The Problem With Models

The problem with models is not their accuracy, but their tendency to be uniform in their assumptions;

Computers, which are universally used in modeling, have to be programed with uniform assumptions. It’s in their nature. Thus we assume sea level rise will be x centimeters per year over the next few decades and get a renewed coastal map which predicts the coastline in, say, 2050. Most likely very accurately. The decadal rate is calculated from the century rate.

But uniformity was discredited when I was still a young man as the way of the world. The coast line in 2100 will very likely be as predicted by the models, but nature won’t make it so in incremental steps.

Like erosion, the changes will be accomplished in fits and starts. Storms will cause flooding, some of which will not be cleared by the simple passage of time and good weather. Gradually, probably, the new sea level will be recognized by you and I as higher than it had been.

Ice sheets, say in Greenland or Antarctica, currently sitting on land, will melt over a period of weeks or months, and slide into the sea, injecting massive amounts of water into the ocean essentially all at once, causing sudden sea level rise of small amounts. Whether this happens before 2050 or not is immaterial to the modeling assessment. As viewed by a Fair Witness, in the sense of Stranger In A Strange Land,  the sudden rise would not reflect a continuous centuries-long average. But for a computer to model such an effect over the passage of time, it has to. The way most climate scientists try to acknowledge this inconsistency in their methods is to refer to tipping points.

Tipping points are those places where the gradual change that has built up in some natural system causes some noticeable, usually thought of as related, change in another natural system. So, for example, the global temperature inching up will, presumably, eventually pass the tipping point of 32 degrees Fahrenheit near Greenland for an extended time, causing the sudden collapse of the Greenland ice sheet, and a relatively sudden sea level rise that eats most of Florida and the East Coast. So, despite knowing seas are going to rise (if you believe the forecasts) the actual rise is apt to take you unawares–perhaps even by surprIse.

What is most unfortunate about this conflict in the way scientific predictions are made and the way in which nature fulfills them is the illusion we get in our minds of non-threat from the one way of perceiving while the reality is quite different. When Florida floods, people are suddenly going to move. They will bring their guns with them. The same kind of thing is going to happen worldwide, though not necessarily simultaneously, and most of the world will not be so well armed. But it’s really the number of migrants that counts anyway.

Perhaps the worst thing about all this is the conflict presented to the scientists. They are so familiar with the phenomena that they may not realize how incongruous their strident warnings must seem to the less scientifically familiar general public. A scientist is not apt to understand how a politician can just ignore, let alone deny, sea level rise predictions.

It is not so much that the politician doesn’t believe the prediction (though that may also be true), it is that the rise in sea level is not perceivable. When it is, of course, it will be too late to do anything but try to deal with the flood of (armed) refugees as they present themselves on your borders. In other words, it will largely be those who occupy the high ground who will actually have to react to sea level rise, and they will have to do so on a short time scale, not a long one.

While the sea awaits the reality of having passed the tipping point, however, the scientist has to live knowing it is out there, but not yet reached. Faced with a public that is dubious, the scientist, like Cassandra, must live with knowing the future, but being ignored. To maintain some semblance of credibility with the public most scientists hedge their predictions in public forums, making their case less convincing, not more. Only after the tipping point is past will they all, with a single voice, shout, ”We tried to tell you so!”

Well, I’m trying to tell you.

Bookmark the permalink.

2 Responses to The Problem With Models

  1. Fred Drake says:

    The real problem with models is the accuracy of the assumptions programed into them. Do you recall Y2K?

    • George says:

      As I recall Y2K it was mainly about fear of the unknown and driven by the average man, not mostly by scientists. They didn’t know what was going to happen any more than we did.

      You may, or may not, think that climate change is coming, but whether you think it is or is not should, I think, rely more on what measurements tell us than on what someone with no scientific background, like almost any conservative commentator, says. The scientific community is virtually unanimous, despite conservative commentator’s claims to the contrary, in support of the impending nature of climate change.

Leave a Reply

Your email address will not be published. Required fields are marked *