Radon and Climate

When indoor air is warmer than outdoor air, the lighter indoor column creates vertical pressure gradients. The pressure difference is ΔP ≈ 0.04 Pa per meter per kelvin. For a basement 3 meters below the neutral plane: at 20°C indoor vs 0°C outdoor (ΔT = 20 K), depressurization is about 2.4 Pa. At −20°C outdoor (ΔT = 40 K), it reaches about 4.8 Pa.

Arvela's model of 3,074 Finnish dwellings found an average convective radon entry rate of 50 Bq·m⁻³·h⁻¹ at a mean indoor–outdoor ΔT of 17 K, plus diffusive entry of 7 Bq·m⁻³·h⁻¹. Korean national data showed monthly geometric-mean indoor radon correlated with outdoor temperature at r = −0.88 for detached houses.

The stack effect both draws soil gas in and drives ventilation. Arvela notes "the same pressure difference is also the driving force for airflow from soil to indoor spaces," so winter ventilation partly offsets winter entry. Houses whose radon comes mostly from building materials by diffusion show little seasonality.

When atmospheric pressure falls, soil gas (which equilibrates more slowly) expands and flows upward along the pressure gradient, carrying radon-rich air from depth. When pressure rises, radon-free atmospheric air is forced into the ground, diluting near-surface soil radon.

Clements & Wilkening (1974): "pressure changes of 1–2% associated with the passage of frontal systems produce changes in the ²²²Rn flux from 20 to 60%, depending upon the rate of change of pressure and its duration." At sea level, 1–2% equals about 10–20 hPa (1,000–2,000 Pa). The rate of pressure change matters more than absolute value.

Typical synoptic variation is 10–30 hPa over 1–3 days as weather systems pass. Xie et al. (2015) found indoor radon correlated most strongly with the indoor–outdoor pressure difference (R = 0.67), more than with absolute pressure. Hahn et al. (2023) tested 66 rural Kentucky homes and found that "with higher atmospheric barometric pressure during testing, observed indoor radon values were lower."

Moisture affects radon through multiple mechanisms with opposite signs. Between 0 and 10% volumetric water content, the emanation factor roughly doubles. For sand, emanation jumped from 0.13 to 0.28 at only 0.5% pore moisture. At 1 m depth, modeled soil-gas radon rises from 30 to 62 kBq/m³ as water saturation rises from 0.3 to 0.7, mainly due to partitioning and reduced diffusive loss.

However, moisture sharply reduces soil permeability and diffusion. Rogers & Nielson's relation: normalized air permeability K_n = exp(−12m⁴), where m is saturation fraction. At 50% saturation, permeability drops ~53%; at 70%, it drops ~94%. Bulk diffusion coefficient falls from 1.2×10⁻⁶ to 2×10⁻⁹ m²/s as saturation rises from 20 to 90% — a 600-fold reduction.

The net effect depends on geometry. If only the upper 0–30 cm is saturated, "air flows towards the footings may come from greater distances, which increases the radon concentration of sub-foundation air" (Arvela et al.). Wet surface soil can cap the ground and redirect soil gas toward the building footprint.

Stoulos & Ioannidou (2023), northern Greece: "Clear radon peaks appeared due to 50–100 mm rainfall during a few days," appearing a few days after heavy rains. Mose et al. (1991), northern Virginia: the wetter summer of 1988 produced about 30% higher regional indoor radon than 1987.

Schumann et al. (1992) found soil-gas radon varies "by as much as an order of magnitude seasonally and as much as severalfold in response to changes in weather." Precipitation and barometric pressure were the main factors. A single mid-October rain on dry soil "had virtually no effect" — antecedent soil moisture matters more than storm size.

Measurable soil-gas responses appear with multi-day totals of about 50–100 mm, or with rain on soils that are already moist. Light rain of 10–15 mm mainly suppresses surface exhalation. Some studies (Sundal et al. 2008, Norwegian glacial sediments) found no apparent precipitation effect, so responses are site-dependent.

The popular model holds that frozen or snow-covered ground caps the soil and diverts radon to the building, whose warm footprint becomes the easiest exit. USGS supports a capping role: "A capping effect caused by frozen or unfrozen soil moisture is a primary mechanism for preventing radon loss to the atmosphere."

However, site-specific observations vary. At Tomakomai (porous volcanic ash), winter soil radon averaged 570 Bq/m³ vs 350 in summer (+63%). At Sapporo (alluvium), soil radon fell from 5.0 to 2.6 kBq/m³ (−48%) under persistent snow. In Finland, dry autumn followed by 5 months of frozen surface kept coarse soil dry, and modeled soil-gas radon was "only ∼50% of its normal long-term average."

Snow reliably reduces escape through open ground, but whether that increases building entry depends on permeability beneath the footprint and whether soil below the frost is moist or dry. The stack effect and closed windows are better-documented winter causes than frozen-ground capping.

Wind produces positive pressure on windward faces and negative on leeward sides and roofs. It affects radon through three mechanisms: depressurizing or pressurizing the building interior, increasing infiltration and dilution, and creating ground-surface pressure fields that move soil gas.

Riley et al. (1996) combined wind-tunnel ground-pressure fields with 3D soil-gas modeling and found a wind "flushing" effect on soil radon around the house, with "consequent sharp decrease in radon entry rates." Fluctuating wind direction dominated the impact; fluctuating speed had little effect. On Finnish eskers, highest indoor radon occurred when wind blew toward the slope side.

The net correlation is usually negative. Wind speed correlated at R = −0.25 (Xie et al.) and r = −0.41 (Spasić & Gulan) with indoor radon, indicating wind generally reduces levels through dilution and soil flushing.

Large surveys show mean winter/summer ratios of 1.23 (Italy, 4,742 houses) and 1.28 (Finland, 3,074 dwellings). UK, Irish, and French seasonal factors imply peak-to-trough ratios of about 1.9. In soil-gas-dominated Finnish houses, the summer minimum is about 50% of the winter maximum (ratio ~2).

Arvela (2004 review): in houses where soil gas dominates, "a summer minimum is typically 50% of the winter maximum." Miles et al. (2012) followed 91 UK homes over two years: two-thirds showed the winter-high pattern, most of the rest showed little seasonal variation, and a few were reversed.

Health Canada early results (377 homes, 5 regions): summer 3-month tests were −8% to −31% vs annual; winter tests were +2% to +17% (coastal BC showed −19%). The "2–3× higher in winter" claim applies only to high-radon, soil-gas-dominated houses; population averages are closer to 1.2–1.5×.