Monday, 22 June 2015

Two gull migrations - visualised

Lifewatch INBO (Belgium) are monitoring Lesser Black-backed Gulls (Larus fuscus) using GPS 'backpacks' developed by the University of Amsterdam. They've already created some great visualisations, and data from previous years has been made available as open data, for others to explore.
In this visualisation, follow Gull 719 (aka "Harry") in red, and Gull 623 in blue, on their movements over one year. See below for how the map was made (although I'll probably keep tweaking it!). Full screen map here
I've used Leaflet to setup the map backdrop combining two different sources, then added two separate map visualisations already created within Cartodb by taking chunks of Lifewatch INBO's data. The visualisations are made using the Torque function. Using the visualisations loaded separately onto the backdrop map, I've got two Torque animations working at once in the same map, which isn't supported within Cartodb itself.

One problem - clicking the pause button only pauses one of the gulls! I'm going to keep tweaking it to improve..

Gull tracking data has been published to the Public Domain CC0 1.0.
Stienen E, Desmet P, Hernandez F, Bouten W, Lens L (2014): Bird tracking - GPS tracking of Lesser Black-backed Gull and Herring Gull breeding at the Belgian coast. v5.2. Research Institute for Nature and Forest (INBO). Dataset/Occurrence.


  1. Nice trick with having to Torque visualizations above each-other. :-) The reason this is not supported by CartoDB (yet) is that they cannot guarantee that the time is synchronized between the two layers and - as you mention - the interaction only works for one. An alternative would be to use the "Torque category" wizard, where you assign a category based on bird id and then give colour to each category. The only disadvantage really is that for categories close together in time and space (e.g. both birds at the colony), one colour will prevail. For an example of Torque categories see what we did in this post: CartoCSS is referenced at the bottom of the post.

  2. Belated thanks for the comments (been away doing seabird surveys)! I'd like to try "torque category" but will need to combine two birds in one dataset - in this case I'd originally saved out details on a per bird basis as I was getting near my data limit on my account! I've had a look at Lifewatch's SQL API too but could only load up a limited number of features at once - will have another look shortly. Plan to try some more visualisations using the gull data so will keep you posted..!

    1. p.s. love the migration visualisation by the way!