Yes, It's in the Air

University of Colorado-Boulder chemistry professor Jose-Luis Jimenez, who studies aerosols and air pollution, recently summarized his thoughts so far on COVID-19 transmission on Twitter. It's exciting (and sometimes a bit confusing) to watch scientists talking to each other about all of this in real time on Twitter, but it beats hearing from the current occupant:

My current estimates on modes of transmission and mitigation measures. From work on this and talking to a lot of scientists, my thinking is evolving. I thought it'd be useful to present my updated estimates, with hope of generating debate, finding new info to further polish this.

I have talked to many scientists, who have their own estimates, but prefer to not make them public. My estimates are not an outlier, many people are in same ballpark, not everyone of course.

Transmission varies with setting, preventive measures. Attempt at averages for non- or low-symptomatics.

These are the things that we know need to avoid, which I later relate to my estimates:

• Indoors
• Crowded
• Low ventilation
• Long duration
• No masks
• Talking / shouting / singing
• Close proximity

 My estimate for non/low- symptom “talking” transmission:

~75% aerosols: 1/2 (37%) at close proximity, 1/2 (37%) at room-level, 1% long range
~20% surfaces / fomite / direct contact between people
~5% ballistic droplets, mostly accidental sneezing and cough. (generate aerosols too)

These correspond with the amount of available evidence. Aerosols have a ton of evidence, and are the explanation for close proximity transmission for non-symptomatic people. Fomites have some evidence, but are not major. Droplets are for sure more important for transmission from symptomatic people who are coughing

Now let’s examine the things to **avoid** and what modes of transmission they eliminate, assuming my estimates are approximately correct:

• Close proximity: this is directly one of the modes
• Not being outdoors: no room-level and long range, less close proximity (wind, UV), less fomites (UV)
• Not Crowded + long duration: reduced close proximity, room-level, fomites, and droplets
• Low ventilation: reduced room-level and long range
• No masks: reduced close proximity, room-level, long-range, fomites, and droplets
• Talking / shouting / singing: reduced close proximity, room-level, long range, droplets (also less generation of fomites)

So that’s my take. When all is said and done and we know the answer a lot better, I'd be extremely surprised if this doesn't look like a good guess based on what we knew at the time.

The notable thing there is the high percentage of transmission Jimenez sees coming through aerosols and particularly through what he calls "room-level" — not from close proximity with another person. It's not measles- or Legionella-levels of airborne transmission, where the disease travels through the air across great distance and even through air-handling systems, but it's a lot more than we in the general public are generally thinking or the WHO or CDC are communicating.

Another scientist I've recently been following on Twitter, Harvard epidemiologist Eric Feigl-Ding, has been calling on these public health bodies to be clearer about this messaging, and that improving ventilation (in addition to masks) is key. 

Tell us some truths about this.