Unprecidented warmth in Arctic

I TA'ed undergraduate statistics while I was at graduate school at the University of Texas. I have a friend that is a math education professor at Georgia, who teaches statistics and does research on how to teach math. The main thrust of her research is how people learn, and why people have a hard time understanding mathematical concepts. One area that the majority fail to understand even at the university level is the statistics of small probabilities. We had a case recently with the tesla fires. The probability of a car in a given year catching on fire is slightly less than 0.1%, 172,500 in 2012, out of roughly 250 million vehicles.

Highway vehicle fires

Of those fires less than 1% result in major injuries or death (the crash is more likely to injure or kill you than the fire). We have had 3 crashes out of the approximately 20,000 teslas S's on the road and no major injuries or death. Now because of sample size, the cars being newer, and all being on the road less than a year, we can't say with confidence that a Tesla S is less likely to catch fire than a gasoline car, but the statistics are in that direction. Say that we divide by age, we should get a probability of about half of gasoline cars for fires, and lower for injuries and death. The media repeating things about the fires though, leaves the impression that you are far more likely to get injured or die in a BEV fire than a gasoline car, and that is simply against the probabilities.

Yes the tail at 2 sigma is quite plausable, that is why we typically go out to at least 3 sigma for which gives us 99.7% confidence. Spencer's numbers are clearly within this range.

In manufacturing quality control, which I have worked in, and consulted on, six sigma is the goal. Toyota tries for six sigma manufacturing in their cars. The odds of a tesla catching fire is much less likely than 3 sigma, but it does happen, as we see from the news reports. It is quite plausable.

I will not even go into how this confidence range for ghg sensitives gets created, but we have a number of papers that say that over 5 (degrees Kelvin for each doubling of carbon dioxide equivalent concentration for direct and fast feedback response) is implausible because we would have had many more ice ages. On the low side below 0.5 papers say is implausible because it is too low to end ice ages. The tails in a well defined probability distribution here should be cut off, but the Spencer number is well before the cut off. Perhaps I should have specified that, and it is clearly possible that the sensitivity is not a constant, as feedback say with lots of ice may be different than feedback with low ice levels.

I hope this helps.

 

The range of plausible values for CO2 (more broadly, all IR-absorbing gases) has not narrowed much over decades of research. This has been touted as a research failure: Money has been spent, lots more paleo-proxy records exist, so why can't the number get nailed down?

I think it is related to the paleo-proxies themselves. They are never exact. Cannot be. So we have more of them and they are generally consistent, but the breadth of the probability distribution persists.

This leaves us in the situation where, if you are inclined one way, you emphasize the upper tail of the distribution.There CO2 is hazardous. If you are inclined the other way, the lower tail, and CO2 is rather benign.

The unfortunate folk who emphasize the peak of the distribution suggest that we really ought to be doing more to limit fossil C release. but they get no love from folks out on either tail!

So we plod along. Continue to heavily rely on fossil C. Dabble in renewable energy and energy efficiency. It really could be otherwise, you know. We could strongly limit CO2 release and learn a lot more about climate system (not to mention technological innovation). If the low tail turns out to be correct, then heck, burn the stuff later. It's not going anywhere you know. It would 'happily' wait in geological storage for us to get the probability distribution function narrowed.

 

The ice does help tell a story about ghg concentrations and when ice ages and interglacials occured. IRRC there was a paper in 2001 that had the top of the plausable range as 10, now because of calculations IMHO this has been lowered to 5. Hansen who was one of the proponents of possibilities higher than 5, in his last couple of papers agreed that fast feedback and ghg forcing were more limited. This does not mean that tipping points don't occur, we can see these clearly in the ice proxy, but we should not confuse that with sensitivity. The lower limit, as you say has stayed stable.

We have time lags with temperature/ghg feedback, and this leaves large areas of uncertainty. Even if the proxies were better, most of the feedback is radiation changing temperature leading to positive ghg feedback. There isn't much in the proxy record with ghg leading temperature.

The solution is better models that can compute feedback, but unfortunately these are not good enough yet. Key areas of modeling failures have to do with ocean oscillations and feedback. If we can model natural variation better, then sensitivity should be easier.

These are political battles, and because politics means funding, we often have research that exaggerates the number high or low. That does not mean this is bad research, but when its reported on one of the affinity blogs it often gets distorted for the general public.

By peak, if you mean the centeral area (1.5-4, peaking in probability (mode) between 2.8 and 3)
Many of them advocate for less, some for more, I don't know. If you mean the higher sensitivity tail (4-5) these folks often think we have passed the tipping point. If we are at 5, then at 400ppm the temperature increase baked into the system is 2.5 degrees C over the average of the period 1850-1900. If 2 degrees is the typing point its already to late.

Some at a sensitivity of 1 still would like to burn less coal and oil, but because of pollution not carbon dioxide.

What if the heads of the organizations calling for less carbon, tend to use the most, and profit most by government programs to reduce it? Perhaps we need to do more than reduce CO2 by buying mansions and flying on private planes and preaching about it. Oh and if we are going to leave some in the ground, let's make good decisions, and not say you can burn as much as you want if you are a senator or former senator in the US, or a rich englishman, but pox to you if you are in the chinese middle class. Although many would like to burn less, often the medicine to burn less is really a payoff.

I would like to burn less, but with some easy fair programs. Just a tax on oil and coal, and some incentives for gas ccgt, wind, and solar. I don't want to burn less by taxing everyone to build a fisker, cause the guy that made a movie wants one, or to tax everyone to give Duke energy money to close down a coal plant and build a nuke. We need good programs to use less.

 

We have been working at least a century on how bacteria work and there is still a lot of unknowns still to be understood. The entire climate system of the planet would be a smidgen more complicated I would think. Similarly, we have an excellent understanding of what is important, where the additional CO2 is coming from, where it is going, and what the possible range of effects are in a gross sense. What many think is needed is an ability to predict exact results of future effects down to the mm change in sea level and the number and strength of every storm. That's never happening.

It is impossible to change the entire world's energy infrastructure significantly in a couple of decades. Period. It might be possible to change the world's energy infrastructure in a century if we plan intelligently. Fortunately, that is actually starting to happen. Cars are becoming more efficient at an increasing rate. Coal Plants are phasing out at an increasing rate. Recycling is becoming more prevalent. Product sustainability and environmental impact is now very much part of any technical engineering design course.

The real change has to come from the bottom up. What you describe is a political system that forces the "right" answers upon a population that does not understand what is being done to them. That's not happening, nor is it going to happen. Don't think showing them a horror story ending works. Showing them how to save lots of money and increasing their standard of living using sustainable energy does work. We all all talking to each other here because we took the step of buying a really efficient car. Think about that.

 

FL_Prius_Driver, if your interest in bacteria is substantial, I'd love to discuss it in another thread. I think that the details about how they (including fungi please) could respond to a range of changes are quite poorly understood. They and climate dynamics both have the potential for simplified descriptions, after fundamental mechanisms can be accurately stated...

Pace of fundamental changes: I suggest that there have been examples of prompt fundamental changes. but the fossil energy system is not the place to look for them yet. Maybe enhanced extraction (frakking) is one though? It came on rather rapidly.

I completely take the point about top-down vs. bottom up, and framing the issue in terms of economic benefits. We have talked about McKinsey here before


Unlocking energy efficiency in the US economy | Electric Power & Natural Gas Practice | McKinsey & Company


Energy efficiency: A compelling global resource | McKinsey on Society

but somehow, they don't make the news nearly as much as the latest studies on 'it's really bad' or 'it's a strongly overstated risk'.

 

It took me a while to figure out how to get my point across without inferring any criticism for those wanting more to be done than what is being done. Here is my point:

Let's say that extremely rapid progress is made and we can model with incredible accuracy and agreement what the future warming effects of CO2, etc. will be upon the climate. Let's say that we are so good at this that the heating per ppm of CO2 in the atmosphere is "X" within a tenth of a degree. Let me take this fantasy one step farther and say even Mojo agrees.

Would governments and organizations significantly change their behavior as a result of this precise and agreed upon scientific knowledge of high precision (vs. what we know now)? That is my key question.

When faced with this question many can and will offer opinions. I prefer to look at history such as CFC pollution, Auto emission controls, and Acid Rain responses. A couple of clear lessons are clear to me. Here they are:

1) Smart people start figuring out mitigations as soon as they know the general issue, not when widespread consensus is achieved, which is much later. Panicking the larger population just ends in a circular stampede. Right now a whole bunch of unheralded smart folks are working a number of issues in the background. They don't need more accurate models to provide more motivation.

2) Any mitigation effort that does not completely factor in economics is worthless. Mitigation efforts that do harness economics are extremely powerful. For CO2 mitigations, this is the do or die issue. Carbon Markets and many other government schemes are often flawed and corrupt, so this is where the real failure or success of proposed mitigation strategies is determined. Carbon Credits can make Goldman Sachs rich and do nothing for reducing CO2. We cannot have that. Government rushing to action are much more worrisome than encouraging. Politicians have no scientific training or discipline and scientist rarely have a clue of the difference between effective and ineffective economic policies. Careful, incremental implementation is needed, not immediate grand schemes.

Nothing above should affect climate science research. Only good things come from learning more of how nature works. But it is the crew figuring out how sustainable solutions can be more profitable that will make the most difference at this point.

 

Back to the arctic, a recent publication suggests that reductions in sea ice coverage allows more carbon sequestration by the plankton there. Can't get at the atmospheric CO2 when you are wearing an ice hat. So, that's good news.

It's this
doi 10.1002/2012GB004491

but I can't get a copy w/o playing a library game. Later.

If that is so, then logically, increased sea-ice coverage in the Antarctic Sea should have the opposite effect. I have not seen anything published about that.

To save time at the expense of being OT here, another bit of good news. Releasing nitrogen and phosphorus into coastal oceans stimulates phytoplankton there, and they pull down more CO2. This is in Nature
doi/10.1002/2012GB004491

When the N and P transfers are too large, then you get the dead zone (like mississippi) and that's no good. But lower inputs are good news. Maybe we should make coastal water-treatment plants just a little bit leaky? Just a thought.

Also there was a paper long ago about UK water-treatment plants discharging 'too-clean' water. phytoplankton reduced, effect continued up the food chain, and seabird populations crashed. Don't remember any other details.

But here the theme is that N and P in fertilizer runoff, pee and poop are highly valuable ecological resources and we'd do well to think about how to manage them comprehensively.