Tuesday, July 29, 2014


Noah Smith is having fun with automatic priors, pointing out that the reasonable person's prior on personal immortality is that you got it because you ain't dead yet.

Consider Proposition H: "God is watching out for me, and has a special purpose for me and me alone. Therefore, God will not let me die. No matter how dangerous a threat seems, it cannot possibly kill me, because God is looking out for me - and only me - at all times."

Suppose that you believe that there is a nonzero probability that H is true. And suppose you are a Bayesian - you update your beliefs according to Bayes' Rule. As you survive longer and longer - as more and more threats fail to kill you - your belief about the probability that H is true must increase and increase. It's just mechanical application of Bayes' Rule:
Frequentists may disagree, bodycounts tell a bunny something, Noah points out that while there are other reasonable priors, how to determine which one should be used appears to be a religious matter.  OTOH, there are people, particularly teenage boys who use Proposition H daily. 

Eli and Socrates debated similar issues a while ago and Andrew Gelman about eight years ago pondered same
The fundamental objections to Bayesian methods are twofold: on one hand, Bayesian methods are presented as an automatic inference engine, and this raises suspicion in anyone with applied experience, who realizes that di erent methods work well in different settings (see, for example, Little, 2006). Bayesians promote the idea that a multiplicity of parameters can be handled via hierarchical, typically exchangeable, models, but it seems implausible that this could really work automatically. In contrast, much of the work in modern non-Bayesian statistics is focused on developing methods that give reasonable answers using minimal assumptions.

The second objection to Bayes comes from the opposite direction and addresses the subjective strand of Bayesian inference: the idea that prior and posterior distributions represent subjective states of knowledge. Here the concern from outsiders is,  first, that as scientists we should be concerned with objective knowledge rather than subjective belief, and second, that it's not clear how to assess subjective knowledge in any case.

Beyond these objections is a general impression of the shoddiness of some Bayesian analyses, combined with a feeling that Bayesian methods are being oversold as an allpurpose statistical solution to genuinely hard problems. Compared to classical inference, which focuses on how to extract the information available in data, Bayesian methods seem to quickly move to elaborate computation. It does not seem like a good thing for a generation of statistics to be ignorant of experimental design and analysis of variance, instead becoming experts on the convergence of the Gibbs sampler.
Lars Syll riffs off Smith with a little less patience noting that as a religion Bayesianism is dangerous.  The right to carry automatic priors can produce nonsense.  James Annan showed this with Myles Allen's way to wide uniform prior, and the recent discussion about Nic Lewis' attempt to mangle C14 dating priors should caution the Bayesians.

Long live frequentism, at least when there is enough data. 

Monday, July 28, 2014

Curious Rabett

The British House of Commons Energy and Climate Change Committee has issued its report on the AR5 WG I.  While the report strongly supports the IPCC (see below), there is opposition within the Commons and the Murdochsphere.  An important issue is how does one know the opposition is blowing smoke.  There are tells. 

In commenting on climate change, Eli has always been of the opinion that the thin bench is the weakness of denialism.  In this, it was interesting to note that the opponents of the IPCC on the House of Commons Energy and Climate Change Committee invited Richard Lindzen and Nick Lewis accompanied by one Donna Laframboise to give oral testimony.  Bunnies can go back and forth about Lindzen and Lewis.  Lindzen has been wrong about a lot of things.  Lewis has a hammer.  Laframboise, well, it is difficult to say what she is, at least politely, and she was a challenge to the Committee.  A bit of the oral testimony shows that they knew they were being insulted  and oh yes, Richard Lindzen was maybe not held too seriously either.

Q75 John Robertson: I remember a movie that I think was called The Core, and the scientist who was a bit of a pain in the backside made a statement that “science is a best guess”. Nobody knows exactly what it is, but it is a best guess. We have to go on the best guess. You need to help us.

Professor Lindzen: Yes, but the question is, how do you decide on policy? In other words, it is kind of a point of amusement. Somebody says, “There is a problem”, and the politician says, “We have to do something”.

John Robertson: There lies your problem. Your problem is the science. Our problem is the policy. I do not get involved in the science and perhaps probably you should not get involved in the politics.

Donna Laframboise: If I might make a comment. I am a journalist with no scientific background whatsoever and I had left journalism for seven years. I got laid off at the National Post with 130 other people. I decided to do something else with my life. Then I started doing some independent research on climate change and decided I should write a book because I realised that there was a wide diversity of scientific opinion about climate change, but that diversity was not being reflected in the news reports that I was reading. I thought I was writing a book about 10 reasons to be calm, cool and collected about climate change because there are diverse points of view. Instead, I ended up writing an expose of the IPCC, quite unexpectedly, because the more I did some basic fact-checking about what we were told about how the IPCC works as an organisation, the more concerned I got.

I compare it to a criminal trial. If we find out that there was bias among the jurors, then we have to throw out the verdict and start again. What I see is there are so many questions and biases and potential problems with the IPCC process that I do not think it is trustworthy. I do not think any reasonable person can look at the IPCC in depth and say, “We can trust this decision”. It is unfortunate. I wish we could because it would be much clearer what we should do.

John Robertson: With the best will in the world, you are one person and a lot of other people would disagree with you and you have had your chance to sell your book.
One must thank the members of the committee in denial, Graham Stringer and Peter Lilley, for inviting Ms. Laframboise.  It would be hard to think of a clearer marker that denial is not to be taken seriously.  The fact that she was invited shows that there is no there there.

The Committee report issued just now is quite strong:
80. The conclusions of this inquiry are very clear: the WGI contribution to AR5 is the best available summary of the prevailing scientific opinion on climate change currently available to policy-makers. Its conclusions are derived with a high confidence from areas of well understood science. Uncertainty remains in a small number of important areas but these are diminishing. It is important to consider all lines of evidence together when assessing climate change rather than focusing on particular aspects of the report. The overall thrust and conclusions of the report are widely supported in the scientific community and summaries are presented in a way that is persuasive to the lay reader.
81. The size and scale of the report reflects the huge effort by the international climate science community, who volunteer their time and expertise. We can now be more confident than ever that human activity is the dominant cause of the warming witnessed in the latter half of the 20th Century. The most significant human impact is through the release of carbon dioxide, which is predicted to continue to cause warming in the coming decades and centuries.
On the issue of climate sensitivity the report states that
The WGI contribution to AR5 has considered the full range of both Equilibrium Climate Sensitivity and Transient Climate Response and given the best assessment possible within the constraints of the evidence available at the time. It does not appear that a consistent pattern for higher or lower sensitivities than that stated in the WGI contribution to AR5 has emerged since its publication. (Paragraph 48)
and on the global temperature anomaly
Periods of hiatus are consistent with earlier IPCC assessments that non-linear warming of the climate is to be expected and that forced climate changes always take place against a background of natural variability. The current period of hiatus does not undermine the core conclusions of the WGI contribution to AR5 when put in the context of the overall, long-term global energy budget. Despite the hiatus, the first decade of the 2000s was the warmest in the instrumental record and overall warming is expected to continue in the coming decades. (Paragraph 53)
and on models
The models used in the IPCC’s Assessment Reports have a successful history of simulating past climate and their future projection of substantial warming over the next century in all but the most aggressive mitigation scenarios is well founded and overwhelmingly clear. (Paragraph 64) 

Sunday, July 27, 2014

Climate change, desal and astronaut water

As Eli points out, California has a water crisis, and much of the rest of the country needs to be much more water efficient. Water efficiency is the obvious place to start but then the next-step question comes up.

While plenty of people don't live near the ocean, lots do. Oceanside areas with large populations are going to have wastewater treatment plants (in developed countries, anyway).  These places therefore have two potential sources of new water supply:  ocean desalination or potable reuse of recycled wastewater.

Potable reuse of wastewater is nothing new. Almost any city drawing water from a river which has another city upstream is already doing it; the question is can we do it without lying to ourselves. In the case of astronauts and the International Space Station, they can do it outright, but the rest of us have to catch up.

Or not. Ocean desal actually uses the same technologies that potable reuse requires, either distillation or more commonly through reverse osmosis. The difference is that ocean water has a lot more stuff in it (mainly salt) than wastewater which already has to go through some purification before it reaches your reverse-osmosis system. That means a lot more energy and cost is involved in ocean desalination than potable reuse, so we've got a climate change issue.

The other climate change issue is that the lack of water currently stops a lot of unwise sprawl development, but ocean desal could change that, or maybe even mandate it - a very expensive desal system could be built on the expectation that there will be a lot more development to pay for it. I suppose there's some sprawl risk from potable reuse as well, but because it functions best in an existing populated area, starting at the wastewater treatment plan and then spreading from there, the risk is lower.

Many other factors involved of course, but these are the main climate issues. All but one of the factors weigh in favor of potable reuse. The one factor favoring ocean desal is psychology and political acceptance. People hesitate to drink this water, and that hesitation killed an earlier potable reuse project in San Diego (p. 17).

I view desal and portable reuse as being in a race. Money is limited so communities are going to prioritize. As much as I can I've supported potable reuse and opposed desal.

First step for potable reuse is Indirect Potable Reuse, achieving psychological acceptance by making the treated water sit somewhere for a while before reuse, either in a reservoir or underground. It's good but maximum flexibility and less cost require Direct Potable Reuse, shunting the water to your drinking water plants.

At my water district we've set up a reverse osmosis system. Currently it's just to improve the quality of non-potable recycled water which will help with certain types of uses, but the goal is potable, if we can get public acceptance.

Note:  stumbled across this - Los Angeles actually constructed an indirect potable reuse plant in the 1930s, but shut it down when they acquired Colorado River water. Back to the future, like with electric cars.

Also, desalination sometimes refers to desal of brackish water, usually groundwater. This water is much less salty than ocean water so a lot of the energy concerns are reduced with brackish desal. But brackish water and even potable reuse require a fair amount of energy, just nowhere near as much as ocean desal.

Tweeting science over the last few thousand years

Twitter conversations by scientists, beginning with one of Pythagoras' tweets. Be sure to read the ones from Jonas Salk.

Wednesday, July 23, 2014

Meeting Michelle yesterday

(Not very substantive post on meeting Michelle Obama, click here to read if you're idly curious....)

Tuesday, July 22, 2014

The Future Today

Over at ATTP, Rachel, playing while Wotts is away, asks about community driven solutions to climate change.  Pekka misses the point in one comment (there are others in the thread which is worth reading, by Pekka and others)

Governments have had some role in those technologies as well, but what has made exactly these fields to grow at an exceptional rate, has not been the government. It’s worthwhile to ponder, what leads to the huge success of some developments, but not of the others. One factor is certainly that technology development has revolutionized data and signal processing, while most other technologies have not shown potential for anything comparable.
Eli was not as nice as he could be, pointing out that government is a vital player in anything that has to do with anything
Government had nothing to do with the coming of the railroads, road traffic or air travel. OK Pekka?
You could add electrical distribution networks and pipelines to that.  In this regard Eli would like to point to a map of the future today

More than there are gas stations inside the Périphérique.

Saturday, July 19, 2014

Stigma for fossil fuel companies, the reverse for the churches that dump them

World Council of Churches, Unitarian Universalists, United Church of Christ, and many smaller/regional church denominations and affiliated organizations have established climate divestment policies. Others are percolating - the Methodists are studying their investment policy, the Presbyterians are first going to try to persuade the companies to give up their core business model (good luck with that!)* and then we'll see them and others consider this issue.

People involved in climate divestment and had also been involved in South Africa divestment a generation ago say that climate divestment is moving faster. An Oxford study backs that up (p. 11).

The same study acknowledges limited direct financial impacts of divestment except for coal industry, but then focuses on the stigma issue:

As with individuals, a stigma can produce negative consequences for an organisation. For example, firms heavily criticised in the media suffer from a bad image that scares away suppliers, subcontractors, potential employees, and customers. Governments and politicians prefer to engage with ‘clean’ firms to prevent adverse spill-overs that could taint their reputation or jeopardise their re-election. Shareholders can demand changes in management or the composition of the board of directors of stigmatised companies. Stigmatised firms may be barred from competing for public tenders, acquiring licences or property rights for business expansion, or be weakened in negotiations with suppliers. Negative consequences of stigma also include cancellation of multibillion-dollar contracts or mergers/ acquisitions. Stigma attached to merely one small area of a large company may threaten sales across the board.
(p. 14, citations removed)

The stigmatization from divestment will have financial consequences. These companies will have to pay more for employees and for other businesses to work with them. Companies with a toe-hold in the fossil fuel sector will find it better for them to get out.

Most important is that stigmatized industries will find it tougher to manipulate the political sector. That's one reason why they disguise their funding, but the disguise is imperfect, and the difficulty gets worse with the stigma.

Two other points. The study acknowledges political restrictions resulting from the climate divestment effort could destroy the perceived value of reserves that end up staying in the ground. When the carbon bubble pops is hard to predict, but any downward pressure increases the possibility of it happening soon.

Second, when companies divested from South Africa they weren't required to physically blow up the businesses they left behind - they sold them. The argument that it had no financial impact was around then, but we see what happened in the end.

*I think there is a business case that fossil fuel companies should 1. stop wasting money exploring for new reserves, 2. sell the reserves they're not going to be allowed to develop before the carbon bubble bursts, 3. play out the remaining and cheapest reserves and 4. either distribute the profits and wind down their companies, or invest in another business model. Not bloody likely to happen, though.

I'm ignoring the complications of when natural gas can substitute for coal. 

Saturday, July 12, 2014

Down the Drain

Eli has long been an advocate of the five fold way
  • Adaptation to deal with the damage already done
  • Amelioration, eliminating harmful effects of our actions
  • Conservation with needed and desired but not wasteful usage
  • Substitution of green systems for destructive ones
  • Mitigation reversing our thoughtless abuse
and not just for dealing with climate change.  Neat and tidy are two virtues Mom Rabett was strong on, and Ms. Rabett, well at times she is just plain cheap, Eli being thrifty.  Thus waste offends.

There has been considerable noise in the climate set about leaks from natural gas pipes and wells.  An  article in the NYTimes by David Bornstein points out that water systems are sieves. 
The Environmental Protection Agency estimates (pdf) that public water systems lose, on average, one-sixth of their water — mainly from leaks in pipes. The E.P.A. asserts that 75 percent of that water is recoverable. (In truth, the volume of leakage in the nation’s 55,000 drinking-water systems is unknown, because few conduct water audits.
with loads of consequences
It’s been widely reported that California is experiencing its worst drought in history. But take a look at the United States Drought Monitor: much of the country is abnormally dry or in drought. Internationally, the problem is even more serious. The World Bank reports that, over the next decade and a half, water availability may fall 40 percent short of global need (pdf).

Meanwhile, utilities in the developing world are hemorrhaging water. The World Bank estimates that water systems have real losses (leakages) of 8.6 trillion gallons per year, about half in developing countries (pdf, 11MB, p.6). That’s enough to serve 150 million Americans (and we use a lot of water!)
Bornstein describes how the Bahamas are dealing with the problem in collaboration with a consulting firm Miya and the obvious progress they are making

Sealing up systems can not only save money, but because water systems are often limited by water supply, it can insure delivery.  Doing so requires commitment, technology and funding, but pays multiple dividends.
One study (pdf) conducted for the California Public Utilities Commission examined audits done by 17 water utilities and found that losses were 1.6 to 6.6 times higher than optimum levels.  Assuming that 40 percent of the losses could be recovered economically, the study’s lead author, Reinhard Sturm, estimated potential savings at 113 billion gallons per year — equivalent to the annual production of six Carlsbad projects.

It’s vital to consider the impact on energy use and the environment. Water is often lost between the main pipe and the customer, which means it has already been extracted, treated and transported a very long way. That’s expensive. All that energy is lost — and more has to be used — and that, of course, increases carbon emissions. California’s water system is already the state’s largest single energy user. At the same time, desalination plants are energy intensive. Electricity accounts for roughly half the cost of their water.
Oh yeah