The influence of temperature, humidity, and simulated sunlight on the infectivity of SARS-CoV-2 in aerosols

Abstract

Recent evidence suggests that respiratory aerosols may play a role in the spread of SARS-CoV-2 during the ongoing COVID-19 pandemic. Our laboratory has previously demonstrated that simulated sunlight inactivated SARS-CoV-2 in aerosols and on surfaces. In the present study, we extend these findings to include the persistence of SARS-CoV-2 in aerosols across a range of temperature, humidity, and simulated sunlight levels using an environmentally controlled rotating drum aerosol chamber. The results demonstrate that temperature, simulated sunlight, and humidity are all significant factors influencing the persistence of infectious SARS-CoV-2 in aerosols, but that simulated sunlight and temperature have a greater influence on decay than humidity across the range of conditions tested. The time needed for a 90% decrease in infectious virus ranged from 4.8 min at 40 °C, 20% relative humidity, and high intensity simulated sunlight representative of noon on a clear day on the summer solstice at 4°N latitude, to greater than two hours under conditions representative of those expected indoors or at night. These results suggest that the persistence of infectious SARS-CoV-2 in naturally occurring aerosols may be affected by environmental conditions, and that aerosolized virus could remain infectious for extended periods of time under some environmental conditions. The present study provides a comprehensive dataset on the influence of environmental parameters on the survival of SARS-CoV-2 in aerosols that can be utilized, along with data on viral shedding from infected individuals and the inhalational infectious dose, to inform future modeling and risk assessment efforts.

Figure 1.

Schematic of rotating drum aerosol chamber. A 16-L environmentally controlled rotating drum aerosol chamber was used to assess the effect of environmental parameters on the decay rate of SARS-CoV-2. The temperature of the air inside the chamber was controlled by adjusting the temperature of polypropylene glycol circulating through the walls of the chamber. The humidity inside the chamber was controlled by varying the mix of dry and humid air used to supply the chamber during aerosol filling and sampling. Simulated sunlight was introduced into the chamber through a quartz glass window located on one end of the chamber, as indicated by the arrows in the figure. Aerosols with a target MMAD of 2 μm were generated with an air assist nozzle into a humidity controlled stainless steel plenum that was exhausted through a HEPA filter. A portion of this flow was diverted into the rotating drum for 20–30 s during filling to achieve a quantifiable concentration of infectious virus. Multiple samples were collected over the duration of the test to estimate infectious viral concentrations and total aerosol mass concentrations over time from sampling ports located on a non-rotating band in the wall of the rotating drum.

Figure 2.

Comparison of simulated sunlight UV spectrum to NCAR TUV model spectrum. The spectrum generated by the solar simulator (black) was similar to that predicted by NCAR’s TUV radiation model for 40°N latitude on 21 June at noon at sea level under cloudless conditions (gray). Integrated UVB irradiances were 1.91 W/m² and 1.84 W/m² for the solar simulator and TUV radiation model, respectively. No irradiance above the background level of the spectroradiometer measured in darkness (dashed line) was present in the spectrum generated by the solar simulator below approximately 295 nm.

Figure 3.

Representative decay plots. The slopes of the best fit lines for time-series log₁₀-transformed infectious and mass aerosol concentrations from each test were used as estimates of the total and physical decays, respectively. Total decay rates calculated from filter samples were normalized for physical losses by subtracting the physical decay rate calculated from APS samples to estimate the decay rate for viral infectivity. Decay plots from tests at (a) 30 °C, 1.9 W/m² integrated UVB, and 20% relative humidity and (b) 10 °C, 0.0 W/m² integrated UVB, and 70% relative humidity are shown. The decay rate for viral infectivity is greater for (a) 30 °C, 1.9 W/m² integrated UVB, and 20% relative humidity relative to (b) 10 °C, 0.0 W/m² integrated UVB, and 70% relative humidity, as can be seen from the larger difference between the slopes of viral (black) and mass (gray) aerosol concentrations.

Figure 4.

Decay constants for viral infectivity as a function of temperature and humidity without simulated sunlight. Both relative and absolute humidity levels are shown on the x-axis, along with temperature. Decay constants were near zero for most conditions, but increased at both 40 °C/20% relative humidity and 30 °C/70% relative humidity. Lines indicate arithmetic mean ± one standard deviation.

Figure 5.

Decay constants for viral infectivity as a function of temperature and humidity at 0.9 W/m² UVB irradiance. Both relative and absolute humidity levels are shown on the x-axis, along with temperature. Lines indicate arithmetic mean ± one standard deviation. Data at 20 °C and 45% relative humidity are from Schuit, Ratnesar-Shumate, et al. (2020).

Figure 6.

Decay constants for viral infectivity as a function of temperature and humidity at 1.9 W/m² UVB irradiance. Both relative and absolute humidity levels are shown on the x-axis, along with temperature. Lines indicate arithmetic mean ± one standard deviation. Data at 20 °C and 20% relative humidity are from Schuit, Ratnesar-Shumate, et al. (2020). The variability associated with measurement of k_infectivity was significantly increased relative to measurements at lower simulated sunlight levels, especially at higher relative humidities.