Thursday, April 16, 2015

Transportation Data Visualization 1: Origin-Destination and Migration Flow Visualization

Migration Flow Data Visualization Using Chord Diagrams

The origin-destination (O-D) matrix is a type of migration flow data. The complication of the migration flow visualization is that you need to show multi-dimensional data such as origins, destinations, volumes of movement, and direction of flow in one single plot. Chord Diagram, which is able to show directed relationships among a group of entities, offers a very efficient way for migration flow data visualization:
  • Chord Diagram is a circular migration plot.
  • Origins and destinations of the migrations are represented by circle segments.
  • Flows are encoded as curves; flow movement is indicated by the curves between two circle segments.
  • Volumes of movement are represented by the width of the curve at the origins.

Figure 1 shows the morning home-to-work O-D pattern changes over a 15 year period (1996 - 2011) in the selected 16 municipalities in Southern Ontario. The O-D matrix data is from the Transportation Tomorrow Survey (TTS), a travel survey program collects information about how and where people travel in Toronto and the surrounding areas.

Figure 1

Tool used: d3.js

Migration Flow Data Visualization Using GIS

Chord Diagram is an excellent approach to visualize migration flow. However, Chord Diagram is not able to provide geographical locations of origins and destinations. If you want to include geographic information in your migration flow visualization, you need to use mapping software tools. In this example (Figure 2):
  • The origin and destination locations are shown as "donuts".
  • Centers of the "donuts" represent centroids of the zone boundary polygons.
  • The inner and outer circle sizes are proportional to inter-regional flow volumes and they represent number of trips (from other zones) to the zone and from the zone (to other zones) respectively.
  • The number of internal trips is coded by colours.
  • Flows between two zones are encoded as curves and the curve width represents the flow volume.
  • The flowing direction is clockwise.

Same TTS data sets (as the Chord Diagram) were used to create this migration flow map.

Figure 2
Tool used: PostGIS and ArcGIS 

Truck Movement Data Visualization:

Massive amount of truck movement data made by millions of commercial vehicles (trucks) are collected widely nowadays using on-board GPS units. If we visualize truck movement by connecting the origins and destinations over a large geographic area, the end product will look like spider web. The example in Figure 3 shows the North America wide flow pattern using one week of truck data from one service provider.
 Figure 3

Figure 4 is another technique to visualize truck movement. It shows how we can turn the truck movement pattern into artistic works. In this example:
  • Flows between any origin-destination pairs are encoded as curves.
  • Use the actual origin and destination locations as the curve starts and ends.
  • Curves (flows) are colour-coded by origin cities.
  • The flowing direction is counter-clockwise.

Figure 4


Tool used: PostGIS and ArcGIS