Thursday, April 25, 2013

How soon could the brimstone part of fire and brimstone actually come?

A Congressman recently said: "We shouldn't worry about global warming leading to sea level rise. After all, the Bible says God promised he will not destroy the world by flooding any more..."
But in that same passage in Bible, he is said to say "Next time it will be by fire and brimstone."
There are lots of serious "fire" hazards we all know about (form nuclear fire to tradiation from new 'ozone holes'), but what about brimstone? Could the stench of brimstone (the poison H2S) be closer at hand than we imagine?
Possibly so.... to the best of my meager knowledge.... not representing the folks I work for...

I recently posted some new details to a couple of lists:

Good morning, folks!

On again, off again, I have been asking myself and others: what would
be the objective indicators that could tell us how far we are from the
"green sky" scenario of mass extinction
(including extinction of humans) which Peter Ward warns about in his book?

Since it is not my job (or anyone else's?) to answer these kinds of
questions, it's been very much a part time thing, though I have
certainly tried to get input from domain experts
(and will continue to do so).

For the moment, it is looking a whole lot scarier to me than it was a
year ago, when I had less information.

In essence, there are two steps to extinction:
(1) "stratified ocean";
(2) "the trigger," which seems to be mainly a matter of low oxygen levels in the
lower layer of the stratified ocean.

Ward and Kump have different views about what the trigger has been and can be.
Ward suggests acidity due to rapid CO2 buildup (such as what we are seeing now).
Kump hints at other mechanisms for anoxia. Agricultural runoff at
today's levels may well be all we need for the trigger. The trigger
warrants further study, but if we get to stratified ocean, chances are
pretty serious that our fate is sealed. More precisely, that H2S and
radiation will rise inexorably to levels that kill us.

So for months I have been looking for hard information on the cutoffs
which would lead to stratified ocean in the northern hemisphere. (If
it's just the Northern hemisphere, that would
still be enough to kill us, e.g. as per the eocene discontinuity
discussed by Ward.)

There is a huge NEAR TERM literature on "overturning meridional
currents," some of it
replicating the obvious biases of political appointees to EPA during
the Bush administration and such. Digging deeper into the physical
realities, I now pay most attention to two sources: Physico Chemistry Pdf version
Chs 1 to 7/Chapter 2 Salt Temperature and Density.pdf


Complex as the overturning ocean currents are, they basically depend
on an input of energy from convection, just as an electrical device
depends on a source of voltage. Cut the voltage, and they go off.

For currents of wind in the atmosphere, the primary source of
"voltage" is pretty simple. The sun heats up the land and the air AT
THE BOTTOM of the atmosphere. The heated air becomes less dense,
because of heating, and wants to rise. When stuff at the bottom wants
to be at the top, that generates all kinds of convection currents.

For "normal" ocean, at room temperature, it is quite different. The
sun heats the water at the TOP of the ocean; it becomes less dense,
but it does not want to rise, since it is already at the top. Thus
there is no overturning or convection current. That's why we had
stratified ocean for most of the history of the earth - and why we
were at the knife edge of extinction, with twelve actual break outs
(events of mass death) when the "trigger" appeared.

Our current benign period (the past 2 to 25 million years, I haven't
tracked down)
is due to a very special circumstance at both poles. When fresh water
is colder than 4 degrees centigrade, it has a rather unusual physical
property. Heat it, and it becomes DENSER. As a result, when it is
warmed, it SINKS. That is the primary source of energy for
the currents which have kept the deep oceans clean and oxygenated for
millions of years, and have also created the warming currents on the
surface  which keep UK and France
warmer than the corresponding latitudes in Siberia. Ice in the Arctic
also has a big effect
on currents there.

So what is the cutoff?

Logic suggests that the cutoff is the temperature where water has
maximum density.
Once it gets that warm, further warming reduces density, and the
currents die (as do we).
Is it 4 degrees C?

The curve in the first URL above suggests that the cutoff is less than
4 degrees C, because
of salinity. On first looking at the curve (actually, two straight
lines in the graph), I thought
it would be close to 4 degrees C, because I read the horizontal axis
as percentage salt. But the example makes it clear that that is not
the units on that axis. The second URL suggests that the prevailing
salinity in north Atlantic is high enough to imply a cutoff less than
zero degrees C.

As the ice melts in the Arctic, lots of changes are coming. If we end
up with a normal level
of mixing, making it like the North Atlantic, it suggests that the
time when all the ice has melted in the Arctic may already be the time
when stratified ocean begins. You can find ideological tracts saying
that the ice of the Arctic cannot possibly melt within the next four
centuries, but empirical reports of what's happening in the Arctic are
quite different.

And so -- it's not my job, and I know that triggers have some role too
here, but on balance,
I wonder how much time we have left.

Some in the lifeboat foundation have said at times "Of course organic
humans are
virtually dead already. BUt can we build intelligent machines?" We
just had good discussion of that in an IEEE conference in Singapore.
At the rate we are going with real design (as opposed to the writing
of visions and traditional overpromising), if we really accelerate
things, we might be able to produce
a mouse-brain level of intelligence in about a century. (Of course, we
already have the required hardware density, according to common
assumptions about complexity of neurons,
but the systems level issues are far more difficult). Not soon enough,
and not high enough
intelligence to give any kind of "singularity."

Where does that leave us?

Best of luck,


From a res;pondent: 
> Other GEE'd are less predictable.

Some say the cutoff will come as soon as 2020 (see above).

Your sources deal with SOME aspects of the problem, not those which I
raised, based on the simple physics of it. I really doubt other
serious sources would plot different curves for the maximum density or
freezing point as a functon of salinity!
No need to debate details of huge models when there is simple physics at work.

Best of luck,


Saturday, April 20, 2013

vast scandal in the economics of austerity

Luda just showed me a rather interesting story about an economics grad student named Thomas Herndon at UMass Amherst who has just had a huge abrupt impact on the world
of economic policy:

I tried to post a comment, but the site is one of those with creeping demands, so I just post my reactions here:

The story is quite   interesting  for  additional reasons beyond the  obvious.    Of    course it was essential to  science and the progression of knowledge that the truth should come out. Yet   Carmen should not have a strong incentive to withhold data; there is severe moral hazard  in  punishing her on that. As  information is not a private good, we face huge dilemmas of this flavor, seriously inhibiting our ability to learn more about reality in many fields of science. (NIH's  open access policy is one good step to deal with some aspects, but perhaps journals should support efforts to require data to be open.)
       Also... it seems that this particular world of policy making relies on stuff like mass clinical trials, versus the underlying "physiology" of the world economy. I have wondered for some time why all but a few people involved in austerity decisions seem to be unable to understand obvious basic concepts like Pareto optimality and how it plays out concretely here. The problem with the clinical trial approach (done right or done sloppily) is that the reality is very multivariate, and we need to understand nonlinear relations whose projection into simple linear statistics is essentially random. If people could think more analytically (or for that matter if they had used rational algorithms to assess mortgage packages), we wouldn't be having the problems right now. Just my personal opinion...