We were monitoring the crowd build up from 0600 (Washington DC time) from 7 different live TV feeds. These are the images of the crowds 45 minutes before the Oath of Office. From the front, the crowd density looks similar, but note the white areas near the back of the 2017 crowds.
From the Podium – the low angle would give the illusion of a packed area stretching far into the distance. This is a common crowd illusion and often leads to overestimating the crowd size. You need to see ALL angles to understand the actual attendance.
From a lower angle, we can see the spaces more clearly.
From the rear view we can see that the front 1/3rd is full (ticketed/seated areas) but the back 2/3rds are empty.
When we use a plan image, mapping the areas, we see the relative crowd size on the Mall.
Using both the visual analysis and a plan/area analysis we have a 1/3rd of the space occupied in the Trump (2017) inauguration compared to the Obama (2009) inauguration.
This is not a political statement, just a crowd estimation. Our initial, real-time estimate of 1/3rd was later confirmed with the area estimate.
From the New York Times (based on our analysis)
We can produce a comparative analysis of area very quickly and therefore a comparative crowd size (given crowd density for similar types of events). So the 20th Jan 2017 had approx. 1/3rd the crowd occupancy of the 2009 event. The 21st Jan had approximately 3x the crowd of the 20th Jan event. When high-resolution images become available, we will update our analysis.
Analysis by Keith Still and Marcel Altenburg – Manchester Metropolitan University.