Spatial Awareness #4 — Maps & Spatial Newsletter by Robin Hawkes

This hypnotic map by John Nelson shows the seasonal ebb and flow of sea ice as it freezes and thaws over the past 40 years. He also looks at alternative methods of visualising the same data to explore how sea-ice extents have changed in 40 years.

This beautiful monochrome map from Lauren Tierney contains all sorts of interesting details and techniques. I’m particularly fond of the text and the transparency effect towards the edges of the map. Very nice! 👌

Andrew Tyrrell has used LINZ LIDAR data to create a crisp map with detailed shadows where the elevation changes. If you’re interested in this effect then check out the go-to tutorial on the technique by Daniel Huffman.

I’m a sucker for 3D spatial visualisations and this one of Super Typhoon Nepartak by Craig Taylor is no exception. He’s taken NASA precipitation radar data and used reflectivity values to define both colour and intensity of the 3D points.

The Bartlett Centre for Advanced Spatial Analysis (CASA) has released their Colouring London project, which attempts to use crowd-sourcing to collect information on every building in London. So far the collected area is small, though this will likely change as more people get involved. I’m curious to see how this turns out.

It’s now possible to create isochrone visualisations in QGIS using a new plugin by TravelTime. The plugin supports multiple modes of transport and is free to use for the first 5,000 requests per month.

Earth Lab at the University of Colorado have released earthpy, a Python package to support working with spatial data. It can be used for both raster and vector data and can do things like creating hillshades and other types of analysis.

Vladimir Agafonkin (creator of Leaflet) has released v4 of Delaunator. It’s a super-fast library for Delaunay triangulation and the latest update improves performance by up to 10%. His other libraries are worth checking out as he’s obsessed with making them as fast as possible — handy for animation and 3D visualisation.

Uber have integrated their kepler.gl 3D spatial visualisation library into Jupyter Notebook. It’s now possible to plot complex maps directly inside a notebook by using Python as you would normally. I’m excited to see what sorts of things people create using this.

NASA are using machine learning to detect particles emitted from ships as they work their way across the oceans. They do this using data from the Moderate Resolution Imaging Spectroradiometer (MODIS) on their Aqua satellite. It’s even possible to detect the ships at night!

This fantastic article by Datawrapper covers the various things to consider when using choropleth maps. It’s a useful reference for understanding when to use a choropleth and how to make them communicate things better.

Mapbox collect over 300 million miles of anonymised location data every day to produce speed estimations for routing. This article by Saba El-Hilo goes into detail about how they use Apache Spark to calculate 30 billion of those speed estimations each week.

SharedStreets have created an open standard for curb restriction data (eg. parking restrictions) and have been working on integrating this standard into OpenStreetMap. The idea is that governments can release curb-restriction data and it can be directly integrated into tools and maps using this standard.

If you’re at all interested in data related to shipping then you should check out the MarineCadastre.gov vessel traffic data. It includes minute-by-minute snapshots of vessel positions off the coast of the US since 2009.

Storing spatial data in a GeoPackage is particularly hot at the moment. You can now access OpenStreetMap planet data in GeoPackage format, updated every two months.

Do you know someone who has had a big impact in Free and Open Source Software for Geospatial (FOSS4G)? The Sol Katz award is given to people who have demonstrated leadership within the FOSS4G community. Nominations close at the end of August 8th.

Are you interested in the future of Earth Observation? The ESA Earth Observation Phi-Week is taking place in mid-September and aims to look at the latest developments as well as what’s coming next.

Are you working on something related to remote sensing? The Geological Remote Sensing Group are currently accepting proposals to present at their annual conference in December. They’re particularly interested in presentations that cover longwave infrared and radar / InSAR.

Planet are looking for someone to work on end-to-end validation of their satellite imagery data — monitoring trends, detecting issues, evaluating performance, etc. The role is in Berlin and you need to have experience with remote sensing data.

The University of Hong Kong are looking for someone at PhD level with expertise in transport geography. The role will be to teach undergraduate and postgraduate students and conduct research in fields related to transport geography.

The University of Bremen are looking for someone at Masters level in a field related to geophysics. You’ll be working on the SLATE project, researching submarine landslides and creating a 3D slope model.

I’ve been a fan of Craig Taylor for a while now — his work in 3D spatial visualisation is second to none and he is a constant source of inspiration. He works at Ito World and creates all sorts of beautiful visualisations; ranging from infrastructure and air traffic, to tracking his dogs using GPS as they tear around a local field (above).

Decentralised maps are becoming more and more important as we move towards centralised platforms for spatial data and maps. Peermaps and Mapeo — suggested by @dietrich and Karissa McKelvey — are two methods for putting the control back in the hands of users. Instead of relying on centralised servers, you’re able to share spatial data and maps using peer-to-peer technology, which means that it’s possible to use even offline. This is particularly powerful for humanitarian projects and those in disaster areas where an Internet connection can’t be guaranteed.

Zoran Čučković (QGIS Whisperer) is raising funds to work on a new QGIS plugin to create realistic shadows from elevation data. It’d be amazing to be able to achieve this sort of effect directly inside of QGIS instead of using software like Blender.