Tuesday, August 28, 2007

Regional climate forecasts

Rasmus Benestad wrote a long article about the predicted impact of a hypothetical climate change on individual regions. In the context of the IPCC, this question is discussed in chapter 11 of the report of the first working group. Rasmus argues that it is very hard for the existing models to predict regional changes but he overwhelms us with a lot of unreliable information collected from random modelers at random places anyway.

The IPCC panel is an arbitrary conglomerate of local scientists whose work was sufficiently convenient to be included. There is not a single person in the world who has studied and checked the whole IPCC report and we shouldn't pretend that this particular conglomerate of statements and ideas is more valuable than it is. If you want to construct your own IPCC report, tear random pages from climate books on your bookshelves, throw away those that contradict a coming cataclysm, and ask your Marxist friend to write a summary of these pages. ;-)




There are some facts that are not stated too often even though I find them rather important. Let us start with a discussion of the importance and general expectations about the regional climate change:
  1. The impact of a conceivable climate change, regardless of its origin, on individual regions is way more important than its impact on the global average. Unfortunately, current climatology is obsessed with global averages - a few quantities that are both ill-defined as well as largely inconsequential. For example, the climate in Canada will be much more important for Canadians than the global averages.
  2. The fact that the existing climate models are not able to reliably predict regional climate change is a huge problem for these models. The Earth is not a zero-sum game. If a region is getting cooler, it doesn't mean that there must exist an equally large complementary region that is getting warmer. Dynamics of the atmosphere above individual regions and continents follows the same physical laws as the dynamics of the whole atmosphere of the Earth and if we don't understand the regional dynamics well, it is obvious that we can't understand the global dynamics well either which seems to be the case. This whole point turned out to be controversial for Alexander Ač and a more detailed discussion appears in the fast comments.
  3. Regional climate offers many more quantities where models may be verified against reality and where the actual drivers of various changes may be isolated more clearly than if we use global averages. That's why the regional climate and the weather is a much more appropriate context in which our ideas and equations about the climate should be verified and refined. A similar comment holds for short-term weather and climate dynamics. It seems that there are way too many climatologists these days who want to avoid such a thing: they really don't want to know the right answers. They only want to talk about a few global numbers exactly because it is easier to adjust a few numbers and pretend that we have already understood them. If we look at the regional data, it is pretty clear that we don't understand the dynamics well, not even in the long term.
  4. Whenever it is possible to divide the impact of climate change on various regions and forms of life into two groups, good and bad, it is extremely likely that about 50% of the influences will be good and 50% of the influences will be bad. Examples will be discussed later. This point turned out to be controversial for Prof Peter Shor who started a discussion in the fast comments.
  5. Individual countries and their scientific institutes should be primarily interested in the conceivable changes that will occur on their territory because climatology is, whether you like it or not, an applied science and the local climate is simply more relevant than the climate in distant regions. The fact that the temperature and other quantities depend on the location should also influence the rational behavior of countries and the international definition of justice.

Winners and losers: counting

As we have stated above, for any conceivable kind of climate change, either natural or man-made, we can find approximately 50% of regions that will benefit and 50% of regions that will lose. This zeroth order estimate is a consequence of a Z_2 symmetry relating cold and warm weather, a symmetry that holds pretty accurately in economics because the current life is relatively well adapted to the existing conditions. About 50% of subjects would prefer a warmer weather and 50% of subjects would prefer a cooler weather. Analogously, 50% of subjects would prefer increasing precipitation (Sahara) and 50% of subjects would prefer less precipitation (Londoners?). By subjects, I mean nations, people, species, regions, animals, corporations, or anything else: the vagueness is deliberate.

However, if you think for a little while, you may see that warming is actually better than cooling in average. Why? It's because we know that when the Earth was a snowball if it ever was, there wasn't too much life here. On the other hand, when Earth was covered by tropical forests and when it didn't tolerate any ice sheets, there was a lot of life here.

If we believe that the average 21st global temperature will be slightly below 15 Celsius degrees, it is sensible to talk about the interval between 0 Celsius degrees and 30 Celsius degrees. I find it obvious that 30 Celsius degrees is better for life than 0 Celsius degrees. We deliberately talk about the interval that is 30 Celsius degrees long because even this interval is compatible with life. In reality, people will only be talking about temperatures between 14 and 17 Celsius degrees - an interval that is shorter by an order of magnitude - throughout the whole century.

Moreover, the impact around a stationary points goes as the second power which is why one order of magnitude mentioned previously translates to two orders of magnitude on the impact scale. Even if the 3 Celsius degree change of the temperature occurred instantly, its impact on life could be quantified as roughly "1 percent of life", assuming an appropriate measure. Because this change is distributed over a century, its impact will be much smaller.

Let us summarize: slightly more than 50% of the regions or subjects will benefit and slightly less than 50% of the regions or subjects will lose. Whoever wants to insert the assumption that much more than 50% of the regions will lose - an assumption that others are supposed to treat as an unquestionable dogma - is simply acting irrationally. The relative effect of the "good news" may be increased further as long as we can plan and adapt.

Winners and losers: examples

Fine. So who will lose and who will win? Rasmus admits that no one knows reliable enough answers. Their usual self-confidence about the global predictions goes away. They are only self-confident about the global predictions because it is much easier for anyone to fool himself into thinking that he understands something as soon as his only task is to "understand" one number only. When you see millions of diverse numbers with patterns that make some sense but you don't know how to derive them, it is much harder to fool yourself.

Because of this comment, all the examples that will follow should mostly be viewed as thought experiments that may turn out to be right or wrong.

So who are the winners? I find it extremely likely that if we assume that the average climate will be getting warmer, the civilized nations located in the cold regions are going to be clear winners. Large regions of Russia, Canada, and Scandinavia may become fertile. The plants grown in these currently frozen or cool regions will be able to feed millions of people who are not yet born. The Arctic region will offer its resources that will be primarily available to the nearby countries. New traffic routes may appear.

I would say that whoever lives in Russia, Canada, and Scandinavia and wants to fight against global warming because he or she thinks that it will harm his country is simply acting irrationally. These obvious thoughts make the hysterical behavior of Stephane Dion who attacks poor silent boy Stephen Harper even more illogical and alarming. (Dion's party seems to be really obnoxious.) The worst thing that can happen to Canada is that it will look much like the U.S. today. Note that the territory of Canada is greater than the U.S. territory. Why does Canada support 10 times less people than the U.S.? Surely it could have something to do with the weather, right?

We're not guaranteed that Siberia, Norway, or Northern Canada will become warmer and more friendly. But I think that common sense dictates that it is likely.

Winners and losers: rain and deserts

What will probably influence the friendliness of various regions much more than temperature will be precipitation. We have been to Sahara a few years ago. It's been fun but it's not a great place to live - except for the hotel located many miles from the borders of Sahara that had several swimming pools in it. ;-) The negative influence of Sahara is one of the main reasons why Africa is such a poor continent. I fully agree with Freeman Dyson's "heresy" that a wet Sahara is something that good people should prefer. And a wet Sahara is more important than some global temperature averages moving by half a degree in either direction.

I've seen a paper claiming that Sahara will get wet if we extrapolate the current warming trends. I don't know whether these predictions are true or whether the opposite predictions are closer to reality.

But I find this question much more important than the questions about various one-degree global warmings. And I find it simply stunning that Rasmus' 24-kilobyte-long article about regional climate projections doesn't contain the words "Sahara" or "desert".



Why? Well, many people expect that China may become the world's #1 superpower in 50 years. Great changes in Africa could also hypothetically promote Africa to the leading region even though I guess that Africa would have to allow remaining races to participate in their economies. At any rate, the extent of Sahara used to be one of the main big questions about the climate and it is completely irrational that climate scientists such as Rasmus no longer think in this way. Or does he know that there is Sahara in Africa? Maybe he finds this fact unimportant for his climate models.

Rasmus' behavior is a part of a more general pattern.

The proponents of the global warming ideology want to talk about the climate at the centennial scale but what they really care about is how to influence politics in the very short run. If they were thinking about the long term political questions, they would know that their assumptions about the most prosperous regions of the world may become invalid by the end of the century.

In Europe, winters will get milder and wetter in the North and summers will get hotter in the South. I don't see any spectacular changes that could be classified as "very good" or "very bad". If there will be changes in Asia, they will be generally irrelevant. What can happen is that the type of life that exists in certain regions will be moved to the North or the mountains. Because one expects small changes distributed over a century and because the right answers are not really known, I don't think that this is anything that any blogger should analyze in more than two sentences.

Finally, some predictions argue that Southwest portions of the U.S. may get dry. It's very questionable whether they are trustworthy in any way but this scenario will nevertheless play a key role in the last section of this essay.

Non-local quantities and sea level

I emphasize that one should focus on the regional behavior of the climate because if there exists not a single location where climate change is expected to be harmful locally, it follows that climate change can't be harmful globally either.

Is the previous sentence robust? I think it is. There are no non-local influences that could damage your society even if your future local climate looks fine to everyone. One might say that the sea level is such a global quantity that matters. Yes, it is except that we kind of know that it won't change in any significant way. It's been increasing by 2 millimeters per year in the first part of the 20th century and 1.5 millimeters per year in the second part. It's actually slowing down. But even if the rate triples, it remains insignificant.

But even if you want to consider speculative scenarios of a huge sea level rise, it is still true that it will have both negative and positive consequences. For example, if the sea level rises by 100 meters as James Hansen is going to predict in his new paper, the Czech tourists won't have to travel as far to the sea for vacations as they have to do today. They will save gasoline, too. More seriously, even though some houses would have to be sacrificed if the sea decided to play Hansen's game, the houses that would survive and get closer to the sea would become more valuable.

Anyway, the set of the threatened buildings doesn't contain too many of them that were supposed to be serving people for centuries. It is irrational to impose international regulation just in order to prevent the natural redistribution of wealth. It is equally irrational to be afraid that a small percentage of people won't be able to run 10 meters away from ocean in 100 years.

Non-uniform climate change: justice

Fine. So believe me for a while that regional changes are much more important for the people who live there than the changes of the global averages. Imagine that such changes will occur. How should the world react?

A priori, I think that no lucky nation has any obligation to compensate another nation whose climate became less friendly. But we have the United Nations and various other tools of international compassion. Let's imagine that the whole mankind is one huge family. How should we react if someone benefits and someone else loses?

Analogous thought experiments could be performed in many situations but let us consider a specific situation. Sahara will become a fertile land and some additional states in the U.S. Southwest will become a desert. Should Africa pay compensations to the U.S.? I don't know whether it should but what is clear is that it won't. No serious person will even ask for them. Even if Sahara gets fertile, it won't be richer than America for a long time. Moreover, a fertile land in Sahara is likely to be grabbed by foreign investors. ;-)

Now imagine that a poor country will become a desert and richer countries will relatively benefit. In this case, it is much easier to imagine that the poor country could be compensated, financially or territorially. All kinds of emotions, laws, conventions, and political arguments would play role if such things were ever being decided. At the end, such a compensation would be just an addition to the existing aid to the third world.

Another thought experiment: imagine that the mankind can actually prevent a particular climate change whose regional impacts are more or less known or - analogously - it can artifically initiate such a change by technological means. Such a decision would create winners as well as losers. Should it be done? A rational decision at the level of the U.N. would clearly be a decision whose net benefits significantly exceed the net costs. If such a decision - to prevent or initiate a climate change - can be made and justified, its asymmetric impact on various countries may be compensated according to an international treaty.

There are many important scientific, legal, ethical, and especially economical questions associated with all these scenarios. My text was only meant to encourage people to think about these actual questions. In the current atmosphere controlled by intellectually limited, fanatical people who only talk about one global quantity, a quantity that is decorated by quasi-religious connotations and irrationally correlated with all of the evil in this world (the global mean temperature or climate sensitivity), it shouldn't be too surprising that a serious discussion about our attitude to climate change hasn't yet started.

Challenge: propose a fair system of rules

I encourage everyone to try to be more specific about what they consider a fair international reaction to various regional forecasts or actual regional climate changes that have either occurred, or are likely to occur, or can be artificially initiated.

And that's the memo.

RealClimate.ORG - previous replies. I haven't responded to two previous texts on that blog. One of them attacked FoxNews, Bush's response to Katrina, and Schwartz's calculation of the climate sensitivity (where James Annan's imported arbitrary comment was the closest thing to an argument: RealClimate.ORG wasn't able to manufacture a better criticism).

The other RealClimate.ORG article asked the journalists to call Schmidt and Mann because they have successfully predicted that the Atlantic circulation wouldn't stop. Congratulations to their extraordinary prediction! ;-) The previous text on RealClimate.ORG that deserved at least some reaction was about tipping points, sweet spots, and model ensembles.

1 comment:

  1. Thinking about climate change on a local scale is a wise thing. Climate is influenced by many factors in different areas. In some areas is climate cooling, in others, climate warming.
    Perhaps the areas influenced depend on their position relating the ocean / sea near them. There have been several climate changes in the past century, as described at http://www.1ocean-1climate.com. All of them have been influenced mainly bu the same factor: what happenbed tot the oceans near them.

    ReplyDelete