How Bloomberg News is Breaking Boundaries | by Mapbox

Behind the scenes with Interactive Graphics Journalist Cedric Sam

By: Ryan Mannion and Cory Guillory

More than six months after the global pandemic occurred, data on cases and deaths from Covid-19 is still available everywhere. Last week, the team at Bloomberg Graphics News shared a new take on how important it is that we think about location and boundaries with proper context in their article “Mapping the Coronavirus: How Many Covid-19 Cases Are Near You.” Their 🔥 visualizations are partly built with the Mapbox API, so we took a moment to talk to them Cedric Sam from the Bloomberg Graphics team (@BBGVisualData) to learn more about their journey to spread this story.

What inspired you to tell this story?

We started with the belief that the number of Covid-19 cases in each state, let alone nationally, was too broad and that data at the county level was often too little. We saw what would happen if we expanded aggregation beyond the administrative or metropolitan boundaries we usually read about to areas that are more volatile and reflect how people actually move in everyday life.

The result is an interactive visualization where readers can obtain information that may tell a different story than what someone hears on the local or even national news. They can enter a location of their choice and then switch between seeing daily deaths and new cases in areas within a two-hour drive of that location. There’s a time slider at the bottom that starts on March 12 and continues through today, then there’s a dynamic line chart above the map that compares a given region to other regions in the number of cases and deaths.

Study some of the most interesting examples you find that show how important this type of location context is.

The interactive really lets you dig into connected areas. For example, Ripley County in Indiana, a rural area with a population of 28,000, but within a 2-hour drive can reach 7.3 million people who live in urban centers such as Cincinnati, Indianapolis, and Louisville.

Another country is California. The Fresno area of ​​the San Joaquin Valley and the areas surrounding San Francisco and Silicon Valley are located next to each other, but the Fresno area was hit harder this summer.

Explain how you built this visualization.

We use a collection of county-level Covid-19 case and death data from our friends @nytgraphics across the city, which we divide into sections per region that are updated at least once a day.

To identify the most reliable driving distances between countries, we queried the Mapbox Matrix API for each county or county equivalent in the United States compared to its 100 or so closest neighbors. Then, using Svelte and Mapbox GL JS, we created a map interface with a custom search box that allowed users to select specific places, which retrieved up to several dozen historical data files for each region within a 2-hour drive.

We performed heavy geoprocessing on the base map using a variety of tools, including Dirty Reprojectors as well as Mapshaper, a tool from NYT Graphics Editor Matthew Bloch.

What challenges did you encounter in this article and how did you overcome them?

Getting the driving distance for each county was one of the more difficult technical parts of the project.

To determine proximity between regions, we were unsure whether using the centroid of each region or its outer boundaries and calculating the straight line distance would be the best measure. We determined that the representative city in a county, that is, the county seat in 98% of counties, was the point per county we wanted and used a data set from DOT. We also determined that driving distance is a better metric for quantifying connectedness, although it is more difficult to obtain (no pre-computed data appears to exist).

Now, how do we measure connectedness across the 3,000 or so counties or county equivalents in the US? We initially looked at the Mapbox Isochrone API, but the maximum distance was 60 minutes, which is not far enough for a region. So we looked at the driving distance APIs out there that we could use and ultimately settled on the Mapbox Matrix API. To be able to measure driving distance between the 3,000+ counties and the other 2,999 counties, we did some pre-processing to reduce the number of API calls.

After getting the distance data, we parsed it and created a compact structure, then listed the counties within 7,200 seconds for each county, using its FIPS identifier.

The Dirty Reprojectors technique mentioned above also adds another layer of complexity in handling user interactions. The points on the map that the reader interacts with are not the same as the actual pairs of latitude and longitude numbers. On the other hand, these points are points around Null Island, the term for the point on the earth’s surface where the prime meridian and the equator line intersect. So, every user interaction (every map touch) is a query that must determine which polygons on the Map Grid map layer were touched. Then, a series of functions (involving using Turf to create contours around the selected area) will visualize the selection on the map and a line chart above the map.

Did you have any unexpected learnings while putting it together?

Building this interactive gives us the possibility to start with high-level external observations and be able to drill down to determine whether there are indeed real differences between adjacent regions. For example, we’ve been hearing that for about a week, cases in northern Wisconsin have been increasing. The visual shows that on September 21, there were twice the number of new cases per capita around Wausau than around Madison. Instead, you can start with an area where you live that may not be known to have many cases per capita, and then see from the interactive that you live within close driving distance of an area that is seeing an increase in cases. I think that is a very strong feature of our project. We looked into it further on Twitter.

Is there anything else we didn’t ask that you’d like to share?

To encourage readers to explore and share what they find interesting, we added hashtags to the main URL as they interact with graphics and search for points of interest. For Chicago, you can save and share these views, for example. (Hint: share!)