Maps of the Week

I've been thinking a lot about the potential of Terrapattern this week. Terrapattern uses deep learning machine vision techniques to search for similar patterns in satellite imagery.

It is a technology which could truly revolutionize a number of fields. For example in archaeology aerial imagery of historical sites is often used to determine sub-surface features based upon features which can be seen on the ground - especially when viewed in detailed aerial imagery. If you fed in the satellite image of a round barrow into Terrapattern then it could identify other locations with similar features.

At the moment Terrapattern only works for Pittsburgh, San Francisco, New York and Detroit. However other cities are coming soon. The Terrapattern about page has examples of things that you might want to try searching for, such as baseball diamonds, airplanes or solar panels. Part of the fun of Terrapattern however is just clicking on the map to see how quickly it finds other similar looking locations.

That 'about' page also includes a lengthy 'How it Works' explanation of the neural network behind Terrapattern.


Blast Map is an interactive map showing underground nuclear tests carried out by countries across the globe since 1963. 1963 was the year the Partial Test Ban Treaty came into force, which prohibited all test detonations of nuclear weapons except underground. Using data from the Northern California Earthquake Data Center the map shows all seismic events which have been determined to be not geological in nature but were instead caused by either quarry blasting or nuclear testing.

Blast Map shows the location of these quarry blasts and nuclear test sites around the world. The chart beneath the map shows the magnitude and date of each of the blasts. The chart and map are synchronized together so that the chart automatically updates to reflect the data in the current map view. You can also use the chart to refine the data shown on the map by range of magnitude and date.

The map sidebar provides links to significant nuclear testing events. For example, if you click on the 'Soviet Nuclear Archipelago' link you can view a map and chart view of soviet nuclear testing from 1964-1991. You can read a little more about the significance of these highlighted testing events and how the map was made on this Adventures in Mapping blog post.


Mark Evans has used the Google Maps API to create a hypnotic visualization of commuting flows, showing the distances and 'journeys' that American's make to and from work.

Using the ACS Commute Map you can zoom in on any U.S. county and view an animated map showing where people live and work. The maps don't show the actual journeys that commuters make but do give a great sense of how town and city centers suck in commuters from surrounding suburbs.

The data for the maps comes from the American Community Survey. You can learn more about how the map was made from this data on Mark's blog post ACS Commuter Data Visualizations. Mark's ACS Commute Map was originally inspired by Alasdair Rae's mapped visualizations of commuting in the Bay Area.