Category 6 has moved! See the latest from Dr. Jeff Masters and Bob Henson here.
Redux: Arrogance, Models, Responsibility
Redux: Arrogance, Models, Responsibility
I am motivated by the remarks to my last blog about the use of models that cannot predict next month's temperature to develop policy. The basic idea was that if the model cannot predict next month's temperature, then why should we give it any credibility predicting next century's temperature? This is a comment or criticism that is often stated in the climate change discourse.
Give this a run: Imagine that you live in a place with reliable rain. At the beginning of each year, you have some confidence that it will rain over the course of the year. You do not know in advance exactly which days it will rain, but you are confident that the rain will come. Suppose further, that you have a very good weather service, which can predict rain 2 days in advance. You decide to build a cistern, or a pond. I like the cistern better, because we can imagine closing it so that we don't have to worry about the loss from evaporation. If on New Year's you wanted to predict the filling of your cistern, then you would find yourself in the following situation. In any two day period you could predict with high accuracy that water would accumulate in the cistern; you have excellent 2 day weather forecasts. Over the course of the year, you can expect your cistern to fill up; you have reliable rain over the course of a year. In between two days and a year, your forecasts are not so good; you do not know beyond 2 days exactly when it will rain.
Therefore, you have a model with good short-term forecasts, reliable long-term forecasts, and that in the intermediate time an exact prediction is not so good. Still this is a useful model, and a model that has been used by farmers for thousands of years.
From a scientific point of view, this brings forward the notion of separation of scales. We often are able to know with some precision about the short-term and the long-term. In the short term, we have deterministic predictability. In the long term we know the nature of the balance that must be maintained from the fundamental physics--in the case of climate, the conservation of energy. In between, we are more reliant on probabilistic forecasts, and it takes some skill to know how to handle the errors in a probabilistic forecast.
It is equally as interesting to me how people use deterministic and probabilistic knowledge in decision making. There are many situations where knowledge does not rise to the top of the decision making process. An example that is often used is buying a lottery ticket--by a vast probability you are not going to win the lottery; you are giving your money away. This example can be dismissed as low-risk/high-reward or as simple entertainment. But people make decisions all the time based on their belief system or what they want to happen. I was recently introduced to the literature of how people make decisions by one of my students. (There is a literature on everything! (Perhaps too many academics in the world? Arrogant academics?)) One book this student referenced was, Dawes, R. M. (1988), Rational choice in an uncertain world (Fort Worth, Harcourt Brace College Publishers). Dawes discusses irrational decision making, and why we do it. If you Google search on the title you can find whole worlds of critical decision making.
It is my evaluation that the information from climate models is reliable enough that we need to limit our emissions of carbon dioxide. I think that the long-term costs will be much less if we act today. On the other hand, I recognize that the environmental impact of reducing carbon dioxide emissions will not be immediately realized. That does not lead me to the decision that nothing should be done. It leads me to the decision that there are a set of problems that exist in the near term, that require our attention. This is a separation of scales. If you choose to focus on only the short-term or the long-term, then you are faced with a set of seeming contradictions. These can be used to justify inaction; they can be used to discredit the other camp. As I said in the last blog ... inaction, seems to me (belief system here), as avoiding our responsibility. We have knowledge.
The views of the author are his/her own and do not necessarily represent the position of The Weather Company or its parent, IBM.