While the novel coronavirus ran its first lap around the Earth in early 2020, making history, researchers worked hurriedly to understand the nature of the disease and the best means to restrict its spread in the face of uncertainty, with world governments in shock and the global economy in freefall.

Other human coronaviruses exhibit some seasonality in their rates of spread. Courtesy of Tamma Carleton

Researchers from UCSB, Tamma Carleton and Kyle C. Meng, along with collaborators at Harvard and ENS Paris-Saclay, scrambled to collect regional data and run analyses to better understand the role that climate may have on the spread of COVID-19.

Their study, published in the Proceedings of the National Academy of Sciences of the United States of America, uncovers one dimension of this: climatic variation in the intensity of UV  radiation, or sunlight, and corresponding effects on COVID-19 transmission.

Having previously studied the effects of climate change on social and economic outcomes, such as mental health, civil conflict and agricultural output, Carleton and her collaborators made use of a toolkit they developed prior to answer such questions as they pertained to COVID-19 transmission rates across various regions during the beginning of the pandemic.

Close relatives of the SARS-CoV-2 virus, such as other human coronaviruses, exhibit a marked seasonality in their rates of spread in temperate regions, with a global peak occurring during the winter season.

“We’re like, ‘Well, everyone really wants to know how seasonal this is, because we’re all trying to project into the summer and into the winter, what this might look like,” Carleton said. “And, you know, we’ve built up this toolkit to link changes in the weather with different social and economic outcomes. Why don’t we bring our tool kit to this problem?’”

This toolkit, reliant on a supply of observational data to link weather and social and economic outcomes, previously looked at climate variation across different time spans within the same region to tease out how weather affects a population of people.

For instance, Carleton had previously looked at data within single locations in sub-Saharan Africa during drier or hotter growing seasons relative to the average and used this to ascertain how a shift away from normal conditions in turn impacts mortality rates, conflict or productivity, among a range of other issues. Understanding these impacts is particularly crucial as climate change accelerates.

With this particular research, Carleton and her collaborators pivoted, using data to isolate and statistically identify the role that individual climate conditions may have in contributing to the spread of COVID-19.

She and her collaborators began a concerted search for information from various countries’ statistical agencies across the world. The researchers collected data across over 3000 different geospatial units, or locations; this data contained a treasure trove of information: real-time weather estimates, temperature, humidity, precipitation, ultraviolet radiation and most crucially, COVID-19 cases.

“As you might imagine, in the United States, weather conditions in Santa Barbara look very different than they do in Minneapolis, Minnesota. And so, to ensure that we could capture differences within a country as opposed to just across countries, we needed a really big data collection effort to collect as much as we could within countries,” Carleton said.

“Essentially you’re estimating using a big statistical model linking those changes and COVID-19 cases to change in weather conditions.”

The researchers discovered a significant connection between negative correlation between UV radiation and COVID-19 growth rates, though the strength of the relationship was not quite as strong as others have proven to be, such as social distancing policies. Other measures the researchers looked at, like humidity and temperature, yielded no relationship.

“A sunnier day is going to lower transmission that can happen through multiple different channels and all those channels are going to be accounted for here. So, one is just that we know that UV breaks down the virus. But we also know that there’s all sorts of human behaviors that change. We all do different things on sunny or on hotter days than we do on other days,” Carleton said.

However, Carleton also stressed possible issues with the data they’re drawing from, which may be clouding important relationships between COVID-19 spread and with other weather conditions.

“The world is scrambling to improve testing and reporting. Meanwhile, we are stuck running analyses on the data that we have. And in many cases, the data that we have here is imperfect, although we go to really great lengths to try to adjust them for really obvious inconsistencies,” Carleton said.

Because of this, Carleton speculates that temperature and humidity may still possibly have a significant effect on COVID-19 growth. Perhaps, as Carleton said, rather than being the only driver, UV is the strongest driver of those that the researchers measured.

“I think the other thing to be aware of this in this paper is that like I said, we really rushed to try to answer this question during the beginning of the pandemic so the data behind these estimates are early pandemic estimates,” Carleton said.

“And so to the extent that the disease changes over time — we’ve been hearing about new variants, for example, social distancing policies are different, which may influence the relationship between weather and COVID-19 — we should really be thinking about these estimates as early pandemic estimates.”


Sean Crommelin
Sean Crommelin is the Science and Tech Editor for the Daily Nexus. He can be reached at science@dailynexus.com