Light Pollution in North America

There's a well-known rule of thumb for estimating light pollution called Walker's Law, formulated by the professional astronomy Merle Walker based on measurements in California. It postulates — among other things — that the level of light pollution is proportional to the population of the town or city that's producing it. As I said in my last blog, this is clearly untrue when comparing cities in different countries. But how well does the rule hold up within Canada and the United States?

If the Light-Pollution Atlas is to be believed, not very well. As the clips at right demonstrate, the Atlas shows a much smaller light-pollution blob for Toronto than for Montreal, though Toronto's population is significantly bigger. And Chicago's light-pollution blob is a smidge bigger than that of Los Angeles, which has nearly twice Chicago's population.

The six biggest metropolitan areas in the U.S., according to the 2000 census, are listed below in order of the size of their light-pollution blobs. This is the number of red and white pixels in the Light Pollution Atlas's world map (as best I can figure it after editing out coastlines) divided by the cosine of the latitude, to compensate for distortion in the map's projection. The number of (corrected) pixels per million varies from 39 for Los Angeles to 105 for the Washington/Baltimore conurbation.

Light Pollution Blobs in the U.S.
Metro Area (2000 CMSA)	Pixels (Corrected)	Population (Millions)	Pixels per Million
New York	937	21.2	44
Chicago	745	9.2	81
Los Angeles	650	16.6	39
Wash/Balt	610	5.8	105
Philly	539	6.2	87
S.F. Bay	347	7.0	50

And here are the four largest metropolitan areas in Canada according to the 2001 census:

Light Pollution Blobs in Canada
Metro Area (2001 CMA)	Pixels (Corrected)	Population (Millions)	Pixels per Million
Montreal	612	3.7	165
Toronto	419	4.7	89
Ottawa	121	1.1	110
Vancouver	94	2.0	47

Numerous caveats apply. As I've said many times, it's not at all clear how well the Light Pollution Atlas corresponds with reality. Maybe the apparent darkness of Los Angeles is due to the fact that it was blanketed in smog when the relevant satellite pictures were taken. Maybe the relative brightness of mostl Canadian cities is due to snow. Maybe there are errors in the methodology. Population statistics are also problematic.

But let's pretend for a moment that the data are correct. What could cause such big discrepancies? One obvious answer is population density. One would expect the New York metro area to have relatively low light pollution per capita because such a large fraction of its population lives in multifamily homes, which are far more efficient in every way than single-family homes. But Los Angeles is the archetypal sprawl city for the entire world, and my figures give it even lower light pollution per capita than New York.

Truth be told, Los Angeles is actually quite densely populated compared to the exurban sprawl that grew around many U.S. cities in the last few decades. L.A. did much of its sprawling in the 50s and 60s, and then was stopped short in most directlons by mountains and federally protected land. Many of the houses date to an era when a 1,500-square-foot home on a quarter-acre lot was considered extravagant. Streetlights are probably the single biggest source of light pollution, and they're proportional to the number of street-miles. The more closely houses are spaced, the fewer streetlights per house.

How about Montreal versus Toronto? A respondent to my earlier blog suggested one possible explanation. Montreal is in Quebec, whose vast hydropower resources make it a net energy exporter. The respondent claims that Hydro-Quebec encourages everyone to use as many lights as possible. Public policy can certainly have a big effect on light pollution, but so can the policies of individual electric utility companies.

Click here to download an Excel spreadsheet with data for many more metropolitan areas in North America and other continents.