Analysis 1: This gives a range of 240,000 to 320,000 marchers down Pennsylvania Ave, and is based on the time-lapse march route video plus measurements of the route taken from Google Maps. The time-lapse sequence was taken from a webcam over Freedom Plaza at 14th St and E NW, looking ESE down Pennsylvania Avenue toward the Capitol. The time-lapse video is 41 seconds long (0-40), is labelled as covering 8 AM to 1130 AM on 9/12/09, and thus scales at just over five minutes real-time per second of video. The video shows the march starting out of Freedom Plaza at 17 seconds (~9:30 AM) with the tail end leaving the plaza at 36 seconds (~11:10 AM). Google Maps sat view, meanwhile, shows Pennsylvania Ave. to be just over 220 feet wide (eight traffic lanes plus sidewalks) along the march route, and the length of the march route from the exit of Freedom Square to where the road reaches the Capitol West Lawn to be just over 4400 feet. The video seems to indicate marchers filled both road and sidewalks for the entire march route for most of the time of the march, FWIW.
The video does not seem to provide enough resolution, spatial or time, to directly measure the rate of advance of the march. The speed with which the route initially fills up shouldn’t be used (the fast ones always end up out front) nor the speed with which it empties (the slow ones are at the back). I’ll make the assumption here that the march averaged 1 mph - large dense peaceable crowds tend to have much internal friction and move slowly, 1-2 mph in my experience; I think 1 mph is mildly conservative.
So we know the crowd took ~100 minutes to march past the east end of Freedom Square on a route ~220 feet wide, and we’re assuming they averaged a 1 mph rate of advance. That would make the entire crowd about 8800 feet long by 220 feet wide (had the route of march actually been that long). In other words, the crowd occupied about 1.94 million square feet at their march density, at an assumed march speed of 1 mph.
We now need a second assumption: how tightly packed the crowd was as they marched.
I estimate that a really tightly jammed crowd (stage-front at a concert) takes about two square feet per person. A dense elbow-to-elbow crowd on the move is three to four square feet per person, and a polite relaxed crowd on the move is six or more square feet per person. This marching crowd seemed very densely packed in the overhead video, but from various closeups I’ve seen it seemed more like 6-8 square feet per person. 1.94 million square feet divided by six square feet per marcher gives about 320,000 marchers; divided by 8 gives about 240,000 marchers.
To sum up, the width of the march route and the time for the march to pass one point are known. The two main assumptions in this crowd estimate are the average speed of the march and the average spacing of the marchers. 1 mph and 6-8 square feet per person seem mildly conservative estimates; based on those we get a crowd size of multiple hundreds of thousands.
Note too that this march-route estimate does not cover anyone who arrived at the march destination by other routes.
Analysis 2: This one gives a range of 330,000 to 500,000 demonstrators in front of the Capitol, and is based on the NYT description of the crowd (”A sea of protesters filled the west lawn of the Capitol and spilled onto the National Mall) plus the first photo in the story at the Daily Mail, showing the Capitol West Lawn during the protest, plus measurements taken from Google Maps sat view of the area.
The overall West Lawn of the Capitol area is a square just over 1500 feet on a side. Between the Reflecting Pool, the Botanic Gardens building, and various other obstructions, I estimate about 75% of that area is actually available for a crowd. That’s comes out to about 1.7m square feet.
The National Mall, meanwhile, consists of eight squares of about 600 by 600 feet each, (about 360,000 square feet each) in a line west to the Washington Monument. I will assume for this analysis that the crowd described by the NYT extended only to the first of these eight blocks, giving us 1.7m + 360K square feet, or about 2 million square feet of crowd.
Crowds at rest take less room than crowds on the move; I therefore assume a range of 4 to 6 square feet per person in this crowd. The above-pointered picture in the Daily Mail seems to support the 4 square feet end of this range, but it was taken from just east of the Reflecting Pond and shows the front portion of the crowd - the crowd density likely drops further back. However, there is also uncertainty about just how far back the crowd goes - at least one attendee claimed in a blog comment that the crowd described by the NYT extended the entire length of the Mall. Applying the 4 to 6 square foot range to the West Lawn plus first block of Mall area seems reasonable for now, absent better data. This gives us an estimated 330,000 to 500,000 demonstrators in front of the Capitol yesterday.
Obviously there are significant error bands in both of these estimates. There’s also room for better data; in particular I’d be interested in any marchers who accurately timed their march from leaving Freedom Square to first arriving at 3rd St where the West Lawn starts, as well as any more info on how far back along the Mall the crowd extended and how dense the crowd was.
I think it’s already very clear, however, that “hundreds of thousands” is the correct description of the size of the 9/12/09 DC protest.
