If you’re interested in how GIS is applied to the world, this episode of 99% Invisible is for you, but the topic predates modern GIS. The US Army Corps of Engineers built a 1:2000 scale model of the entire 1.25 million acre Mississippi River watershed on 200 acres of land - and it could predict flood stages on real levees within two inches. We can do much of this (but not all of it) with modern GIS, but it’s a fascinating podcast to listen to.

My office ran an analysis using Sentinel 2A satellite imagery to find out just how much of Baton Rouge flooded, and how much of that flooding was outside of floodplains, or in areas designated as protected by levees. The word is that lots of flooding there is due to backwater effects (water not being able to clear out quickly enough from rain), and it’s extensive. In our incomplete analysis, we found 30 square miles (78 sq. km) of flooding in the unclouded part of the image. This web map will let you explore.

A few months ago, NASA’s image of the day highlighted a process they use to automatically detect landslide activity around the world using remotely sensed imagery.

An incredible animated and interactive visualisation of global commercial shipping, created using GPS data for ship locations. Watch ships move around the globe or idle outside of ports. It’s pretty mesmerizing, and demonstrates what you can do with a well-constructed visualization.

The gold standard of Landsat data viewers. It’s been out for a long while, but it’s never been in this newsletter! USGS maintains an excellent viewer of its own, but the EOS viewer retains power and is incredibly easy to use.

I don’t have an Apple device, but this demo for a Landsat data viewer for iOS is pretty impressive and probably worth checking out for those of you who have access. There are many companies pushing the limits of data display and analysis and bring remote sensing into reach for people without access to traditional GIS tools now!

In a past issue, I highlighted a way to learn PostgreSQL - well, now you can bring it into the GIS realm with this great tutorial from Boundless.

Spatialite is a version of SQLite for spatial data - effectively implementing similar functionality to PostGIS in a serverless form. I’ve used it a bit in the past and am about to use it on a new project - maybe some of you will find it useful too.

Esri has a nice connector from R to ArcGIS, and now there’s a connector for R and QGIS. R already has some of its own spatial capabilities (through GDAL/OGR, most often on the analysis side, and then Leaflet for the web), but having a package to directly access the more than 1000 QGIS processing algorithms should provide far more advanced capabilities.

Want to have some fun building a maps with basic terrain and territories? This page has a writeup of how the author built the tool (in both Python and Javascript), as well as the ability actually generate maps, and play with the specific portions of the tool to see how each behaves. It’s part mapping thought experiment, part fun tool to make a map.