AccuWeather to offer users innovative flu forecasts from University of Virginia's Biocomplexity Institute
A new agreement between AccuWeather and researchers at the University of Virginia's Biocomplexity Institute will soon bring cutting-edge flu forecasting technology to AccuWeather app users, helping to increase the level of flu knowledge and awareness for people across the United States.
The strategic partnership will allow AccuWeather to provide its users with information about projected flu outbreaks in specified regions before an outbreak occurs.
Prior to the agreement, brands were limited to targeting regions already experiencing flu outbreaks.
Through the partnership, brands can now partner with AccuWeather to deliver messages to consumers in U.S. counties weeks ahead of an outbreak.
This is made possible through EpiCaster, a disease surveillance tool that enables University of Virginia scientists to forecast how certain regions could be impacted by the flu.

The disease surveillance and forecast platform was originally developed to help guide the U.S. government’s response to West Africa’s Ebola outbreak in 2014.
“EpiCaster gathers data from lots of different sources, combines it, runs it through a couple of simulation models and then fuses those together to offer forecasts relevant to particular stakeholders,” said Bryan Lewis, associate research professor at the Biocomplexity Institute of U. Virginia.
Biocomplexity Institute researchers, who are experts in fields including epidemiology, computer science and physics, use surveillance data gathered by the Centers for Disease Control and Prevention (CDC) and also monitor Google health trends and social media posts related to possible flu cases.
“If you see people starting to Google a bit more about a disease, it might be that the disease is on the uptick there,” Lewis said.
Researchers use that information to run different simulation techniques for predicting how and when the flu could further spread in the future.
The flu forecasts intend to help people take actions that will benefit the health of themselves and family members, including getting the flu vaccine.
“You’ll get a two-week lead time for the vaccine, or you can make better decisions, much like from a weather forecast, like if you need to pack an umbrella or take that warm coat with you if it’s going to be chilly in the afternoon,” Lewis said.
However, unlike with weather forecasting, members of the public will have the potential to influence the outcome of flu forecasts.
“If you make a good forecast for an infectious disease, people are able to control something about it,” said Zane Reynolds, advanced software innovation architect at the Biocomplexity Institute.
“For example, if we were to say, ‘the flu is going to be really bad in your town,’ you might not send your kids to school, you might wear a mask or you’re careful about washing your hands; you can actually affect that forecast, so it won’t become as bad as it was predicted,” Reynolds said.
As disease forecasting is still in its infancy, U. Virginia researchers hope that the partnership will help increase the accessibility of forecasts for the flu and other diseases for members of the public in the future.
“We want to make disease forecasting have the same levels of vigor, quality and pervasiveness that weather forecasting has, because we think it will actually change the way people are likely to behave,” said Madhav Marathe, computer science professor and the Biocomplexity Institute’s lab director.
“People almost always look at the weather to decide what clothing they’ll wear and decide to make changes in their activity, and we think that disease forecasting can have a similar effect as it gets better,” added Marathe, who played a key role in developing the EpiCaster system.
Flu forecasts are expected to be made available to AccuWeather users by spring of 2018.
Report a Typo