For my most recent Dear Science column, a reader asked:

Dear Science,

Is reading The Stranger online actually any greener than reading the printed-in-Yakima hard copy? Doesn't it take a shitload of electricity to run the servers and keep them cool? How would one even figure out how to compare the carbon footprint of, say, going to the coffee shop once a week and reading the print version versus reading it online, as well as checking in with Slog on a regular basis? Folks talk about the internet as being green, but part of me suspects that all it does is put its pollution somewhere out of sight.

It was time to roll up my sleeves and do some real, primary, research on the question. (If you just want the answer, go read the column for the condensed answer.) Allow me to show my work.

For print on paper, I assumed the two major carbon impacts would be the manufacturing of the paper itself, and the physical distribution of the printed copies.

The EPA maintains a fantastic online calculator intended to help manufactures figure out ways of reducing their carbon impact by using recycled materials. Newsprint is one of the categories. Kevin (at The Stranger) was kind enough to tell me the amount of recycled paper (pre- and post-) consumer: 40% pre-consumer, 40% post-consumer recycled and 20% pulp from freshly cut down trees. As my column notes, only the use of post-consumer recycled paper reduces the carbon impact. Both pre-consumer recycled paper and pulp require the cutting down of trees. As I noted in another column, trees actively sequester carbon. Cutting them down, if you're accounting properly, has a really nastly net impact on the atmosphere.

I independently calculated the tons of paper needed each week by weighing a single copy (150 grams) and multiplying by the total circulation. The total weekly weight was about 13,500 kg (about 30,000 pounds) of paper. Kevin confirmed this was about right. For the mix of recycled paper used, that worked out to 5.2 metric tons of carbon equivalent (MTCE) released into the atmosphere each week for just the paper. If The Stranger used (much more expensive) 100% post-consumer recycled paper, this would drop to a mere 0.30 MTCE per week—17.5 fold less than currently emitted.

For the distribution, I first started with the semi-truck from Yakima to Seattle—140 miles at about 5 miles per gallon, or about 28 gallons of diesel fuel consumed per trip. Burning a gallon of diesel fuel releases about 2.8 kg of carbon into the atmosphere, so 28 gallons is about 0.08 MTCE emitted.

The in-town distribution consumes another 76 gallons of gasoline per week. Burning gasoline releases about 2.4 kg of carbon per gallon, making the total emissions from the in-town distribution 0.19 MTCE per week.

The total for the physical delivery of the paper? 0.26 MTCE per week. The overall total (paper + distribution) carbon impact for the paper each week worked out to about 5.5 MTCE per week, almost all of which coming from the newsprint itself. Divide by the current circulation of The Stranger, and that works out to about 71g of carbon equivalent per printed paper: 67.4 g for the paper itself, 3.4g for distribution.

Were my assumptions valid? I'm ignoring the energy costs of running the printing presses, figuring they are probably predominantly powered by non-carbon emitting hydroelectric and wind power. I'm also ignoring the carbon impact of manufacturing the soy-based ink, assuming it's a small contributor. That might be dangerous, as farms are massive contributors to atmospheric carbon emissions. I couldn't find a good source for the MTCE per gallon of soy-based ink. If anyone knows, I'll be glad to incorporate it into my analysis.

For online, we have a few things to consider: how much energy does it take to serve, deliver and read the content.

A nicely done study figured it takes about 12.5kWh per gigabyte of data served and delivered on the internet. On average in the US, generating one kWh of electricity emits 0.00012 MTCE into the atmosphere.

Anthony was kind enough to provide me with hard numbers for the number of visitors and views on The Stranger's website in a week. I measured a bunch of pages to calculate the average pageview on The Stranger's website weighs in at about one megabyte. The total weekly carbon imact of serving and delivering the content on The Stranger's website is about 1.7 MTCE; interestingly, that's more than the weekly carbon impact of distributing the physical paper. Per unique visitor, that works out to 9.4 grams of carbon equivalent each just on the delivery.

The carbon impact of reading things on the internet really is dependent upon which computer you are using—and how many watts the computer uses. A relatively modern laptop, consuming 45 watts, emits 5.4 grams of carbon equivalent per hour to operate. A big honking desktop PC, weighing in at 250 watts, emits 30g per hour.

I have no clue how many hours a week people spend reading and commenting on The Stranger's website, nor the mixture of computers. So, I cannot make an honest estimate of the total carbon impact of the online presence of the paper. I can tell you about 11.4 hours of online reading on a laptop, per week, has about the same carbon impact as a single paper copy. Reading on a desktop PC? Only two hours equals the carbon impact of the paper.

Want to calculate for your own PC? Here's the formula:

(Watts of your PC) * 0.00012 * 1000 = your grams of carbon emitted per hour.

For the number of hours until reading online on your PC equals the carbon impact of a single paper:

61.6 / (grams of carbon per hour from your PC) = number of hours.

How are my assumptions here? I'm not considering the carbon impact of manufacturing the laptops and computers. But, I'm not considering the carbon impact of manufacturing the roads, trucks either.

And, as I end my column noting, reading isn't even close to your biggest carbon impact. A single cheeseburger emits the equivalent of a kilogram of carbon. Driving the average car on the road today one mile emits more carbon equivalent into the atmosphere than a single paper.