hence the equation is expressed as [19]:

$$ {\text{LCC}} = {\text{E}}_{{{\text{Rn}}}} \times {18} \times {1}0^{{ - {6}}} $$
(4)

Statistical analyses of data

The statistical analysis of the results was done with IBM SPSS Statistics version 26. The normal distribution of the data was tested using the Kolmogorov–Smirnov test. A Q–Q plot was similarly used to examine the normal distribution of the data. Log-normal distribution was also done to check normality. Mean, standard deviation, annual mean, geometric mean and geometric standard deviations were determined from the data.

Radon mapping

An indoor radon map is a powerful statistical tool with accuracy and high resolution which helps educating the public on environmental radioactivity. Indoor radon maps are constructed with indoor measurements and are precise since the radon levels are measured at the point of exposure. Radon map is an important resource for any government seeking to reduce the exposure of its populace to radon gas. It influences government policies particularly on building infrastructures. Thus, it has an enormous value in both the economic and social aspect of the public.

An indoor radon concentration map was produced for this survey by means of the ArcGIS software version 10.4.1 over a grid of 5 × 5 km2 dimensional unit. An Arithmetic Mean (AM) was used to estimate the average indoor radon levels for cells with no data observation. This was to obtain concentrations for grid cells where no data was available. The grid cells for the AM were of the dimension 5 × 5 km2. Again, three different interpolation methods were tested for this survey. These were the Inverse Distance Weighting (IDW), Empirical Bayesian Kriging (EBK) and Ordinary Kriging (OK) methods [19]. Afterwards, the accurateness of the above mention techniques was examined using the mean absolute error (MAE) and average standard error (ASE).

Results and discussion

For this work, the indoor radon concentrations in 95 dwellings were measured for a total exposure period of 90 days using the SSNTD technique. The frequency distribution of the indoor radon concentrations for the 95 dwellings evaluated in the Greater Accra for the measurement period are shown in Fig. 3.

Fig. 3
figure 3

Frequency distribution of the indoor radon concentration (Bq m−3) for dwellings

Full size image

Normality of the data

A graphical approach was used to test the normality of the data by means of the cumulative frequency distribution and the normalizing Q-Q plots. These were used for the first analysis of the normality of the measured data. The data was additionally examined with the Kolmogorov–Smirnov normality test.

Figure 3a tested for the normality of the radon concentrations data with the cumulative frequency distribution. At a first look at the graph, one may conclude that the cumulative data is normally distributed. However, a further examination with the Q–Q plot (Fig. 4b) showed the data deviated from the diagonal line indicating they are not normally distributed. A similar situation was observed in Fig. 5a, b in the case of the age of the houses sample.

Fig. 4
figure 4

Normal distribution of indoor radon concentration (a) and (b) Q–Q plot of indoor radon concentration in Bq m−3

Full size image
Fig. 5
figure 5

Normal distribution of the age of houses (a) and (b) Q–Q plot of the age of houses in years

Full size image

Kolmogorov–Smirnov test for normality was performed on the data. The radon concentration and the age of houses data unfortunately failed the test as their p values were less than 0.05 and therefore rejecting the null hypotheses. The absence of the normal distribution of the data could be due to soil compositions, ventilation and the different types of building materials.

The fact that the geometric means (GM) of the radon concentrations (50.0 Bq m−3) and age of houses (12.4 y) were close to their median values of 48.5 Bq m−3 and 12.0 y, respectively is an additional suggestion for a log-normal distribution of the indoor radon concentration. Figures 6 and 7 showed the plot of the log transformed values of the radon concentrations and the ages of the houses involved in this survey. The Kolmogorov–Smirnov test proved the age of the houses and radon data to be normally distributed with significant values of 0.2 and 0.1, respectively. Thus, the log-normal transformation of these data made it possible to accept the null hypotheses.

Fig. 6
figure 6

Log-normal distribution of indoor radon concentration (a) and (b) Q–Q plot of log-normal indoor radon concentration in Bq m−3

Full size image
Fig. 7
figure 7

Log-normal distribution of age of houses (a) and (b) Q–Q plot of log-normal of age of houses in years

Full size image

Indoor radon concentrations

The indoor radon concentrations for the individual homes ranged from 36.1 ± 2.7 to 92.0 ± 5.2 Bq m−3. The determined indoor radon concentrations were below 100 Bq m−3. The major sources of indoor radon concentrations as considered in this study could be pointed to the soil composition and geology, the various types of building materials used in construction, indoor air ventilation within the buildings, soil-gas infiltration and the lifestyle of occupants [19, 27, 32]. Several international organizations including UNSCEAR, WHO, ICRP, U.S. Environmental Protection Agency and the European Union have set out recommended values as guidelines for indoor radon exposure. The UNSCEAR has proposed a worldwide average indoor radon concentration of 39 Bq m−3 [33]. For this study the annual average indoor radon concentration was 50.8 ± 3.4 Bq m−3 exceeding the UNSCEAR reference level. This point out that the inhabitants of the study area are vulnerable to internal exposure of indoor radon and its associated lung cancer effect. However, the results of this study did not exceed the recommended annual average values of 100 Bq m−3 and 300 Bq m−3 proposed by WHO and ICRP respectively [4, 6]. Same was realized for recommended values set by U.S. EPA (2004) at 148 Bq m−3 [26, 34]. Data from this survey were likened to the global geometric mean (GM) and geometric standard deviation (GSD) specified by the UNSCEAR (2000). The GM for the indoor radon in Greater Accra was found to be 50.0 Bq m−3 which is approximately 13 Bq m−3 more than the action limit 37 Bq m−3 and the GSD was 1.2 as compared to 2.2 suggested by the UNSCEAR [31].

The age of buildings was also regarded in evaluating the levels of indoor radon exposure. Some older buildings were identified to have high radon concentrations especially in the densely populated areas as observed in a 25 years old building recording an indoor radon concentration of 82.1 ± 2.5 Bq m−3. This can be attributed to cracks in the building walls and floors aiding in emission of radon gas from the underlying soils into the rooms. However, average concentrations such as 44.8 ± 4.5 Bq m−3 and 56.0 ± 4.0 Bq m−3 were observed in a few older dwellings which were 100 years old. In the case of newly built homes, some recorded high radon concentrations and this can be due to the geology of the area. For example, the highest indoor radon measurement of 92.0 ± 5.2 Bq m−3 was observed in a building which is 9 years old and a concentration of 55.1 ± 5.1 Bq m−3 was obtained for a 1 years old home. Nonetheless, the lowest measurement of 36.1 ± 2.7 Bq m−3 was found in a building which is 13 years old. These observations mean that the indoor radon levels determined in this study did not show a relationship with the age of the dwellings surveyed.

The geology of the study area is known to be composed of rocks such granites, quartzite and gneiss which are mostly used for building constructions in the region. Studies have shown that these rocks have a strong correlation with 222Rn [29, 35]. Yeboah et al. in 2001 studied the natural radioactivity of rocks and soils in the Greater Accra region. Their findings showed that the radioactivity of the soils and rocks depends on their geology. Places like James Town, Weija, Dodowa and Shai Hills had activity concentrations of 2.4–62.7 Bq kg−1 of 238U. Rocks sampled from these places had activity concentrations of 0.7–40.0 Bq kg−1 of 238U, the maximum being from those of the igneous and granite compositions [36]. Previous studies also have shown that rocks with significant radioactivity levels and tectonic faults with the excessive emanation could lead to a major surge in radon levels in dwellings and its impact becomes prevailing in the common dose of the populace [37]. Therefore, it was observed that outcomes from this study corresponded to the findings of Yeboah et al. For example, in this study low to medium indoor radon concentrations were recorded around the Shai Hills which agreed to the low activity concentrations reported by Yeboah et al. for that same area.

Another reason for high indoor radon levels detected in some parts of the study area is the type of building materials used in the building of the homes. Humans are exposed to ionizing radiations such as radon from the radionuclides found in building materials particularly the use of tailings in construction [13]. This exposure is significant especially when such building materials have radiation doses above the normal background levels. Building materials like cement, concrete, bricks etc. are produced from raw materials with significant concentrations of NORMs including 238U, 232Th and 40K. Thus, radon and its progenies can accumulate inside the buildings and their inhalation leads to internal radiation exposures [13, 29]. According to Otoo et al. (2018b), building materials used for constructions in the Greater Accra region contained some levels of these radionuclides. In that study it was also realized that building materials from the Greater Accra region had radon concentrations between 17.4 ± 0.7 Bqm−3 and 42.1 ± 2.8 Bqm−3. The activity concentration of 238U, 232Th and 40K were below the UNSCEAR reference levels of 50 Bq kg−1 (238U and 232Th) and 500 Bq kg−1 (40K) except for activity concentration of 40K in gneiss rock from Shai Hills [29, 38]. Thus, the low NORMs activity concentrations corresponded to low radon levels in the building materials indicating that indoor radon levels to a large extent depends on the building materials used in construction.

Again, as most of the population spend more hours indoors they are more susceptible to inhaling the radon decay products and the situation worsens when there is limited ventilation as frequent air flow into homes help reduce the radon concentration in a building [19]. Hence, in both old and new homes, the kind of building materials used in the constructing of the homes can increase the radon concentrations in the rooms and are the major entry route of radon into homes [8, 18]. Since the radon concentrations in building materials from the study area are within average levels it can be said that the high indoor radon levels measured in this survey is due to lack of adequate ventilations in buildings as few and small window sizes were observed in most dwellings during the survey especially within populated areas. Also due to the lifestyle of residents, windows are mostly when no one is at home, during rainfalls and at night limiting the amount of airflow in the homes. The preference for usage of air conditioners by residents particularly when temperatures are high also can increase the levels of indoor radon. Therefore, the ventilation systems in the region must be improved to reduce indoor radon levels in homes and its associated health risk.

Radiation dose and risk assessment

The study assumed the mean indoor radon concentrations to be the same as the annual mean indoor radon concentrations. The inhalation of radon’s short-lived isotopes is liable for almost half of the total annual effective dose humans receive from the natural sources of ionizing radiation altogether [39]. To evaluate this, Eqs. (2)–(4) were applied to estimate the radiation dose and risk assessment of radon inhalation. The annual effective dose, excess lifetime cancer risk and the estimated lung cases per year caused by inhalation of radon gas in the various homes were calculated. The resultant annual mean effective dose was found to vary between 0.9 and 2.3 mSv year−1 with an average of 1.3 mSv year−1. Given the data above, the average annual indoor radon concentration value of 50.8 Bq m−3 corresponded to an annual effective dose for the entire body of 1.3 mSv year−1. This outcome is below the ICRP reference value of 3.0 mSv year−1. High annual effective doses ranging between 1.4 and 2.3 mSv year−1 from the inhalation of radon were found for the dwellings in the central and South-Western parts of the Greater Accra region. Average values for indoor radon concentration as high as 92.0 ± 5.2 Bq m−3 corresponding to 2.3 mSv year−1 were recorded in these areas. This could be attributed to the fact that the area is very populated and dense with very poor ventilations than in other parts of the region. Sometimes buildings are built closely together with few windows having limited access to airflow. Dwellers prefer to have air conditioners in their homes with doors and windows often closed and in some cases for security reasons than to open them for aeration. Also, lots of building projects are noted to take place in these areas and as a results radon gas is emanated as soil and rocks are being disturbed.

With the aim of evaluating the influence of residential radon exposure on lung cancer, the excess lifetime cancer risk from the indoor radon were determined to range between 3.5 and 8.9 with 4.9 as its annual mean value. For the lung cancer cases per year per 1,000,000 people, its range and mean were calculated as 16.4–41.7 and 23.1, respectively. It has been demonstrated in previous studies that there is an increase in the risk of lung cancer even in the circumstance of normal indoor radon levels as low as 40 Bq m−3 [39]. According to the Global Cancer Observatory, an international agency for research on cancer under WHO, fact sheet for Ghana as of 2020, lung cancer was among the other category of cancers affecting the Ghanaian population with a percentage of 45.1%. Again, lung cancer ranked 8th among the top 10 cancers in the country. The country recorded approximately 535 new cases in 2020 with a total lung cancer death incidence of 487 [40]. Accordingly, in this study, almost 5% of lung cancer deaths annually is attributed to indoor radon exposure in the Greater Accra region and this estimate falls within the WHO estimate of 3–14% [4]. The risk of lung cancer has been recognized to be proportional to indoor radon exposure as made known by the WHO that lung cancer cases increases 16% by every 100 Bq m−3 of radon concentration [4, 19]. Thus, there exist a correlation between the risk of lung cancer and indoor radon exposure.

A comparison of the indoor radon concentrations reported in other countries are provided in Table 1 below.

Table 1 Studies of indoor radon concentrations, geometric means and geometric standard deviations reported in other countries
Full size table

Spatial distribution of indoor radon

In Ghana, explicit guidelines regarding indoor radon levels in either homes or workplaces are not yet in place, consequently a national radon map is required to aid categorize the geographic areas where the populace is at risk of radon exposures and to increase public awareness on the related health risk of radon. The spatial map of the indoor radon levels for the study area was generated with the following interpolation techniques: Inverse Distance Weighting (IDW), Ordinary Kriging (OK) and Empirical Bayesian Kriging (EBK). The first 2 techniques are dependent on the distance between the two points, which is those of observation and estimation in the interpolation [19]. The IDW technique weights the impact of the observed points over the estimated interpolation with respect to the distance only. Thus, its overall idea is built on the theory that the characteristic value of an unsampled point is the weighted mean of known values in the zone [41]. Alternatively, the relationship between the points and formation of a preliminary function, variogram, which can repetitively be updated was done using the OK technique. The OK calculates the semivariogram from identified data points and uses this single semivariogram to make estimates at unidentified points, but in the case of EBK it takes into account the uncertainty as well. Consequently, spatial distribution maps of indoor radon concentration for the residences surveyed in Greater Accra were plotted over a grid with the dimensions of 5 × 5 km2 using ArcGIS software version 10.4.1, as illustrated in Fig. 8a–c.

Fig. 8
figure 8

Predicted indoor radon map by means of a IDW, b OK and c EBK interpolation methods

Full size image

The accuracies of the interpolations employed in the mapping were verified using the Mean Absolute Error (MAE) and Average Standard Error (ASE). The IDW was realized to give more accurate interpolation since it estimates unknown values by means of known values by identifying distance and closest points and assumes that measured values adjoining to prediction points have more influence on the predicted value than those farther away. This interpolation perfectly describes the indoor radon data measured from the dwellings within the region. It also had the lowest MAE and ASE of 7.7 and 11.6, respectively. The map generated by the OK method did not really fit the data since the method depends on the normal distribution of the input data. Furthermore, the measured data was limited compared to the broad size of the study area. Similar situation was realized for the EBK map which needed extra parameters like soil gas, types of soils and faults [42] to make it more suitable for the study area. Hence, the IDW technique demonstrated to be most appropriate for predicting average indoor radon concentrations.

According to Fig. 8A, the IDW spatial distribution map revealed that high indoor radon concentrations were found both in dense and less dense environments. In dense urban cities and towns such as Tema, Ashiaman, Michel Camp, Kaneshie, Pokuase, Amasaman, Madina, Adentan, Dome, Kwabenya, Haatso, Legon Hills, Osu, La and some parts of Accra central including James Town and Palladium had concentrations ranging 58.7 ± 2.6–92.0 ± 5.2 Bq m−3. This can be related to poor ventilations in these areas. Some homes in these areas surveyed were between 60 and 100 years old and such houses had smaller window sizes compared to modern houses. It was observed that some of these houses were initially built with red mud and bricks and later plastered with cement. Less dense towns including Osudoku, Dawhenya, Afienya, Abomeyi, Azakope, Dedukope, Afiayengba, Azizanya and Toje also showed high radon concentrations. These areas are known to have medium ventilations as dwellings there are not crowded as seen in the first category but the reason for their high concentrations can be liken to the former since they have similar building characteristics. Again, as already mentioned building materials also serves as source of entry paths for radon gas into homes.

Conclusions

The annual indoor radon concentrations for dwellings in the Greater Accra region of Ghana were studied. Radon concentrations ranged between 36.1 ± 2.7 and 92.0 ± 5.2 Bq m−3 and the annual mean for the entire region was 50.8 ± 3.4 Bq m−3. Whilst the GM (50.0) was found to be higher than the reference level proposed by the UNSCEAR and the GSD was 1.0 lower. Several factors like soil composition, geology of the area and the type of building materials used in construction were considered to be the cause of indoor radon levels in the homes surveyed. The radiological parameters obtained in all the sampled homes were below the recommended values. The annual effective dose varied between 0.9 and 2.3 mSv year−1 with the lowest belonging to a building which is 21 years old. Subsequently, the excess lifetime cancer risk was evaluated 3.5–8.9. The estimated minimum lung cancer cases for the region was found to be 16 per a million population and an estimated 5% lung cancer deaths annually. In the study it was realized the age of the buildings surveyed did not show a correlation with the measured indoor radon concentrations. Building materials used for home constructions in the Greater Accra region may to an extent contribute to the indoor radon levels in the region. The main source of indoor radon levels as observed in the study is due to lack of ventilation in many dwellings in the study area.

An indoor radon concentration map was created with the ArcGIS software over a 5 × 5 km2 grid to determine high risk areas within the studied area. Three interpolation techniques were employed for this purpose. The arithmetic mean from the indoor radon concentrations was used to predict the averages for the grid cells with no data available. Among the interpolation methods applied, the IDW was identified as the most appropriate in predicting the indoor radon levels in the grid. The other methods could not give suitable predictions due to the lack of normal distribution of the measured indoor radon levels. The spatial distribution of the indoor radon data helped to easily identify areas within the region which are prone or less prone to radon gas and its ensuing health effects. The most apparent remedy for the high-risk zones is to improve the ventilation of the dwellings which permits radon gas to escape.

The outcome of this study is valuable to the government as it would benefit in decision-making processes and drafting national policies and guidelines for indoor radon levels. It is also beneficial to the real estate industry and the general public as it informs them on building decisions and the choice of residences. Given the results of this study and the consequential health impact of radon, it is recommended that more indoor studies be carried out in the region all over Ghana especially in areas where no studies have yet been done by the application of various measuring devices and procedures. It is expected that these studies will take into consideration other aspects such as thoron levels, soil gas, soil permeability, soil and rock types, building materials used, faults, etc. This would aid in the compilation of a national radon map for the country to estimate concentrations and the number of radon related lung cancer cases. Thus, measurement of indoor radon levels in each area is essential in determining the suitability of an environment for a healthy lifestyle.