Flight Maps!

One of my main projects this spring involves conducting an analysis of airport traffic to see whether it makes a difference to have a rail connection to the airport. It's really fun and I'm enjoying the chance to construct and analyze my own personal dataset.

A byproduct of this is that the data I'm using to find out which airports are the busiest (the BTS Origin and Destination Survey) comes in origin-destination pairs, which I'm compiling to get the total number of airport users. But first, it offers me the chance to make my first maps with data!!!! I thought these were pretty cool, and they give me a chance to demonstrate my newfound prowress with Python graphics. First, an image of US flights in 2008:

This shows the top 500 or so actual trips made by planes. The width of the lines is proportional to the number of passengers traveling between two given airports. You can see the importance of hubs like Atlanta, Denver, and Dallas-Fort Worth, which are all really busy even though the cities they serve aren't that large. I thought it was neat to see all the airlines in one place, as opposed to the in-flight magazines which only show hubs for one or two airlines. When flying on Northwest, you would get the impression that MSP, Detroit, and Memphis were the largest airports in the country, and while that obviously wasn't the case, it was sort of hard to compare with airports serving other airlines. Still, it's pretty much what you might expect if you travel a fair amount.

The other map I made was a little harder to predict:

This one shows origin and destination pairs--where people are actually traveling to and from, not counting transfers. It only has the top 350 or so pairs, and I made a few changes to better illustrate some trends, mostly consolidating airports in cities with more than one, like New York and Chicago. You can see how different it is from the segments in the other map: it turns out that the main trend in US air travel is people going from the East Coast to Florida and the Southwest. It turns out that some of the busiest airports, like Atlanta, Dallas, and even Chicago to an extent, get most of their travel from people flying through them on their way to somewhere else. And whereas the flights give a fairly consistent net over most of the country (except Montana), the destinations are much more focused on the major centers of population and tourism.

Anyway, hopefully this info is interesting to other people. I know I had been curious about where people actually travel, and whether Atlanta really is that huge a destination. Now we know it's not, it's just the gateway to Florida (and most other places).

Whining about wine

When you get a chance, check out the "Environment" section of this month's National Geographic.  It's about the carbon footprint of shipping wine from major production regions to various major US cities, and you can see it here.  It's worth looking at for a few reasons:

1). The main point, which it makes dramatically, is transporting goods by truck, which is the most common way by far to move things in the US, is incredibly inefficient.  Moving a bottle of wine from Napa to New York by truck releases 4.4 pounds of CO2 versus just 0.9 to ship a bottle from Sydney to NYC.  It's roughly equivalent in terms of carbon to drink a bottle of wine trucked to NYC from upstate New York as it is to drink one shipped from Bordeaux.  

2). Also interesting is the backlash on some wine blogs, particularly those devoted to producing upscale wines in idyllic parts of California.  I googled the article to try to link to the graphic (they don't have it online! n00bs...), and turned up a bunch of  these blogs ranting about it, plus some very probing comments on the site of the guy who made it.  One of them even has a new graph that adds in the carbon emitted by sending a package of National Geographic magazines across the country :-).

It looks like they have some points--he wasn't entirely fair to California, assuming for example that all Australian wine started in Sydney instead of the countryside, and possibly assuming that wine shipped to the ports closest to its destination instead of checking where it actually goes.  I see their concern that now in the eyes of hundreds of thousands of slightly-informed National Geographic readers Californian wine is Bad For the Environment, while in real life it's less clear.  So I'm going to reconsider my initial impulse to chide them about pouncing on environmentalism the minute it touches fancy small-batch wine from California.  Mostly, though, we should stop flipping out about eating (or drinking, apparently) local for energy reasons, at least until we know what "local" and "far away" mean.  This was brilliantly expressed last year in Time.

I liked the original graphic mostly because of how it made the point that the distance food travels doesn't matter nearly as much as the way it travels.    If we really wanted to help the environment through infrastructure spending, and make ourselves much better set up for future prosperity, we would be seriously thinking about improving our freight rail connections. It's sort of crazy that we've let trucking overtake rail transport in this country, through a combination of auto/asphalt industry special interests and government policies that required railroads to run money-losing passenger trains (this makes me a little nervous about $40 billion high-speed passenger rail projects in California).  They had trains in the 40's that could ship food cross-country with less energy than we use today.  Families used to ship their refrigerators to their vacation homes for the summer, just because it was so cheap!  I really like the idea of taking a fast train from Minneapolis to Chicago, but I have to wonder if that option makes anywhere near the economic and environmental sense that investing the money in freight railroads would.

Specter and Swine Flu

Good reading:

On what to expect from the flu: http://www.nytimes.com/2009/04/28/opinion/28barry.html

And about the impact Specter's switch might have:
http://www.fivethirtyeight.com/2009/04/specters-switch-more-insult-than-injury.html
http://tpmdc.talkingpointsmemo.com/2009/04/why-specter-did-it----and-had-to-do-it.php

Drugs

Our topic in Latin American Politics this week. A bunch of recent developments, basically pointing to increased tolerance and possibly decriminalization of marijuana in the US, but continuing the main thrust of the War on Drugs in Latin America:

Marijuana now essentially legal in CA.
You can even go to school to grow it.
But our policy in Mexico is still more of the same, even under Obama..
Finally, a really detailed description of the whole process as of last summer.

Best line of the whole thing is from the third article: '“By supporting the Calderon administration in their fight with the cartels, we’re keeping it from becoming a fight on U.S. soil,” [Arizona Attorney General Terry Goddard] explains. It’s not hard to see why someone like [Mexican journalist Lydia] Cacho would find that rationale unappealing."

China topic, with normal cringe-inducing US behavior

I hope we don't embarrass ourselves again....
http://www.theatlantic.com/doc/200904u/shanghai-expo-2010

Doga

I'll refrain from comment.
http://www.nytimes.com/2009/04/09/fashion/09fitness.html?em

Change in style: link-dumping

So I am well ensconced into normal school life now, learning about Latin America, the US, and Probability. Part of normal life means not enough time to condense my thoughts into the beautiful prose that you were reading over the course of last semester. Instead, I think I'm going to start using this as a more traditional blog, eg a way for me to disseminate the cool/interesting/important things that I read without having to interrupt all my friends' email lives.

Highlights from my google reader today:

1. This seems sort of disturbing. Did you know that Wyoming and Nevada are huge tax shelters, and even more secretive than Bermuda or Switzerland? Last week's Economist

2. Visualizations! This person's coming to Swat next week :-D