… if you live in the Red River or Minnesota River basins near anything that looks like water.

This morning, I heard a TV weather forecaster, speaking of the potential flooding in the Red River Valley in Minnesota/North Dakota. He said of this possibility, “Flooding is not a certainty.” But he’s wrong. It is a certainty. This raises issues related to your flood insurance if you live in the vicinity, to questions of long term human land use planning in the region and generally, and of the skeptic’s approach to life. And, this discussion applies not just to the Red River but to its tributaries as well as to other rivers in the vicinity, including the Mississippi and the Minnesota.

Most skeptics would probably look at something like the question of flooding from a statistical point of view and say that the chance of a flood happening (or not happening) in a given spring on a given river as an uncertainty. There is a probability distribution and one can never be truly “certain.” I suspect that there are those who consider themselves skeptics who will assume that I’m wrong to say that flooding on a certain river is a certainty, months in advance. There is a probability distribution and the distribution includes outcomes in which the flooding does not happen, and thus calling this a certainty is incorrect and a misuse of the term certainty, faulty reasoning, and so on and so forth bla bla bla.

But they would be wrong. There will be flooding, there will probably be widespread moderate flooding, and there is close to an even chance that we will see flooding as bad as it ever gets in at least some places.

A thought experiment: Imagine placing a dozen plastic cups full nearly to the brim with water on the edges of piece of furniture in a medium size room. Every cup is in reach of a short person, say, a baby who just learned to walk. Then release a baby in the room for ten minutes, unfettered, and ask yourself, “What is the probability of a given cup being spilled?”

There are a lot of ways to calculate that probability, including the ever-popular empirical method of carrying out the test dozens of times and recording the data. Be sure to use a fresh baby with each test, or at least every few tests, because there is a good chance its behavior will change over time. Regardless of how you do it, you can come up with a defend-able probability model and thus ask the original question. The probability of a given cup being spilled by the baby is probably less than 100%.

Now consider this alternative: Consider only cups which have been grabbed by the baby, who is in the process of turning the cup upside down, but in which the water still remains, but only because it has not spilled out yet. What is the probability that such a cup will not spill? Well, we could use the probability model we worked out before and apply this to the cup in question, but that would be wrong. That is what the weather forecaster was saying when he said “A flood is not a certainty.” In order for the baby to NOT spill the cup he has already grabbed some extraordinarily unlikely things would have to happen. We’re talking “all the O₂ molecules moving to the ceiling so we all suffocate by random chance” level of probability. In other words, that the cup will spill is certainty.

As we approach the deadline for purchase of flood insurance for those living in the Red River Valley of Minnesota, we see that the conditions for causing a flood are extreme. In order for there NOT to be a flood, several things would have to happen. There’s more than one way to do this, but the following scenario would work:

1) No more snow falls in the region, or at least, no more than a couple of inches of water (as snow);

2) The snow that exists melts at a rate such that the minimal amount that can melt melts per day (more or less) between a very short time from now and a fairly late date for there to be any snow retained in the snow pack. No bursts of melting can happen.

3) Ice and snow pack that flows down the river has to no get stuck on any turns, bridges, or berms.

Last year the flooding on the Red River was in the range of the historically most severe floods in the region. In 1969, 1897, 1997, and 2009 the river flooded maximally to 37.34, 39.10, 39.72 and 40.84 feet, respectively at Fargo. Last year’s flood did not reach any of those numbers, but did come to 36.95 feet and was considered a close call. The full plastic cup of river flooding here in the north is the snow pack, and the toddling grabby baby is the Spring melt. Last year the maximum snow pack in the region was on or about February 14th, most of which was melted away by March 20th. Last year at this time the snow pack in the region ran from about 18 to 22 inches with a few areas over 24 inches. This year the snow pack in the same area is between 20 and 39 inches.

The chances of a major flood, the flood to beat all floods, or at least, be in the top five or so floods, is probably about 50-50 so far, but there is a good chance that if our winter continues as it has been, that will go up. The ultimate severity of the flood depends on the pattern of melting this spring. The chances of moderate flooding is, according to the National Weather Service, greater than 80 percent. The chances of some flooding is not in question. There will be flooding.

The situation could be worse in the Minnesota River, where the snow pack is about the same as it was during the most recent severe flood (2003).

Homes in some of the floodplains in the Red River Valley need to be purchased and removed. It was not wise to have built them there to begin with, so it may be reasonable to not pay top dollar for them, depending on circumstances.

Sources of information:

A Young River in an Old Valley

Fargo Flood Home Page

Overview of flooding

Flooding predicted

Snow pack