200,000 Rivers Run Through It

The U.S. Geological Survey (USGS) has published a fascinating visualization exploring the language of rivers in the United States - a map that doesn’t just show waterways, but the words we use to describe them. Their Map that Glows with the Vocabulary of Water reveals how terms like arroyo, bayou, creek, and marsh are not randomly distributed, but instead cluster in ways that reflect the geography and climate of the United States.

Inspired by the map my first thought was could I create an interactive version. My USGS Stream Vocabulary Map brings the concept to life by allowing you to explore, filter, and toggle different categories of stream names, to help more effectively reveal the patterns that might be more difficult to perceive in a static graphic.

Mapping the Vocabulary of Water

The underlying data comes from the U.S. Geological Survey’s Geographic Names Information System (GNIS), a comprehensive dataset of named geographic features across the United States. For this visualization, I’ve focused on named water features - rivers, streams, creeks, and related forms - which together represent tens of thousands of individual data points spread across the country.

In total there are 220,316 named streams in the lower 48 states. The map may may therefore take a few seconds to show them all.

An Interactive Take on the USGS Visualization

In the original USGS visualization, the distribution of these names was presented as a static map. The interactive version includes several layers of exploration:

• Individual categories of stream names can be toggled on and off
• Ephemeral and persistent naming groups are visually distinguished
• A secondary layer allows users to reveal more common, “neutral” terms like river, creek, and run
• An information panel updates dynamically to explain what is being shown on the map

What the Map Reveals

When viewed through the lens of linguistic categories, clear geographic patterns begin to emerge in the names of U.S. rivers.

Names associated with ephemeral water — such as arroyo, wash, draw, and gulch - are concentrated in the arid western and southwestern United States. These are regions where rainfall is infrequent and waterways often flow only temporarily after precipitation events. The naming reflects this reality: the language itself encodes the intermittent nature of water in these landscapes.

In contrast, names associated with more persistent water - such as bayou, swamp, and marsh - are more common in the southeastern United States. This region is characterized by higher humidity, abundant rainfall, and low-lying terrain that supports wetlands and slow-moving water systems. Here, water is less transient, and the vocabulary reflects environments where it tends to linger.

When additional, more generic terms like river, creek, brook, run, and fork are included, the geographic distinctions become less pronounced. These names are widely distributed across the country and are less tightly coupled to specific hydrological conditions. 

A Subtle Hydrologic Boundary

With the full dataset visible, a broader geographic structure becomes apparent - one that aligns with a well-known climatic transition in the United States: the approximate location of the 100th meridian. This boundary roughly separates the wetter eastern half of the country from the drier western half, and it has long been recognized in geography and climatology as a meaningful division in precipitation patterns and land use.

Language as a Geographic Signal

What makes this visualization particularly compelling is that it treats language as data. The names we assign to natural features are not arbitrary - they are shaped by how people experience their environment over time. In this case, stream / river names act as subtle indicators of climate, hydrology, and regional history.

By making the map interactive, these patterns become easier to explore and compare. Turning layers on and off, revealing different categories of names, and observing how they cluster across space transforms the map from a static illustration into an exploratory tool.

Also See

The original USGS article on this data credits John Nelson as inspiration for the firefly visualization aesthetic. The USGS map also immediately reminded me of a 2018 map by Nelson - Gap, Notch, Pass and Saddle - which 'explores the regional naming patterns' of mountain passes in the United States.