Saturday, November 26, 2011

The Odds on Global Warming

I have mentioned previously that I am a big fan of Andrew Tobias. Recently, he posted on his blog the claim that climate "deniers" were massively funded by corporate interests.

 I sent him the following note.

"You said, "The connection between the climate-change deniers and the Republican budget enthusiasts is that both are massively funded by corporate interests, now entirely unleashed by the Republican Supreme Court and its Citizens United decision.  That’s the one that stacks the political system even more heavily in favor of the rich and powerful, and allows their influence to be hidden."

"For the record, I am a big fan and we are on the same side politically. Scientifically we differ, and this thing about climate-change deniers being funded by corporate interests is simply not true as far as I can tell. People point to XOM funding someone for 25 Million dollars or so. Compare that to the forces of climate change "proponents" who have received Billions upon billions of dollars. As far as I can tell, most of the major skeptics (please don't call them deniers and I won't call the universities elitist), are private citizens doing their research with their own money and minor contributions from supporters.

"Am I biased? Yes. I work for, but am in no way representing Chevron with this opinion. I am also a citizen of this planet and want our leaders to make good decisions based on reliable information. Most of the AGW data does not meet that standard."

He responded by asking me what probability I assign to the possibility that the scientific community has it right. My response follows:

Hi Andy,

You asked me  “What probability do you assign to the possibility the broad scientific community has it substantially right?”

I am an earth scientist with an MBA. I have been a member of Texaco’s exploration risk team and I do probabilities for a living. So if you’re looking for one number answer like x%, it’s not happening. Also, I normally do not ever assign 100% to anything, but I will on some items just to not look too pedantic.

The key word in your question is “it”. There are a lot of “its” in the field of climate science and what to do about it, so I will just go through those as a chain of events that has to happen for “it”. As you read down the list, remember that the links are dependent, so if the previous link does not come to pass or is not true, the ones following become irrelevant.

One more quick note. I don’t think the probabilities below are “right.” They represent states of knowledge—mine and the climate scientists.  I am pretty comfortable directionally, so if a number is at 0% it’s pretty low, and if it’s 75% it’s pretty high.

Cumulative Probability
The earth is warming.
Probably. It appears that there was a very low temperature period from about 1600 to 1800. It’s not surprising that we are warming from that. The actual data are quite ambiguous. The temperature increases shown on all the graphs can be arrived at only by applying adjustments to the thermometer data. These adjustments are in many cases larger than the temperature anomaly. The signal to noise ratio is therefore quite low. Nonetheless, although I have a lot of questions about the data, on a macro level it seems reasonable. Let’s call it
CO2 causes the earth to warm as a primary effect. Call it
Increase in CO2 causes the earth to heat up as a primary impact. Call it
I think that the above 3 points define the scientific consensus. Going past this there is much less consensus.

Increase in CO2 is the primary cause of the observed temperature increase.
From ice core data, historically, the relationship is reversed. Temperature increase leads CO2 increase by about 800 years. Climate scientists claim that the causality reversed about 50 years ago.

Their claim is something like, we have eliminated all other known causes of temperature increase, so that leaves CO2 as the culprit. Their models leave out a number of possible drivers such as deep ocean currents, land use changes, and solar activity.

I think there is too much that we do not understand about the climate and impacts on the climate for this line of reasoning to yield any kind of certainty. Has CO2 had a positive impact on temperature. Yes. Is it the primary cause (more than 50% of the change)? Not proven, but not totally unreasonable.

If we do nothing to curtail CO2 emissions, temperature in 2100 will be at least 2 deg C higher due to CO2.

A doubling of atmospheric CO2 causes about 1.2 deg C as a primary effect. IPCC projects about a doubling through 2100. So, based purely on the primary impacts of CO2, a number closer to 1 deg C would be appropriate.

The CO2 in the atmosphere has doubled since 1900, yet the temperature has risen approximately 0.7 deg  C, if you believe it.

Getting to a higher number requires computer models with positive feedback impacts that have not been observed in nature. I have built complex economic computer models over long time periods. They tend to be extremely sensitive to assumptions in the model. Even extremely small errors or inconsistencies can lead to huge anomalies because of the large time period. On top of that, the researchers have not made the models available for auditing. There is a near 100% probability that there are errors in the models of unknown impact. Going beyond 2 deg C is not high certainty. The role of the water cycle and cloud formation are still very much poorly known.

The models do not backcast correctly and they have not demonstrated ability to forecast. The models are wrong.

They rely on extremely complex computer models with huge assumptions about how variables interact, that have not been audited. Possible, but until those models are opened and audited by critical assessors, I can’t put much faith in them. I’ll give this one a VERY generous
This is the end of the climate science. The rest gets into economics and social sciences. At this point we're looking at a 10% probability that the climate science is "right."

So if all the above comes to pass, what will be the net effect on the world? There is a mixed bag. There is some possibility that the earth will reach a tipping point leading to global catastrophe. There is no consensus that catastrophe will ensue, so this is not part of “it.”

Consider more moderate scenarios. In general, heat is better than cold for people. Higher CO2 is better than low CO2 for plants. Low lying areas could flood. Weather patterns could get more severe (although this is not a slam dunk either). All else being equal, if there is no economic growth between now and 2100, I think the overall consequences would be negative to mankind. However there will be economic growth. Today’s poor countries will not be so poor 100 years from now. People will not be engulfed in their beds by the melting glaciers. Earth’s climate has always changed, always. Cities and countries grow prosper and go away due to climate. There is a lot of time to respond to slowly developing trends by adapting. Maybe marginally better, maybe worse.

We should change economic incentives to encourage less carbon emissions.
By the IPCC’s own analyses, if we were to implement Kyoto, it would delay the warming in 2100 to 2106. So let’s just forget that, except possibly as a small incremental step.

I have seen the argument that investing in the lower carbon alternatives is smart economically. Well, I have a lot of faith in the greed of my fellow citizens. If it were a good investment it would be happening without massive government incentives.

You understand opportunity cost, so even if green tech were not the worst investment that could be made, it would slow overall growth of the economy. As has already been discussed, we will need the economic growth to help us adapt to whatever the world throws at us. Furthermore, think of all the bad stuff happening in the world today. Non-potable water in the developing world, starvation, torture, slaughter, and mutilation in Africa. In a world with limited resources to spend, how many people are you willing to let die, in order to delay global warming by 6 years in 2100? Should spending to lower CO2 be a priority? At best

To reiterate, the numbers above are not correct. Ultimately, either each of the events will happen or not. The thing to think about here is that in order for AGW to be correct, there is a long string of events, each of which must happen. It does not take many items for the odds to become ridiculously small. For example if the odds of events come in at 90%, the chances of 10 of them happening in a row is around 32%. As the odds of an individual event decrease, the cumulative odds plummet as seen in the graph.

Finally, let’s briefly address the precautionary principle. The idea is if there is even a small chance that something bad or unacceptable will come to pass, we should do something to protect ourselves. This is always dependent on four things: consequences of doing nothing, likelihood that doing nothing will result in the consequences, cost of the intervention, and improvement after the intervention.

Consequences are questionable, but there is some possibility that the consequences of doing nothing will be large. The likelihood that consequences will be catastrophic are vanishingly small. There is no consensus on catastrophe, only imagination.

Cost [of prevention] can run into the billions or trillions including the opportunity costs. And there is no guarantee that it will improve anything. It might not even improve anything, if we reduce CO2 and the climate change is due to other factors.

So I am not dismissing the possibility that carbon dioxide emissions are having an impact on global warming or climate. They probably are. I do not believe the whole tipping point thing, and the fearmongers who claim that the world’s climate is about to go off the rails are thoroughly unconvincing. Everything in between is in the realm of needing more data before taking drastic action.

All in all, the information in my opinion is not reliable enough to commit huge amounts of money into a social policy that may have no net effect.

I wrote a much longer analysis on my blog some time ago. Here’s the link if you’re interested.

Tony Kenck