The Information Architecture of Forecast Products: Leveraging Relationships to Communicate Complex Findings

A forecast cannot effectively communicate risk—or anything else—if people coming to a website are unable to find it. This is a problem that haunts information-dense websites across many sectors. Any organization that produces hundreds of individual pieces of content each day faces the daunting problem of making that information findable online. Information architecture is the practice of organizing and labeling websites to make information findable and usable. A website's information architecture includes both the navigational structure and the underlying relationships between the different content items on the site. The relationship between different forecast products is an underexplored problem in the development of forecast-focused websites.

The Weather Prediction Center (WPC) of the National Weather Service (NWS) is no exception. WPC has a complex mission—it tells the nation's weather story; produces a variety of hydrometeorological forecast products that cover liquid, frozen and tropical precipitation events; and increasingly takes a leading role in forecasting extreme temperature events.

The WPC website was launched in 1997, a year before Google existed. At that time, there were still more than 100 people on the web for each existing website. (Today there are fewer than 4 people online for each website.) A significant percentage of the WPC website still reflects that original 1997 design and navigation structure. A refresh in 2001 brought updated designs and technology to parts of the site, though many sections continue to use the original 1997-era designs and user interface. Like many older websites, the WPC website has grown organically, as new products were added and new technologies (e.g., GIS software, application programming interfaces [APIs]) became available. A natural result of 27 years of organic growth is a website that is difficult to navigate.

User feedback indicated that the variety of WPC products, coupled with the site's long period of organic growth and aging information architecture, have made it difficult to find material on the website. Usability testing of the existing site confirmed these intuitions: on average, regular users of the WPC website could successfully locate materials slightly less than half of the time.

Fountain Digital Consulting—in partnership with the Nurture Nature Center—is leading a project to test the usability of the WPC website and to make evidence-based recommendations for its improvement. The study includes extensive interviews with WPC staff to help understand the relationships between WPC products as well as a series of usability studies to determine the extent to which users of WPC products (a) understand those relationships and (b) can leverage that understanding to find relevant related products on the WPC website.

The Fountain-NNC team developed an extensive set of product mappings that visually encode relationships between products (e.g., WPC's Excessive Rainfall Outlook translates its Quantitative Precipitation Forecast (QPF); its Extreme Precipitation Monitor contextualizes the QPF). A second phase of the project examined the degree to which users had already internalized these relationships. A third phase will test whether integrating those relationships more explicitly into the website’s navigation.

The project seeks to integrate best practices in digital user experience research into risk communications, helping forecasters tell a more integrated story by explicitly linking multiple products that deepen understanding of the full weather picture.