Seasonal Variations in Effective Radiation Dose in Residential Buildings of the Akmola Region: Assessing the Impact of Basement Presence and Proximity to Uranium Tailings

Abstract

Residential exposure to radon and environmental gamma radiation poses a significant public health concern in uranium-rich regions. The Akmola Region of Kazakhstan, home to one of the world’s largest uranium tailings sites, lacks localized data on seasonal exposure variations and associated health risks. This study assessed indoor radon progeny concentrations and gamma dose rates in 62 dwellings across two settlements—Aqsu and Zavodskoy—in the Akmola Region during autumn 2023 through summer 2024. Using RAMON-02 and Alpharad Plus detectors, seasonal equivalent equilibrium volumetric activity (EEVA) of radon progeny and effective doses were calculated, stratified by presence of a cellar. In Aqsu, ambient dose equivalent rates reached up to 0.55 µSv/h, and winter median EEVA levels exceeded 130 Bq/m³ in some non-cellar homes. Seasonal effective doses peaked in spring (up to 8.82 mSv) in cellar dwellings, with annual doses reaching 23.5 mSv—substantially higher than in Zavodskoy. Although mitigation efforts have reduced exposure in some homes, several cellar dwellings in Aqsu exhibited persistently elevated EEVA, suggesting potential structural vulnerabilities or residual contamination. These findings underscore significant seasonal and structural disparities in radiation exposure and highlight the need for targeted, site-specific interventions to reduce long-term health risks in affected communities.

1. Introduction

Radon (Rn-222) is a naturally occurring radioactive gas that emanates from the decay of uranium-238 in soils and rocks [1]. As it migrates into buildings, particularly poorly ventilated and underground structures, it becomes a dominant source of indoor radiation exposure [2]. The International Agency for Research on Cancer (IARC) classifies radon as a Group 1 carcinogen, designating it the second leading cause of lung cancer after tobacco smoking [3]. Epidemiological evidence confirms that long-term exposure to elevated radon concentrations, even at moderate levels observed in dwellings, significantly increases lung cancer risk in both smokers and non-smokers [4,5,6]. However, the health effects of chronic radon exposure are complex, as systematic review of inflammatory biomarker analysis results shows potential anti-inflammatory effects of chronic radon exposure [7]. Given these dual implications and regional variations in radon levels, continuous monitoring and risk assessment are essential for evidence-based public health interventions, particularly in high-risk areas [8].

Kazakhstan, a former hub of Soviet-era uranium production, hosts extensive uranium tailings repositories that pose persistent environmental and health risks to nearby communities [9]. Historical uranium mining operations in Kazakhstan have generated over 200 million tons of radioactive waste, with an estimated total radioactivity of approximately 9.25 × 10¹⁵ Bq (250,000 Ci) [10]. The Akmola Region, including the Zerendi District, lies adjacent to one of the world’s largest uranium waste storage sites, raising concerns about dispersion of radioactive contaminants including sources of radon and gamma radiation into residential areas [11,12]. Open-pit mining and ore processing have further contributed to localized contamination, with documented hotspots of elevated radionuclide concentrations in soils and buildings [13,14]. While current mining operations (including gold, platinum, palladium, and molybdenum extraction) contribute to the regional economy, they also present additional radiation exposure pathways through open-pit mines and associated activities. The combination of historical contamination and ongoing mining impacts creates complex public health challenges. Inadequate remediation efforts, including only partial removal of contaminated materials, failure to comply with sealing standards, and the absence of systematic radiation monitoring, have allowed radioactive byproducts to migrate into populated zones, potentially exacerbating health risks through chronic radiation exposure, carcinogenic and genotoxic effects [11,15,16].

The health risks of indoor radon exposure in the Akmola Region are further amplified by climatic conditions characterized by a harsh continental climate with prolonged, extreme winters, and residents typically remain indoors for extended periods from late autumn through early spring [17]. This high indoor occupancy level significantly increases cumulative indoor radon exposure, especially in poorly ventilated dwellings and those containing cellars, where radon concentrations are notably elevated [18]. These exposure patterns coincide with concerning health outcomes. The Akmola Region demonstrates elevated rates of respiratory and oncological diseases, with lung cancer incidence substantially exceeding national averages. While Kazakhstan’s 2023 national oncology report documents an all-cancer incidence rate of 186.1 per 100,000 population (including 19.5 lung cancer cases per 100,000), the Akmola Region reports significantly higher rates of 240.1 and 32.6 cases per 100,000, respectively, ranking among the nation’s highest [19]. Several epidemiological studies have suggested that environmental radiation exposure may contribute to these regional cancer disparities [20,21]. Nevertheless, there remains a critical lack of site-specific, seasonally stratified data quantifying radon and gamma radiation exposure in rural populations living near uranium waste sites.

The previous studies have identified elevated background radiation linked to industrial sources [11,13], comprehensive assessments integrating seasonal radon variations, gamma radiation measurements, and cancer risk estimations are lacking. Specifically, comparative studies between uranium-affected industrial and background communities are absent, limiting the ability to discern localized exposure disparities. This study addresses these gaps by conducting detailed indoor radon progeny and environmental gamma radiation measurements, calculating seasonal and annual effective doses in two rural settlements (Aqsu and Zavodskoy) near uranium tailings, with consideration for cellar presence and building characteristics.

The aim of this study is to compare indoor radon progeny concentrations between the settlements of Aqsu and Zavodskoy in the Akmola Region, focusing on three key factors: (1) structural differences in residential buildings, particularly the presence of cellars that facilitate radon entry from the subsoil; (2) environmental sources of radon progeny, including proximity to open-pit mines; and (3) seasonal variations in radon progeny levels to better understand exposure dynamics and associated health risks.

2. Materials and Methods

2.1. Study Area

Measurements were conducted in the settlements of Aqsu and Zavodskoy, both located in the Akmola Region, and both are located close 3–4 km to large tailings storage facilities for uranium waste (Figure 1). The distance between these two settlements is approximately 2.5 km. Aqsu is primarily known for its mineral extraction industry, which includes mines and quarries producing valuable resources such as gold, platinum, palladium, and molybdenum. The area contains two active open-pit mines, which are the main sources of radon activity. The population of both Aqsu and Zavodskoy is estimated at around 4000 residents each. Zavodskoy served as the comparative control area, characterized by the absence of open-pit mines and minimal residential cellar presence. A total of 62 indoor and outdoor radon progeny concentration measurements were taken across the two settlements—45 in Aqsu and 17 in Zavodskoy. The survey included single-story private residences in both areas. Notably, fifteen homes in Aqsu contained cellars, compared to none in Zavodskoy.

2.2. Measurement of Environmental Gamma Radiation

Environmental gamma radiation levels were quantified using ambient dose equivalent dose rate (H*(10)) measurements. Outdoor assessments were conducted at a standardized height of 1 m above ground level. Indoor measurements followed a systematic spatial sampling protocol, including: (1) central room positions at 1 m above the floor, (2) floor surface evaluations, and (3) triplicate wall surface measurements at 0.25 m intervals along each wall. All radiation measurements were performed using RKS-01-SOLO dosimeters (SOLO LLP, Kazakhstan). This device uses a scintillation-based gamma detection unit (scintillator type), with an energy detection range of 20 keV to 10 MeV and a dose rate detection range of 0.01 to 500 µSv/h and has a reported instrumental uncertainty of 15% as per the manufacturer. The manufacturer also provides a standardized calibration protocol registered with the Republican State Enterprise KazInMetr (RSE “KazInMetr”) of Kazakhstan. To minimize the influence of short-term environmental factors, each measurement point was assessed three times, and the arithmetic mean of these triplicate measurements was used for analysis. Measurements were repeated across different seasons. The study adhered to International Atomic Energy Agency (IAEA) technical guidance [22] for field measurements.

2.3. Measurement of Indoor Radon Progeny Concentration

Radon progeny measurements were conducted during the autumn and winter of 2023 and spring and summer of 2024. For the present analysis, the dwellings were stratified into dwellings with cellar and no cellar. The equivalent equilibrium volumetric activity (EEVA) of radon progeny was measured using two complementary semiconductor-based detection systems. The RAMON-02 radiometer (Solo LLP, Almaty, Republic of Kazakhstan) was employed to quantify alpha emissions from aerosol-attached decay products. These products were collected on a filter via air sampling at a flow rate of 15 L/min. Concurrently, the Alpharad Plus system (DOZA SPZ, Russian Federation) provided spectral resolution of alpha emissions via a silicon detector. This system also collects radon progeny (specifically polonium-218 and polonium-214) on an aerosol filter using an internal air blower (sampling duration: 180 s), followed by automatic filter transfer (10 s) to the measurement position, and filter counting for 120 s. The total cycle time is 5 min and 10 s. The measurement outputs include EEVA of radon and thoron progeny, the equilibrium factor (F), and the air exchange coefficient (KV). The Alpharad system covers a dynamic range of 1–2 × 10⁶ cps with ±30% measurement uncertainty.

All measurements were conducted under standardized indoor conditions to capture maximum exposure potential. Measurements took place in a frequently occupied living spaces (bedrooms and living rooms) with windows closed for at least 30 min prior to and during measurement, with instruments positioned at breathing height (1–1.5 m above floor level) and away from walls or ventilation sources to minimize airflow interference. The RAMON-02 measurements involved 120 s air sampling period followed by a 120 s filter counting period. Measurements were conducted three times consecutively during the same day, between 9:00 AM and 6:00 PM, and geometric means of triplicate measurements were calculated for each location to account for temporal variability. The selection of this measurement window (9:00 AM–6:00 PM) was determined by the agreement established between the participating households and the research team, which permitted access to dwellings only during daytime hours. We acknowledge that diurnal variations in indoor radon progeny concentrations may influence this timeframe, particularly with respect to night-time exposures in bedrooms. Site selection in Aqsu prioritized residential dwellings near historical uranium and gold mining operations to evaluate potential geological influences on indoor radon concentrations [23]. To identify dwellings with elevated radon levels, we applied a permissible radon gas threshold of 300 Bq/m³, adjusted using an equilibrium factor (F-factor = 0.4) to account for the radon-progeny relationship in homes [23]. Consequently, a derived progeny threshold of 120 Bq/m³ was used in this analysis.

2.4. Seasonal Effective Dose

The seasonal and annual effective doses from indoor radon exposure were calculated based on the dose assessment methodology provided in the UNSCEAR 2000 report [24]. The seasonal effective dose ($E_S$) was estimated using the following equation:

$$E_S=C_S\times T_S\times D$$

(1)

where

$C_S$ is the seasonal radon progeny concentration (Bq/m³) in each dwelling;

$T_S$ is the seasonal residential occupancy time (hours), derived from participant questionnaire responses (1987 h for winter; 1752 h for each of the other seasons);

D is the dose conversion coefficient, set at 9 nSv·(Bq·m⁻³·h)⁻¹ according to United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) recommendations [25].

The annual effective dose ($E_y$) was calculated as the sum of the four seasonal doses:

$$E_y=E_{wint}+E_{spr}+E_{sum}+E_{aut}$$

(2)

This approach allows for a season-specific assessment of inhalation dose and accounts for variations in radon concentrations and time spent indoors throughout the year.

2.5. Statistical Analysis Plan

Data analysis was conducted in RStudio (version 2024.12.1.563) using R (version 4.3.2, released on 31 October 2023). Descriptive statistics were computed. Median radon progeny concentrations in both locations were compared across seasons using the Friedman test (non-parametric, paired design). Post hoc Wilcoxon tests with Bonferroni correction were applied for significant results. Boxplots were generated to visualize the distribution of EEVA values. The top and bottom edges of each box represent the 25th and 75th percentiles, respectively, while the whiskers extend to the 90% confidence interval. The horizontal line within each box indicates the median, and individual dots denote outliers. A linear regression model with Pearson correlation was constructed to examine the association between H*(10) and EEVA, while also accounting for the presence of a cellar in the building where measurements were taken. The relationship is presented using a scatterplot. Across all subplots, shaded regions represent 95% confidence intervals around the regression lines. The regression equations and R² values are reported within each plot to indicate model fit. Annual and seasonal effective doses were converted to units of mSv/year and mSv/season (i.e., year divided by four), respectively.

3. Results

3.1. Gamma Radiation

Table 1. Measurements of H*(10) in Aqsu and Zavodskoy settlements at 1 m above the soil surface.

Settlement	Distance from the Uranium Tailing Dump, km	Mean ± SD, µSv/h	Range (Min–Max), µSv/h
Aqsu	3.5	0.16 ± 0.07	0.08–0.55
Zavodskoy	3.2	0.14 ± 0.04	0.07–0.22

presents the H*(10) measured at 1 m above the ground surface in the two settlements located near a uranium tailings site. In Aqsu, located approximately 3.5 km from the tailing dump site, the mean value of H*(10) was 0.16 µSv/h with a standard deviation of 0.07 µSv/h, ranging from 0.08 to 0.55 µSv/h. In Zavodskoy, the mean value of H*(10) was 0.14 µSv/h with a standard deviation of 0.04µSv/h, with observed values ranging from 0.07 to 0.22 µSv/h.

3.2. Seasonal Distribution of EEVA of Radon Progeny

Table 2. Seasonal distribution of EEVA radon ranges by building types in Aqsu and Zavodskoy.

Radon Progeny Concentration Range	The Number of Dwellings (Fraction in %)
	Winter		Spring		Summer		Autumn
No cellar
	Aqsu	Zavodskoy	Aqsu	Zavodskoy	Aqsu	Zavodskoy	Aqsu	Zavodskoy
below 120 Bq/m3	14 (47%)	14 (82%)	16 (53%)	14 (82%)	21 (70%)	17 (100%)	20 (67%)	15 (88%)
120 and above Bq/m3	16 (53%)	3 (18%)	14 (47%)	3 (18%)	9 (30%)	0 (0%)	10 (33%)	2 (12%)
With cellar
	Aqsu		Aqsu		Aqsu		Aqsu
Below 120 Bq/m3 with cellar	6 (40%)		6 (40%)		12 (80%)		7 (47%)
120 and above Bq/m3 with cellar	9 (60%)		9 (60%)		3 (20%)		8 (53%)

summarizes the seasonal distribution of dwellings with EEVA radon progeny concentrations below and above the permissible level of 120 Bq/m³, stratified by building types (with or without cellar) and locations (Aqsu and Zavodskoy). In Aqsu, proportion of dwellings with no cellars exceeding the threshold decreased from 53% in winter to 30% in summer. Zavodskoy exhibited a similar trend, decreasing from 18% in winter to 0% in summer.

For dwellings with cellars (observed only in Aqsu), exceedances occurred year-round, peaking in winter and spring (60% each), followed by autumn (53%), with summer showing the lowest rate (20%).

Table 3. Median radon concentrations in Aqsu and Zavodskoy.

Seasons	Aqsu		Zavodskoy
	Median IQR, Bq/m3	Min–Max, Bq/m3	Median IQR, Bq/m3	Min–Max, Bq/m3
No cellar
Winter	130 (61.5–422)	23–2397	222 (87–270)	13–231
Spring	96 (48–258)	20–817	140 (40.5–349)	5–255
Summer	63.5 (29.5–184)	11–526	22 (12–96)	2–87
Autumn	86.5 (48.2–196)	19–763	155 (76.5–222)	8–210
With Cellar
Winter	107 (47.6–328)	54–4837	-	-
Spring	22 (6.75–32.25)	13–5327	-	-
Summer	7.5 (5–12.2)	2–341	-	-
Autumn	12.5 (10.75–27.75)	3–1817	-	-

Abbreviations: IQR—interquartile range; Maximum—maximum; Min—minimum.

presents the median EEVA values, interquartile ranges (IQR), and minimum–maximum levels across different building types in Aqsu and Zavodskoy during winter, spring, summer, and autumn. Significant seasonal variations in radon progeny concentrations were observed, with highest winter levels (Aqsu: 130 Bq/m³; Zavodskoy: 222 Bq/m³) and lowest summer levels. Dwellings with cellars in Aqsu exhibited extreme maximum values (up to 5327 Bq/m³ in spring. Complete data including IQRs and ranges are provided. Figure 2 illustrates the visual distribution of seasonal EEVA values described in both locations.

Table 4. Pairwise comparisons of seasonal radon progeny concentrations.

	Cellar Type	Comparison	Autumn p-Values	Spring p-Values	Summer p-Values
Aqsu	No cellar	Spring	1	-	-
		SUMMER	0.037	0.0045	-
		Winter	0.0004	0.012	0.00006
	With cellar	Spring	1	-	-
		Summer	0.002	0.01	-
		Winter	0.332	1	0.01
Zavodskoy	No cellar	Spring	1	-	-
		Summer	0.187	0.349	-
		Winter	0.039	0.019	0.015

presents the pairwise comparisons of seasonal radon progeny concentrations. In Aqsu, dwellings without cellars showed significantly higher EEVA values in winter compared to all other seasons (p < 0.001 vs. autumn/summer; p = 0.012 vs. spring), while summer levels were notably lower than spring (p = 0.0045) and autumn (p = 0.037). For Aqsu dwellings with cellars, summer EEVA values were markedly reduced relative to autumn (p = 0.002) and spring (p = 0.01), with winter differing only from summer (p = 0.016) and no significant variations among other seasons. In Zavodskoy, winter concentrations were elevated compared to autumn (p = 0.039), spring (p = 0.019), and summer (p = 0.015), with no other seasonal differences reaching statistical significance.

Figure 3 displays scatterplots with regression lines, illustrating the Pearson correlation between H*(10) and EEVA of radon across seasons, stratified by building type (with cellar vs. no cellar). In Panel A (Aqsu), a positive correlation between H(10) and EEVA is observed during winter for both building types. The fitted regression equations reveal a steeper slope for dwellings with cellars (blue) compared to those without (red). The R² values range from 0.35 to 0.59 across seasons, with the strongest association occurring in spring (R² = 0.59), followed by winter and autumn (R² = 0.5 each), and summer (R² = 0.35). In Panel B (Zavodskoy), the linear relationships between H(10) and EEVA are generally weaker. For winter, the regression model for buildings without cellars yields an R² of 0.16. In spring, summer, and autumn, the corresponding R² values are 0.14, 0.11, and 0.30, respectively. Influence diagnostics for all correlation models, including Cook’s distance are presented in Supplemental Figure S1.

3.3. Radiation Exposure from Radon Concentration

Table 5. Seasonal and Annual Effective Dose in Aqsu and Zavodskoy.

Seasons	${\mathit{E}}_{\mathit{S}}\left(\frac{\mathbf{m}\mathbf{S}\mathbf{v}}{\mathbf{s}\mathbf{e}\mathbf{a}\mathbf{s}\mathbf{o}\mathbf{n}}\right) \mathbf{a}\mathbf{n}\mathbf{d} {\mathit{E}}_{\mathit{y}}\left(\frac{\mathbf{m}\mathbf{S}\mathbf{v}}{\mathbf{y}\mathbf{e}\mathbf{a}\mathbf{r}}\right)$
	Aqsu		Zavodskoy
	No Cellar	With Cellar
winter	5.03	7.81	1.07
spring	3.48	8.82	0.89
summer	1.80	1.22	0.32
autumn	3.55	5.66	0.77
annual	13.86	23.51	3.05

summarizes the estimated seasonal and annual effective dose across winter, spring, summer, autumn, and annual exposure periods in the settlements of Aqsu and Zavodskoy. The seasonal effective dose in Aqsu in dwellings with no cellar ranged from 1.80 mSv in summer to 5.03 mSv in winter. The annual effective dose was 13.86 mSv. The seasonal effective dose in Aqsu in dwellings with a cellar ranged from 1.22 mSv in summer to 8.82 mSv in spring. The annual effective dose was 23.51 mSv. In Zavodskoy, seasonal effective dose ranged from 0.32 mSv in summer to 1.07 mSv in winter. The annual effective dose was 3.05.

4. Discussion

This study provides a comprehensive assessment of indoor radon exposure and environmental indoor gamma radiation levels in two rural settlements of the Akmola Region of Kazakhstan, namely, Aqsu (located closest to the dump site and house for two open-pit mines) and Zavodskoy, which served as a comparative control area. Our findings confirm the initial hypothesis, indicating substantially higher seasonal and annual EEVA concentrations in Aqsu compared to Zavodskoy, with the highest levels observed in dwelling homes with cellars. Statistically significant differences in median EEVA values across seasons were found for all building types. The highest median EEVA concentrations were observed during winter in both types of dwellings in Aqsu. Furthermore, a moderate correlation was found between H*(10) and EEVA in Aqsu across all seasons, and the estimated annual effective dose values were higher in Aqsu. These results highlight spatial, structural, and seasonal heterogeneity in radiation exposure near legacy uranium mining areas.

The findings of this study are consistent with and extend previous research conducted in Kazakhstan. Earlier studies evaluating gamma radiation exposure in this region reported elevated radiation levels ranging between 0.14 and 0.95 µSv/h [26] and 0.16–0.4 µSv/h [13]. Our measured mean exposure level of 0.16 ± 0.07 µSv/h demonstrates consistency with these established baselines while providing updated spatial and seasonal data. Notably, these results mirror observations from uranium-rich and mining regions globally, where residential proximity to open-pit mines consistently correlates with elevated environmental radiation [27,28,29,30,31]. The observed increased radiation exposure in our study could potentially be attributed to the geographical closeness of these residential areas to disused mining sites and uranium waste disposal zones.

Furthermore, previous radon exposure assessments in Aqsu documented elevated indoor radon levels in residential and school settings, especially in proximity to uranium processing facilities [12,13,14,32]. A study by Tokonami et al. revealed that radon concentrations exceeded 300 Bq/m³ in 70% of the residential dwellings investigated, with concentrations surpassing 1000 Bq/m³ in 5% of these buildings [12]. Similarly, research conducted in schools and kindergartens in Aqsu reported a mean radon concentration of 550 ± 200 Bq/m³ on the first floors, indicating a significant risk of elevated annual effective doses for children [14]. This study builds on that evidence by incorporating seasonal measurements. Moreover, the results affirm earlier observations that structural characteristics, such as the presence of a cellar, substantially influence indoor radon accumulation [33,34]. While the presence of cellars was associated with higher EEVA values in residential spaces during all non-summer seasons, additional factors also contribute to the observed seasonal variations. Studies show that indoor radon concentrations are largely determined by outdoor temperature, with a consistent inverse correlation [35]. Moreover, the widespread practice of sealing windows during the cold period reduces ventilation, thereby increasing indoor accumulation. At the same time, frozen ground surrounding houses restricts radon release to the outside environment [36], channeling soil gas into the softer soil beneath buildings and ultimately indoors. Together, these mechanisms explain the consistent pattern of elevated EEVA in Aqsu across all seasons and confirm continued environmental risks in uranium-adjacent communities [12,14,32].

Compared to other regional assessments of radiation burden, where earlier work focused on annual exposure estimates [11,12,14], our seasonal estimates reveal that winter and spring present the most significant period of internal exposure. Radon progeny concentrations significantly differ across seasons [37,38]. Effective dose was highest during the winter in the dwelling with no cellar at 5.03 mSv/season, and in the spring in the dwellings with a cellar at 8.82 mSv/season in Aqsu. This supports findings from similar contexts globally, where radon exposure intensifies during colder months due to reduced ventilation [39,40,41].

The presence of markedly elevated EEVA readings observed in dwellings with a cellar during winter—and even more pronounced during spring—indicate that radon accumulates in specific micro-environments. These outliers likely reflect a combination of factors: increased stack-effect pressure differentials in colder months, limited ventilation of cellars, moisture-related transport mechanisms and incomplete sealing of foundation cracks or utility penetrations [42,43]. Combustion heating also increases indoor aerosol concentrations, extending the airborne residence time of radon progeny because particle-attached progeny deposit more slowly than unattached ions [44,45]. Together, they suggest that current interventions are not uniformly implemented or sufficiently robust for dwellings with cellars.

These findings carry significant health implications that expand upon existing literature. In addition to the well-established association between chronic radon exposure and lung cancer, regional epidemiological studies indicate a higher prevalence of circulatory, respiratory, and musculoskeletal disorders among residents of this area compared to populations living farther from uranium tailings sites [46]. In Aqsu, where annual effective dose reaches 13.86 mSv/year in dwellings without cellars and 23.51 mSv/year in those with cellars, ongoing public health interventions are warranted—including radon-resistant construction standards, systematic monitoring, and prioritized remediation. Such measures prove particularly crucial for dwellings with cellars, which demonstrated consistently elevated EEVA values. Several low-cost mitigation strategies could be implemented by homeowners with proper education and guidance [47]. Notably, regional mitigation programs appear to be effective, as reflected in the fifteen-year downward trend in overall cancer incidence in the area [20]. However, this trend requires further investigation to establish causality. The exposure disparity between Aqsu and Zavodskoy, despite their similar distances from the tailings site, underscores how proximity to open-pit mines, seasonality and building characteristics may influence radiation risk.

This study has several limitations. First, although we conducted repeated measurements across seasons, the reliance on short-term repetitive monitoring may not fully capture temporal variations in radon concentrations. Future research should prioritize long-term passive radon detection methods to improve temporal resolution and accuracy [48]. Second, Cook’s-distance influence diagnostics (Supplemental Figure S1) identified several observations that exerted a disproportionate effect on the regression coefficients and R² values; while we retained these data to reflect real-world variability, the presence of influential outliers means that the reported associations should be interpreted with caution. Third, while our regression models accounted for a moderate proportion of the variance in H*(10), the overall correlation coefficients (R²) were relatively low, which further underscores the exploratory nature of these analyses. Finally, several unmeasured confounding factors—such as soil permeability, ventilation rates, and meteorological conditions—could influence radon exhalation and subsequently EEVA levels. These variables should be systematically evaluated in future studies to enhance predictive models and exposure assessments.

5. Conclusions

In summary, this study demonstrates that residents of Aqsu—a uranium-adjacent village in Kazakhstan—are exposed to significantly higher indoor radon levels and seasonal/annual effective doses compared to those in the nearby settlement of Zavodskoy. In Aqsu, measured gamma dose rates exceeded intervention levels at multiple locations, with H(10) values ranging from 0.08 to 0.55 µSv/h (mean: 0.16 ± 0.07 µSv/h). The highest EEVA of radon concentrations occurred in winter, with median values of 130 Bq/m³ (61.5–422 Bq/m³) in dwellings without a cellar and 107 Bq/m³ (47.6–328 Bq/m³) in dwellings with a cellar. Seasonal effective doses peaked in winter for dwellings without a cellar (5.03 mSv/season) and in spring for dwellings with a cellar (8.82 mSv/season). The total annual effective dose was substantially higher in Aqsu than in Zavodskoy, reaching 13.86 mSv/year in dwellings without a cellar and 23.51 mSv/year in dwellings with a cellar, compared to just 3.05 mSv/year in Zavodskoy.

Critically, while ongoing mitigation efforts have reduced radon levels in some households, certain dwellings with a cellar in Aqsu still exhibit extreme EEVA values, suggesting localized contamination or structural deficiencies that require targeted remediation. These findings highlight the elevated radiological risk faced by Aqsu residents, particularly in dwellings with a cellar during winter and spring seasons and underscore the need for additional mitigation measures to reduce long-term exposure. Overall, the results confirm the persistent radiological hazard in Aqsu and emphasize the need for optimized interventions, especially in dwellings where current measures fail to mitigate exceptionally high radon exhalation.