All posts by Ben

Dam the Mediterranean

I somehow have managed to live this long without learning about the Atlantropa proposal to dam the Strait of Gibraltar. Every description of the idea starts with the appeal of a tremendous electric power source. As demonstrated by the Lago Mare event, the Mediterranean operates as a net water sink, due to low rainfall and high insolation. Water flowing through the Strait of Gibraltar (and the Bosporus, Nile, etc.) are required in order to prevent the sea level from falling rapidly.

The strait of Gibraltar is only 14 kilometers across at its narrowest point, and about 300 meters deep on average. If the dam were largely built by simply dumping sand from the growing peninsula at its angle of repose, about 1.25 cubic kilometers of fill would be required, or about twice the volume of the largest Syncrude Tailings Dam. It would also be 10 times the material required by the Tarbela Dam in Pakistan, which cost $1.5 billion to build and generates 3.5 GW of electricity.

How much electricity would such a dam generate? I couldn’t find the answer anywhere on the net. I did find this uncited claim that a recent high school physics exam in Germany asked students to calculate the available wattage.

Well, I think I remember high school physics. Let’s give it a try.

According to this study the average net water deficit in the Mediterranean is between 500 and 700 mm per year. The surface area is 2.5 million square kilometers. That’s 1500 cubic kilometers per year, or 48,000 cubic meters per second. As a mass flow, we’re talking about 50,000 tons of seawater per second. Multiply by g=9.8 m/s^2, and you get 490 megawatts per meter. (The electric output is not much lower, because turbines run at >90% efficiency.)

What’s this per-meter business? Well, a hydroelectric dam requires a water level drop from one side to the other. If we want to make power, we can’t just dam the sea … we also have to let the sea level fall. To achieve the same, presumably profitable, ratio of fill volume to power output as the Tarbela dam, we would have to let sea level fall by 72 meters, generating 35 GW of power, or 300 TWh per year … about the total electricity consumption of Italy.

This is a little disappointing, for such an ambitious geoengineering project. The original Atlantropa proposal goes further, lowering sea level by 100 or even 200 meters, thus generating up to 1000 TWh per year … and substantially altering the coastlines. In 1920 I’m sure that was an unfathomable amount of power, but today it’s only a quarter of European electricity consumption … not even enough to entirely displace fossil fuel for electricity generation there.

However, if we cannot eliminate greenhouse gas emissions, there might be one other reason to build such a dam: rising sea levels. The Mediterranean coastline has seen substantial investment in waterfront property over the past, say, 3000 years. Those investments are threatened by the predicted rise of 0.5 to 2.0 meters over the next 100 years. If the construction cost is $15 billion (in proportion to Tarbela), perhaps existing property owners would be well advised to cover the cost of such a dam, and the few hundred megawatts of electric power generated would merely be icing on the cake.


I tried to check out my book from the library yesterday, but it didn’t work. The librarian explained, to my surprise, that you can’t check out paperbacks at the self-scan station … because they’re on the honor system! No checkout, no due date, no late fees. Read it, and please return when finished.

I like this library!

Half a League

I grew up hearing rumors that awesome grown-up boys had invented this thing called a “LAN party”. At the time, if you wanted to play a 3D-ish video game, you could play on a videogame console or on a computer. Consoles plugged into your TV set, and had up to 4 controllers, so that was the most people you could have playing at the same time, and they each only got one quadrant of a low-resolution, fuzzy CRT screen.

Computer games were more advanced. You could play someone else by connecting to them directly using a modem, and then later (if you had an internet connection) you could play with more than 2 people by connecting to the same server. This was nice, since you got a screen all to yourself, but it had some limitations. I spent several frustrating evenings getting crushed in Age of Empires, which I blamed on the handicap of my 56K dialup connection against my friends’ new cable modems.

More importantly, it’s hard to taunt your foes when they’re across town. It’s way more fun to play games together in person … but laptops were still in their infancy. If you wanted to play a multiplayer computer game in the same room together, you had to drag your mid-tower, CRT, and peripherals to the host’s house. You also needed everyone to have ethernet ports (which were rare on consumer PCs), cables, a hub (or switch if you were really fancy), etc. Then you needed to configure the server and client software to communicate. It was quite an undertaking, although I think I may have tried to host one or two in high school.

Yesterday evening was a friend’s bachelor party. I knew he was a fan of computer games, and generally a big nerd, but I did not realize just how deep it went.

The party started off at a BattleTech arcade. (I think they’re better known as “MechWarrior”.) The arcade had 8 simulated cockpits, which provided a reasonably immersive experience with the door closed. Before each round, we chose our virtual war machine, and then spent 5 or 10 minutes (and $4 each) shooting at each other in the virtual world. I was terrible, of course.

The simulators were fancy pieces of hardware, with 5 displays, two foot pedals, a joystick, throttle, and probably a dozen different buttons. The guts, however, were clearly dated, with software that was probably unchanged from the last MechWarrior PC game 13 years ago.

In a sense, this was just a dressed-up LAN party.

After BattleTech, the eight of us headed to two dudes’ house for the next phase of the party. I shouldn’t have been surprised, on arrival, to find the dining room tables covered in a mix of full-size desktop and gaming laptops. They had been assembled, I learned, for the express purpose of playing League of Legends all night, which is precisely what we did. We played the new Ultra Rapid Fire mode, in which the optimal strategy is basically to mash all the buttons over and over as fast as possible.

I was, of course, also terrible at this. It was good fun.

Freedom of $peech

The Supreme Court says that political donations are speech and cannot be limited?

Fine. We can work with that.

We can replace McCain-Feingold with a law that

  1. Tracks the amount of money donated to (or spent on behalf of) each campaign
  2. Immediately gives the same amount to the other side

There’s plenty of precedent for this sort of public election funding at the Federal level, in the form of the Presidential Election Campaign Fund. This fund was used to pay for every presidential election from the 1970s until 2008, and continues to fund the conventions. Candidates who took the federal funding had to agree not to accept any donations. This is essentially the same, except that you agree not to accept more than $X in donations (which happens to be the lesser of the amounts actually raised), and in return you get $Y from the fund (which happens to be the difference).

There’s nothing necessarily federal about this idea. It would work just as well at the state level, which would be a good place to try it out.

The 2012 presidential election is reported to have cost over $2 billion, so this might seem like it creates a large public expense, but

  1. $2 billion is nothing compared to the benefits of an even marginally less corrupt political system
  2. The law would actually cost much less than that! If every dollar I donate to my candidate also provides a dollar to the other side, then I have no incentive to donate. Campaign contributions would therefore fall to pretty much zero, even though they would not be restricted.

What are the odds of any bunch of politicians passing a law like that?


Last night as I got off the bus, Jack and Elaine were just entering a Glasgow science fiction and fantasy convention in the year 2018, in the middle of Charles Stross’s Halting State. Walking home from the bus stop I passed a gaggle of young, oddly dressed hipnerds wearing eyecatching entry badges on lanyards for the Emerald City Comic-Con, right across the street.

Comic-Con was already sold out, but as a local I already knew that the Convention Center itself is still public space during these conventions; only the meeting halls themselves are restricted to badge-holders. This afternoon I walked across the street, bought a sandwich, and found an inconspicuous bench on which to eat lunch and read about Action and Adventure at a convention, at a convention.

To blend in, I wore my company-issued propeller beanie. I don’t suppose I looked out of place, but (not entirely by chance) I had picked a bench in prime photo-shoot territory. I certainly didn’t look much like the people in movie-grade costumes, posing for reserved men in all black pointing kilos of glass.

This event (despite being a mere local chapter of the Real Thing in San Francisco) drew the biggest crowd I’ve ever seen at the convention center, and the most enthusiastically dressed. Inside the hall there were Captains America, a Wonder Woman, a very attractive bright blue Twi’Lek, a dude with some kind of homemade bright red electric-guitar-that-is-also-a-battle-axe, Batwoman (??) but with kind of a Darth Vader Mask (???), V, a blowup of My Little Pony the size of a Clydesdale, and of course hundreds or thousands more that I couldn’t possibly identify, let alone remember.

Heck, just to get in the door I had to walk past four different exquisitely detailed incarnations of Boba Fett.

I wish that chapter in Halting State was a little longer.

Cold Places

Today’s announcement about a newly discovered Oort cloud dwarf planet got me thinking about the statistics of the Oort cloud bodies. Very little is known about the Oort cloud, and most of the speculation regarding large bodies has been focused on objects from Saturn- to star-sized. We don’t even have the data to detect objects that size yet, let alone something smaller.

What if there were an Earth-sized body out in the Oort cloud? (I’m not the first to wonder.)

When we think of reasonable places for humans to live, the pickings are slim indeed. Humans need a hard surface to stand on, a large supply of water (which means Hydrogen; there’s plenty of oxygen if you have rocks), and controllable temperature and pressure. Those requirements alone rule out every planet other than Earth. There are some moons of Saturn and Jupiter that might work, but the surface gravity on e.g. Ganymede or Europa is awfully low, lower than Luna’s.

If you imagine that humans do better around 1 G, then an Oort cloud planet might be our best bet. Such a planet would likely never have been near a star, so all the primordial volatiles would still be frozen on the surface. Solar energy would be negligible.

Traditionally, these kinds of ice worlds are not seen as attractive for terraforming, but the frozen surface could be seen as an endless supply of fusion fuel. Fusion reactors could power human habitats, or even melt and boil the ice sheets into ocean and atmosphere. (Some care would be required to get the chemistry right!)

Travel to such a body would be difficult, but assuming it’s in the inner Oort Cloud (where most of the mass is), the trip would still be child’s play compared to reaching, say, Alpha Centauri. Depending on the chemical makeup, an ice world might be easier to terraform than Mars, which is mostly desert.

For all we know, there could be dozens of Earth-size worlds slowly circling in the inner Oort cloud, or there could be none. Could we build a telescope that could see them?