UV and SARS-CoV-2 (Covid-19)

The paper behind the infamous White House briefing ...

A couple of weeks ago I reported on the now infamous White house briefing of April 23 where a report emerged that the covid-19 virus is inactivated by a few minutes exposure to sunlight (though that study was largely submerged below Trump’s gaffe about ingesting bleach to drive out the virus!)

I promised then to keep you posted if the paper behind that UV study reported ever saw the light of day.

Well, it looks as if it just did. At least a paper very much like it, submitted by a team from the US Department of Homeland Security, has just been accepted for publication in the Journal for Infectious Diseases. I’ve just read a preprint, and asked the lead author for confirmation on that. I’ve yet to hear back from her (* see addendum added 2 June 2020), but you can read the whole paper for yourself here. It concludes that ninety percent of the virus was inactivated in 7 to 14 minutes when exposed to simulated sunlight representative of the summer solstice at 40N latitude at sea level on a clear day. About the same as stated at the White House briefing, but much shorter than the inactivation times we calculated for 45S (where the summer peak UV is similar to that at 40N).

They didn’t use real sunlight in the study, but instead used radiation from a solar simulator. I wondered about a couple of telling sentences in their paper: “A small amount of irradiance above background was also present in the 250-280nm range of spectrum. The integrated amount was constant across the different UVA and UVB irradiance levels utilized and averaged 3.2x10-3 ± 7.5x10-5 W/m2”.

The maximum UV-B they considered was 1.6 W/m2, so the contribution from these shorter UV-C wavelengths that aren’t present in sunlight reaching the Earth’s surface was approximately 0.2 percent of that. That 0.2 percent may not seem much, but you need to remember that the damaging effect on RNA increases rapidly towards these shorter UV-C wavelengths. And that changes things - a lot.

The UV-B wavelengths that matter in sunlight are those near 305 nm. At longer wavelengths the action spectrum for damage becomes too small, and at shorter wavelengths the solar irradiance becomes too small. As my earlier posting showed, the damaging effect at UV-C wavelengths may be anything between 1 and 3 orders of magnitude greater than at 305 nm.

Unfortunately, the action spectrum for RNA damage is not known over the full wavelength range from 250 to 315 nm. However, the action spectrum for DNA is, and some studies show that over shorter wavelength intervals the shapes of the two action spectra match. Assuming the stray light reported in the paper is equally distributed in wavelength, the contribution to viral damage from the UV-C component from the solar simulator is about 50 percent of that from the UV-B. So the inferred time for inactivation in real summer sunlight - which contains no UV-C - would increase (from their reported 7 to 14 minutes) to about 10 to 20 minutes.

The authors state - see the second sentence in their quote in italics above - that the stray light remains the same for all three UV-B irradiance settings used. That seems surprising to me, but if it’s true, the stray light problem becomes a much larger percentage error in each of the other two cases where UV-B irradiances are lower: 0.7 and 0.3 Wm-2 respectively. For the lowest setting (approximating winter sunlight at 40N), the corrected inactivation times would be measured in hours.

So, interesting though the study is, the quantitative result needs be taken with a grain of salt (not bleach) until the inactivation times can be verified using real sunlight, rather than just simulated sunlight. The conclusion from both this study and my own calculations is that, for accurate estimations using artificial light sources, more work is needed to characterise the action spectrum for RNA for wavelengths at least encompassing the range 250 to 315 nm.

I’d like to add that if it takes 15 minutes to inactivate 90 percent of the virus, that means that 10 percent will have survived. It also means that there would still be 1 percent left after another 15 minutes, etc. Normal sunlight won’t ever be the universal panacea.

* A couple of days after posting this, I was encouraged to receive a courteous reply from the lead author of the paper, which gave me added confidence in the validity of the work. Prior to that I had been skeptical because the first airing of the work was in one of Donald Trump’s White House briefings, which we’ve learnt to distrust (I had even wondered if the group really did exist). The author mentioned that they’ve found the source of the stray light in their experimental setup and they are currently planning followup studies to better evaluate their effect.

Subscribe below to continue receiving occasional postings about UV and climate (and occasional digressions) from UV &You.