To continue my harping on the necessary relationship of science to a testable reality, I will consider Governor Bobby Jindal's insistence on building barrier islands to prevent the oil from the BP blowout from reaching Louisiana's shores. It apparently isn't working so well.
Jindal had a bright idea. He didn't bother to test it against some of the rules of thumb I've suggested, and it's not working because it fails the reality test.
It's a very easy and nice thought: Oil in Gulf. Sand islands absorb oil before it reaches the marshes along the shore. Dig up islands after oil problem abates and dispose of oily sand. Marshes remain clean.
The problem is that water moves. Not only that, but it's heavy and capable of moving other stuff, like sand. If Jindal had simply thought of other instances, like the building of jettys and other rock structures designed to change the movement of water so as to retain sand on beaches, he might have come up with the right answer. In many cases, those jettys have actually resulted in the loss of sand.
Or he could have built a sand mountain on a beach, below the tide line, and then watched what the water does to it.
The people who watch this sort of thing for a living said the sand berms would wash away, or, if they somehow stayed in place, they would prevent water as well as oil from reaching the marshes, which also would damage them. Plus it would cost lots of money, but hey, Jindal is Republican, and he was asking for the money from the feds, so that wasn't a problem in today's political world.
And now the sand berms that have been built are washing away (photos here). There is a discoverable reality, and science provides the way to find it.