How to quickly visualise massive spatio-temporal data

Background

Are you collecting lots of geospatial data and need a way to quickly visualise what you’ve got? Read on! I had this problem; I was collecting a lot of information on my daily movements from Arc App: Location and Activity (more on the use of Arc for research later). Additionally, I wanted to quickly see what features of the urban environment (e.g. greenspaces) I was being ‘exposed’ to. The solution was Uber’s open source geospatial analysis tool: Kepler.gl.

Method

Some formatting of your data is required, but it’s pretty simple, just output a CSV with a datetime, latitude and longitude column, alongside any other attribute data that you are interested in (here I’m interested in greenspace exposure, i.e. when I was within 50 metres of a greenspace). That’s the hard bit. After this, you have lots of options for styling your visualisation. You receive an automatic animation of your data over time, which you have control of with a set of buttons to determine playback. A notable, but more time-consuming, alternative to creating animations from data is the QGIS Time manager.
Once you like the look of your visualisation, there are options to share as an image or a URL. If, like me, you want to do furthering editing (e.g. annotate certain frames) then you can do a screen recording and then import to a video editing program such as iMovie (n.b. There is probably a better way of doing this). Anyway, hopefully this blog will give enough information to get started with kepler.gl. I was able to create a visualisation of my contact with greenspaces over a 6 months period, which is shown below.

Visualisation

comments powered by Disqus