As a deluge poured down in Grass Valley, California, in late 2012, software engineer Nelson Minar took to a cerebral rainy-day activity: tinkering with datasets and open-source software. He had recently learned of the National Hydrography Dataset—a collection of digital vectors compiled by the U.S. Geological Survey representing the drainage networks of rivers across the country—and the downpour outdoors inspired him to take a closer look at how the data could be used with mapping technologies.
Minar would be the first to caution viewers that the map doesn't necessarily indicate the location of rivers but rather identifies the elevations where water is likely to flow. The map is far from perfect, with certain areas lacking topographical information, but Minar is intrigued by the sculpting force of water—how, over millennia, rivers have etched the surface of the United States. "I was struck by its beauty," he said of the map. "We read about water systems being fractal, and beautiful—here you can really see that."