West Nile virus forecasts? It might be possible thanks to NSF-funded research that is finding correlations between weather patterns and virus outbreaks

NSF's Weather or Not!

West Nile virus forecasts? It might be possible thanks to NSF-funded research that is finding correlations between weather patterns and virus outbreaks

National Science Foundation


Forecasting West Nile?

Interviewer: Charlie Heck

Interviewees: Micha Hahn and Andy Monaghan

Charlie: Sprays, candles, zappers: what do these things all have in common? If it happens to be summer time, they just might be your only defense against backyard barbeque crashers—mosquitos! While these little buggers can definitely put a damper on outdoor activities, they can also transmit West Nile virus, and there is no vaccine or treatment. I’m Charlie Heck at the NSF, co-editor of Science360’s news service, co-host of the Super Science News Show and science geek in training. What if scientists could “forecast” West Nile outbreaks with weather patterns? New research, funded in part by the National Science Foundation, has identified a strong correlation between weather conditions and West Nile virus in the United States, raising the possibility of a forecasting system in the future. 

Micah: So because there's no vaccine or treatment for West Nile we're really predicting when and where we might expect to see higher risk so we can help target public health messages to high risk regions of the country. And additionally counties will have additional information to use for deciding about when, where and if they should do mosquito control.

Charlie: That’s Micah Hahn, lead author of a new paper on the subject. I spoke with Hahn and a co-author, Andy Monaghan. So, first let's talk a little bit about the West Nile virus. What is it and how do humans come in contact with it?

Micah: Well, the virus itself, West Nile virus, is maintained in bird populations. Over 300 species in the U.S. have tested positive for West Nile. And the virus is transmitted between birds by mosquitoes that are feeding. And sometimes when an infected mosquito bites a human they can transmit the virus between birds and humans.

Charlie: In the U.S., where do we tend to see the most frequent outbreaks?

Micah: West Nile virus was first identified in the U.S. in 1999. And then it spread across the U.S., and by 2004 we saw cases in all 48 states in the contiguous United States. And there have been annual seasonal outbreaks every year since that time. But most of the geographic scope and locations of where those outbreaks are from year to year varies substantially. That said, there are a few regions in the U.S. where we tend to see consistently high transmission including the Northern Great Plains, Southern California and in the Gulf states.

Charlie: Andy, can you talk a little bit about the difference between the Eastern United States versus Western in terms of the outbreaks, the weather patterns and the types of mosquitoes?

Andy: Sure, we see that there's one of the primary types of mosquitoes that transmits West Nile Virus is Culex. And there are many different species of Culex. But there is-- we think that the difference is between kind of needing to have dryer conditions that have higher incidence in the East and wetter in the West has to do with differences between the predominant types of Culex mosquito species. In the East, the Culex mosquitoes tend to be a little bit more urbanized. And they'll breed in things like: sewers, water catchments and whatnot. And so we think that what might be happening there is that in the already kind of relatively wet Eastern States that when we get rainfall events that they tend to maybe flush out the stagnant water that was fairly good breeding grounds for Culex that made lower incidence with higher rainfall. In the West, the Culex mosquito species that is predominant out there prefers grasslands and irrigated areas and whatnot. And we think that in the relatively dry Western states when we get rain it actually improves breeding conditions by creating those stagnant waters out in the grasslands and irrigated fields and whatnot that are conducive for mosquitoes and higher West Nile incidence.

Charlie: Okay, so let's talk about the research now. You're part of a team that has identified a correlation between weather conditions and the occurrence of outbreaks. Can you talk a little bit about those correlations?

Micah: Yeah, sure, so what we set out to do was see if we could find associations between annual temperature and precipitation anomalies and higher than normal West Nile incidents. So what we did was basically take 9 years of data from 2004 to 2012, and we looked at anomalies in annual temperature and precipitation at the county level and all over 3,000 counties in the United States and saw that it was correlated with higher than normal human West Nile incidents. So we found that, yes, there were associations.

Charlie: So let's talk a little bit more about that data. Tell us about all the things you analyzed, how you went about it. How do you study something like weather conditions associated to West Nile?

Micah: Sure, well, I guess maybe it would make sense to talk about sort of the overall goal of what we're trying to do. So, what we've been lucky to be able to do is to create a predictive model to forecast West Nile risk prior to the onset of the transmission season. So we thought we could do that using weather data. And the reason that we were interested in using meteorological information is because, like I said, the West Nile transmission cycle involves birds and mosquitoes and humans. And so each component of the transmission cycle is influenced by weather in some way, as you might imagine. So, for example, the bird populations in terms of fledgling survival when they're born in the winter-- or in the springtime or bird migration habits, things like mosquito biting rates increase, so with increasing temperature. And also the rate of mosquito-- the biology of the mosquito speeds up with increasing temperatures. So mosquitoes are both reproducing more frequently, and they're also biting more frequently when temperature is higher. And then for the actual virus, also replicates more quickly in higher temperatures. And, of course, as we all know human behavior is influenced by weather. So maybe our exposure to infected mosquitoes would vary depending on weather or, on a small scale, whether we have pools of water in our back yard where mosquitoes can breed and bird baths or in plant containers or on a larger scale whether farms are irrigating all depend on weather. And so the idea was that if we can sort of take some of these ecological reasons that weather influences West Nile and find a few significant predictors, we might be able to forecast what the season is going to look like early in the spring. So, you know, the first place to start is trying to think about which weather variables you want to use because there are a lot of them out there. You could use monthly data. You could seasonal data, annual data, temperature, precipitation, relative humidity, soil moisture, snow cover and lots of different things. And so what we wanted to do and the goal of this paper was first to sort of do a proof of concept for ourselves to decide if there are associations between temperature and precipitation and an association with West Nile incidents were strong enough to sort of justify moving forward with a predictive model. We started basically with the very basics just looking at temperature and precipitation and only looking at annual temperature and precipitation. And so that's sort of why we chose these variables for this analysis.

Andy: So this first step that's-- that we just recently published was just looking at the feasibility. And we found that the results were very promising in terms of being able to have associations between climatic variables and West Nile Virus not just annually but also looking at the seasons before the West Nile season. So, for example, looking at the winter and spring conditions proceeding a given summer, which is when we see West Nile Virus cases peak.

Charlie: If I ask you what/was there anything that really stood out to you? What was the most interesting kind of like "huh" that you found in your research?

Micah: That's a good question. I would probably say that difference in precipitation impacts in the different parts of the United States. It's sort of what we had prophesied. We think the ecology of West Nile is really different depending on where you are in the U.S. just because of the differences in the mosquito species that are responsible for transmitting the virus. So we did expect to see that, but we weren't quite sure how those differences would pan out. There's been lots of other research to show that-- that precipitation is different depending on where you are in the U.S. I didn't really expect to see quite such historic difference between the Eastern and Western United States. But in looking back at the map it's sort of where West Nile is in the Eastern United States tends to be more in urban areas compared to the Western United States where it tends to be more rural areas-- it makes sense.

Charlie: So what's next in this research? What are you guys working on?

Micah: So the next step is now we've done this proof of concept. So we've sort of proved to ourselves, yeah, these associations between temperature and precipitation and West Nile are strong enough that we think we can find-- you know, pick out a few good predictors that we can use to make these forecasts. So what we're working on right now is sort of expanding the number of predictors that we are using in the model. So, like I said, in this first descriptive paper we just looked at temperature and precipitation for the whole year. And so now we're sort of refining that by using seasonal weather variables and looking at things like the snow cover in an area. We're also including things like land coverage. So is it predominantly in an urban area. Are there a lot of crops in the area, a lot of grasslands? And then putting in some variables to tell us where each of these mosquito species are living. So, basically, we're just trying to, you know, find a few more variables to sort of refine the predictions and then create something that will allow us to put out a forecast of where we expect to see higher risk in the coming seasons.

Charlie: Hahn is lead author and Monaghan co-author of the paper, “Meteorological conditions associated with increased incidence of West Nile virus in the United States,”… published in the American Journal of Tropical Medicine & Hygiene. I’m Charlie Heck, co-editor of Science360’s news service and co-host of the Super Science News Show, at the National Science Foundation. If you have any questions about this story or suggestions for interviews about super cool NSF-funded science, pick me, pick me…email me at editor@science360.gov