All projects
Next

Wildfire Risk to Communities

Improving a map user experience to assess wildfire risk for vulnerable populations

Project overview

Wildfire Risk to Communities is a free, easy-to-use website with interactive maps, charts, and resources to help communities understand, explore, and reduce wildfire risk. It was created by the USDA Forest Service under the direction of Congress and is designed to help community leaders, such as elected officials, community planners, and fire managers. This is the first time wildfire risk to communities has been mapped nationwide.

Client

U.S. Forest Service (via Headwaters Economics)

My role

Product designer: Risk Explorer

My role

I took over as the design lead on the Explore Risk web application during the implementation phase of a redesign. I worked closely with the client and our developers to resolve a number of unforeseen design challenges during development, such as adding a fourth risk category, improving the Recommended Actions UI and copy, and accommodating long community names for tribal communities.

Improving the Vulnerable Populations map

My main contribution to date has been an overhaul of the ‘Vulnerable Populations’ map UI. This section of the application lets user overlay wildfire-related datasets with U.S. Census Bureau data on potentially vulnerable populations by census tract.

Vulnerable populations (VP) are people who are more likely to be affected by wildfire because they lack access to resources, experience cultural and institutional barriers, have limited mobility, and/or have compromised physical health.

The problem

In their user testing the client had discovered a number of usability issues, including that the VP map layer was hard to see on certain basemaps, and that users struggled with terms like “(Census) Tract.” They also wanted to add population sub-categories, language explaining how populations were sorted and selected by default, and the ability to download a PDF report for each tract.

A few bandaid solutions were tossed around until I convinced the client and development team that it was worth taking a step back to overhaul the map — considering all existing and new requirements holistically.

Improved map style

First prototyping in Figma and then working closely with a developer, I improved the map style to increase the visibility of community and census tract boundaries, highlighted features, and labels against multiple basemaps.

Old map style (left) vs. improved style (right)

Improved map UI and copy

To improve the usability of the map UI, we iterated on the language of legends and tweaked the typography.

Previous design (left) and an iteration (right)

The map used a UI pattern where controls and layer toggles were grouped on one side and legends on the other. This approach is well-established but was causing trouble: It was difficult to label selected VP tracts in the legend without obscure or lengthy language.

I explored an alternate approach where legends are grouped with their related UI controls and conducted a quick ‘think-aloud’ user test with two colleagues. Both were shown prototypes with the old and new layout (each presented with a different version first). I asked them describe their experience while they used the prototype and had them rate the design. Afterwards, I showed the alternate version and asked for thoughts and a second rating. Both colleagues preferred the new version of the map UI.

The final map UI with a new layout and improved, consistent legends for all map feature types

Reimagined map interaction design

Previously, VP data was shown in a tooltip when the user hovered over a census tract on the map. With more categories, the tooltip outgrew the map and was clipped due to technical constraints of the mapping library. Additionally, to add the ability to download reports by tract, there needed to be a select state for tracts.

I designed and prototyped a new solution Figma that would move VP indicators from the tooltip into a dynamic UI panel. I iterated on the map behavior when clicked (Pan, no zoom) and how to give the user feedback when no tracts match their VP criteria.

Recording of the interactive Figma prototype to prototype the map interaction. Note that the map layers don't update in this prototype. Check the production site for a working example.

In the mobile version, I added a delay on the tap event to give the user feedback which census tract they selected.

Recording of the mobile prototype

Outcome

The improved Vulnerable Populations page provides community leaders with immediate insights about the neighborhoods where social and economic factors may make it harder for people to prepare for and recover from wildfire. Users can now download customized reports with detailed data for every tribal area, community, county, and state.

Due to the nature of the contract I could not gather feedback directly from users. The client was pleased, however, and reported that users were too. The page was apparently even discussed in Congress.

Go to websitewildfirerisk.org/explore