For my January 2010 diary, go here.

Diary - February 2010

John Baez

February 1, 2010

He Who Conquers

There is he whose spirit casts shadows
upon valleys
he who drowns thunder with words
who steps heavy and broad
pressing green into soft wet earth

and there is he who does naught
save to unweave the spider's web
and bind it back again –


February 2, 2010

Check out this moonlit fog bow in Hawaii, with Mars gleaming bright.

February 3, 2010

In email, Robert Smart wrote:
There are lots of "might save the world" stuff that need mathematical help, but I'd like to suggest something different.

Resource depletion and environmental destruction, and the measures to counter these things, will play out through the economic system. Understanding of the economic system is shockingly weak. Much of the understanding only works in a growing economy, but oil depletion alone will see economic contraction for a decade or so.

All current mathematical understanding of economics seems to be in terms of money. But money is one of the oddest things in economics. It doesn't have the properties of stability it is assumed to have, when governments can "print" more at the stroke of a pen (quantitative easing).

So how about running some sort of cooperative Internet thing to try to build a better understanding of economics. Certainly money and some other things look like fluids, so there is scope for n-ports. And maybe the work on large sparse random matrices has something to do with very large networks with semi-regular features. Obviously I'm guessing.

When you need to make significant changes to a large complex thing like the economy, it is wise to understand it better. However I think there are many other reasons to do this. For one thing: the Nobel Prize for Economics seems to be a lot easier to get than any of the others :-).

Robert Smart

P.S. I have been writing stuff in an attempt to understand economics better in a qualitative way. However we don't understand stuff till we understand it mathematically.

P.P.S. With the latest feature of Google Reader I can get changes to your Diary as if it had an RSS feed. We'll see if it works.

February 4, 2010

Pretty soon I'm going to start a new blog focused on environmental and related issues. My plan is to learn stuff by explaining it. In the meantime, I've been getting some emails on this topic. For example, in response to Robert Smart's comment above, Dave Tweed wrote:
Dear John,

Since I saw Robert Smart's comment on your diary, I guess an email is appropriate. I wouldn't presume to give you advice, but I'll just suggest that money is not a well-defined fundamental concept in economic theory, and probably can't be uncontroversially defined. Arguably a more important understanding in economics is understanding how feedback/system effects can manifest themselves. From there you can start to figure out what characteristics money needs to play, and hence how it should be "defined". One example of the kind of thing is presented here, where it's suggested that the historic choice of definitions and tools leads to what you consider the problems to be (namely only those that fit nicely within your toolset framework):

I'll leave it there, but if youre interested in money as hydraulics consider MONIAC.

Anyway, best wishes,
Dave Tweed

February 6, 2010

A lot of rain! We need this rain badly... but it's causing mudslides near the coast.

According to Southern California Weather Notes:

The recent enhancement of El Nino convection in the equatorial Pacific by the Madden-Julian Oscillation (MJO) has triggered a strong atmospheric response. A Global Wind Oscillation (GWO) phase space plot shows large increases in relative atmospheric angular momentum (AAM) and AAM tendency. As a result of this increase, the average relative AAM anomaly for the rain season to date is now positive. As mentioned in this post from December 2009, relative AAM is correlated with rain season precipitation in Southern California. This suggests an increased likelihood of wet weather in Southern California in the medium range outlook period.

That December post says:
Will the current El Niño produce the expected seasonal impacts in Southern California? A new tool that can help gain some insight into the linkage of climate and weather is the Global Wind Oscillation (GWO) phase space plot. Ed Berry repeatedly demonstrated the usefulness of this tool in his blog Atmospheric Insights. Although the blog has been discontinued, its content remains a valuable resource.

The GWO is a recurring subseasonal phenomenon that involves the transport and interchange of momentum in the earth-atmosphere system. It encompasses the MJO and occurs on a similar timescale. Analogous to the MJO phase space plot, but based on a framework of atmospheric angular momentum (AAM), the GWO phase space plot is a measure global relative atmospheric angular momentum and its tendency. For details, see Weickmann & Berry, 2008.

Relative AAM is generally positive during an El Niño and negative during a La Nina. Relative AAM is correlated with rain season precipitation in Southern California, and can be helpful in assessing potential El Niño impacts. The following November to March GWO phase plots show the distinctly different behavior of the GWO during the strong El Niño of 1997-98, and the strong La Nina of 1973-74:

In a decade characterized by quirky El Niños, the El Niño of 2009-10 has been acting like another odd one. From a Southern California perspective, the concern has been that it might be like the El Niño of 2006-07 when Downtown Los Angeles recorded only 3.21" of rain over the water year. In the early stages of the El Niño of 2006 strong convection developed in the Indian Ocean during November, but an MJO did not develop until a second round of Indian Ocean convection occurred in mid December. Relative AAM remained negative, and in terms of the atmosphere, the 2006 El Niño didn't make it to 2007.

This year, Indian Ocean convection did spawn an MJO which eventually enhanced El Niño convection near the dateline. Significant momentum was added to the mid-latitudes of the Northern Hemisphere, energizing the westerlies and contributing to the pattern change that resulted in our recent wet weather. However, the increase in mid-latitude AAM has been mostly offset by negative anomalies at higher latitudes. Following are the GWO phase space plots for the current rain season to date, and the quirky El Niño of 2006-07.

Luckily the El Niño seems to have gotten its act together by now and produced some hefty rains!

I'm struggling to understand the graphs shown above. I believe that the quantity called M is defined on page 7 of this paper:

It seems to be an estimate of the total angular momentum of the Earth's atmosphere: AAM means "atmospheric angular momentum". Angular momentum is a vector, but they're only considering the component aligned with the Earth's rotation, since winds tend to blow east-west. A typical value of the angular momentum of the Earth's atmosphere is 1.25 × 1025 kilogram meter2 / second. This corresponds to a uniform wind blowing due east at a speed of 0.5 meters/second over the whole planet's surface.

The graphs above also show the time derivative dM/dt. A typical value of this is 1.8 × 1019 kilogram meter2 / second2.

I don't know what units the graph is using, but you can get the idea: an El Niño tends to boost the overall westerly flow of the planet's winds, while La Niña boosts the easterly flow. According to Rich Monastersky, writing for Science News:

During non-El Niño years, winds in the tropics tropics, also called tropical zone or torrid zone, all the land and water of the earth situated between the Tropic of Cancer at lat. 23 1-2°N and the Tropic of Capricorn at lat. 23 1-2°S. blow from east to west, whereas winds over the rest of the globe travel from west to east. Combined, they give the atmosphere a net eastward momentum.

The atmosphere routinely trades some of this momentum back and forth with the solid Earth as winds drag across the surface of the planet and push against mountain ranges. In the Northern Hemisphere's winter, the atmosphere speeds up and Earth slows. In summer, the reverse happens. El Niño boosts the atmosphere's angular momentum by slowing down the tropical easterlies and speeding the westerlies outside the tropics, says Salstein.

As the atmosphere speeds up during El Niño, Earth itself slows down to conserve the combined angular momentum. John M. Gipson of NASA's Goddard Space Flight Center in Greenbelt, Md., has tracked the planet's spin by monitoring changes in the length of the day. Over a typical year, the day shortens and lengthens by roughly 1 millisecond, mostly because of shifts in atmospheric angular momentum. During the current El Niño, the day has grown longer by four-tenths of a millisecond, he says.

For more, try:

February 7, 2010

In week293 of This Week's Finds, I announced that I'm going to quit writing This Week's Finds in Mathematical Physics after week300, and start a slightly column, This Week's Finds, which will address a new range of issues. I'm going to start by examining the case for human-caused global warming.

My friend Nathan Urban writes:

Strange to hear that TWFMF is ending ... but good to hear there will still be a TWF! And I think it will be a lot of fun for you to explore these new areas. It was for me.

A few things which come to mind as I read Week 293:

Photovoltaics are definitely the most interesting from a quantum physics and optics perspective ... but there's also boring old concentrated solar thermal (i.e. mirrors and steam turbines) which is already a pretty developed technology and doesn't suffer from limited supplies of rare earth elements. (Instead it needs water, which can be problematic in deserts! But some have proposed using molten salt to use for heat storage during night time, although you still need water for the turbines.)

Some papers on the decline and future of Arctic sea ice:

Papers on wind (and solar) spectra:

Jay Apt at CMU has been working on the inefficiencies induced by having to match the wind power spectrum to the electricity grid spectrum.

You talk about mesoscale and large-eddy simulation. This ties into "multiscale simulation", which is about how to represent sub-grid scale physics in a large-scale numerical simulation. Common approaches are to treat the sub-grid scale physics stochastically in some way as "unknown physics with given statistical behavior", either analytically or by embedding small-scale simulations inside larger ones to numerically compute said statistics. There are a number of mathematicians working on this, e.g. at these UCLA workshops this spring:

See e.g. work by Andrew Majda, Ilya Timofeyev, Weinan E, and others.

There is some debate as to how to do this properly, between "heterogeneous multiscale" vs. "equation free" approaches. (I gather the former is more systematically mathematical, and the latter is more purely numerical.)

Multiscale modeling is particularly important in modeling cloud physics since clouds are hugely important to the Eath's radiative balance but are too small to realistically model globally, hence "cloud superparameterization". See:

JB wrote:

I have some ideas. For starters, something like science journalism / blogging, but with a focus on issues where we need a scientist to understand technical details and then explain them clearly. I think I'm good at that.
Yes, that could be a good strategy. There are a number of topics which I think could benefit from a good technical exposition. In fact, I've toyed with writing a little book about this, with simple qualitative reasoning about systems dynamics. You can probably explain these things much better than I could.

1. Why the "urgency"

People need to understand better the ideas of sources/sinks in systems analysis and the concepts of physical and socioeconomic inertia. This is, e.g., one of the themes of the Solomon "irreversible climate change" paper you've mentioned.

For example, people often think that we just have to stop increasing our CO2 emissions and the problem will be solved. But CO2 levels will continue to increase as long as emissions (source rate) exceed the sink strength, and natural carbon sinks are much weaker than current emissions: terrestrial sink takes about 1/4 of what we emit, the oceans take about 1/4, and 1/2 stays in the air. So we'd have to drop emissions to half what they are now before natural sinks can even keep up with them, let alone reduce CO2 (assuming that the strength of natural sinks stays constant). A lot of people get confused between CO2 concentrations and CO2 emissions/sinks:

Related concepts are "airborne fraction" or AF (how much anthropogenic CO2 stays in the air) and whether AF is changing (and how that relates to how CO2 concentrations are changing, some people falsely think that a constant AF means CO2 isn't increasing). Also misunderstood is the difference between carbon atmospheric residency time and the e-folding time of carbon sinks. A given CO2 molecule only spends a few years in the atmosphere before ending up in a sink, but the effective lifetime of atmospheric CO2 is much longer, decades to centuries on average. (Analogy: a molecule is emitted and quickly absorbed by some sink, leaving CO2 concentration unchanged. But how long for CO2 concentration to decrease? For that, maybe you have to grow a whole new tree to absorb the extra CO2 molecules that aren't already being absorbed by existing trees.) This confuses many people:

In fact, there are multiple timescales involved in carbon sinks, and some of them are very slow. This is where I think Mason Inman's analysis and David Archer's work (e.g., The Long Thaw) are more relevant than Franklin Cocks, as you discuss in your diary. A lot of CO2 is going to be around for a long time, and this may affect the ice age cycle itself:

So, it's not just a matter of keeping emissions steady, we have to cut them, and cut them a lot, before atmospheric CO2 will actually stabilize or decrease. And ultimately, even cutting emissions to zero isn't going to necessarily avoid dangerous climate change. For example, if we cut them to zero by simply burning all fossil fuels! If we want to avert the full amount of climate change possible, we have to leave fossil fuels in the ground and never burn them, not just burn them more slowly (e.g., by more efficient use of energy). Otherwise, you're just delaying the problem a little. (There's so much focus on 2100 that people forget that a huge amount of global warming in, say, 2200 is also a problem!) What really matters is our cumulative emissions, the total amount of fossil fuels we will ever emit, and how much less that is than the total amount available. See here:

People don't understand this either. Many think that all we have to do is be more energy efficient, but this isn't going to stop us from burning fossil fuels, it will just slow it a little. To stop burning fossil fuels altogether, they have to become more expensive than alternatives (or regulated). (Also related is the Jevons paradox: if we become more efficient, that just lets us burn more fossil fuels for the same cost! To keep consumption down, economists have argued that you need a price incentive or need to regulate consumption.)

Then there is inertia. If we mitigate our emissions, we're not going to see that reflected in the climate system immediately. Most of the heat capacity in the planet is in the oceans, with a characteristic heat mixing timescale of decades. That's how long it takes to fully adjust to changes in radiative forcing (and the time also depends on feedback strengths, with stronger feedbacks implying slower responses). You can see this by analyzing the simple linear response model:

C dT/dt = F - λT

where T is temperature, C is heat capacity, F is radiative forcing (e.g. log CO2 concentration), and λ is a feedback factor (inversely related to climate sensitivity).

The oceans will slow the climate response to anything we do, sea level rise from thermosteric expansion will continue as long as the oceans continue to heat as they try to re-equilibriate. Ice sheets have even slower response times.

Then there is socioeconomic inertia. People often think we should be embarking on a crash program to develop breakthrough energy technologies. That just takes too long to rely on. It probably takes about 50 years from initial research, to development, to deployment, to widespread deployment for a new energy technology. Breakthroughs could eventually help, but things we can deploy widely now (like simple baseload solar) are the most important, when coupled to the long lag times in the climate system itself.

2. Climate sensitivity and feedbacks

What is CO2 going to do to the climate? Well, we expect about 3.7 W/m2 extra radiative forcing, top of atmosphere, from doubled CO2. (Radiative forcing is logarithmic in CO2 concentration, basically the Beer-Lambert law, so we count CO2's effect in terms of doublings.) This is expected to produce about 1 C of warming for each doubling. That by itself is not too bad. It's the feedbacks which get you: they can increase the response to 3 C or more. These are things like water vapor increases in a warmer climate (water vapor is a greenhouse gas), ice albedo feedback (melt ice, Earth's surface is darker and absorbs more heat), cloud cover changes, changes in the carbon cycle, etc. That's where most of the problem, and most of the uncertainty in how bad the problem is, comes from. I think people could use a good introduction to linear feedback theory in the climate system. Something like a layman's version of this:

3. Climate surprises, thresholds, nonlinearities, and abrupt climate change

These are the big risks which might be unlikely but could be so bad that we ought to take steps to reduce their risk anyway. Disintegration of major ice sheets, collapse of the Atlantic overturning circulation (although that's looking like less of a risk), runaway permafrost or methane clathrate melting, etc. Some of these can occur if you pass a "threshold" (bistable system, can exhibit hysteresis) and can be effectively irreversible. For example, if you melt Greenland and later reduce temperatures back to pre-industrial, the ice sheet won't return to its pre-industrial state. It will be gone. You need an ice age to rebuild it. I bet you could show this analytically in a simple conceptual model.

Some miscellaneous links that come to mind, perhaps not the best selection:

JB wrote:

Then, maybe learning enough and getting to know enough people to tackle some more specific technical projects.
Not sure what I can suggest about that, as most of the mathematical research I know of in climate science is pretty far from "abstract nonsense" and bond diagrams ... more numerical analysis.

JB wrote:

Apparently people need better modelling techniques all over, especially for designing the "smart grid". I can sense an opportunity for abstract nonsense like "bond graphs" here.
I'm no expert on "the smart grid", but you might want to look at CMU's energy research:

Some interesting people in their Engineering and Public Policy department are M. Granger-Morgan, Jay Apt, Lester Lave. For electricity research, see their Electric Industry Center:

At Penn State we have Seth Blumsack (who did his PhD at CMU) who works on energy grid design (although mostly from an economic policy perspective rather than, say, network design and modeling):

February 10, 2010

Another bit of rain.

February 12, 2010

Tim van Beek writes:
Dear John,

I always wondered if climate models that try to predict sudden climate changes could profit from the knowledge that physicists have about phase changes in condensed matter physics. Item 3 in Nathan Urban's email, "Climate surprises, thresholds, nonlinearities, and abrupt climate change", reminded me of this analogy. The article

was particularly interesting to me for this reason.

The following article is a little addendum to this:

That requires a subscription, unfortunately, but the authors are — remote — colleagues of yours. By the way, the Potsdam Institute of Climate Impact Research has a nice homepage that will be a good distraction, should you suffer from insomnia.

Kind regards,

Tim van Beek

February 13, 2010

In South Carolina, terrorists now need to register with the state. Here's the form: There's a $5 filing fee. You have to outline the fundamental beliefs of your subversive organization, attach a copy of its bylaws or the minutes from your last meeting, and list your members.

February 14, 2010

It's Valentine's Day! And last night we had our Lunar New Year party, where the guests help us make jiaozi — the dumplings that are traditional for this occasion in China.

Charlie Clingen writes:

Hi, John-

Recently I've been thinking a lot about your change of direction and what you will be doing next. I'm beginning to form a clearer idea of what I think you could do (and enjoy doing) in order to have a significant impact of the future direction of our country and our planet.

I'll sketch the outline and see if it appeals to you at all.

First, I'm assuming that we do best that which we like most to do (and vice versa). It seems to me that some of your most outstanding abilities include: the ability to summarize difficult, technical concepts and relate them to each other in new and interesting ways; the ability to explain complicated ideas clearly in novel and interesting ways; the ability to organize and display large amounts of interrelated information in efficient ways. You are also good at keeping blog responses under control before they spin off into flame wars.

Given that, and your past history as visible to me, I'm guessing that the following might be of interest to you.

A summary:

So, in a nutshell, that's a summary of my thoughts. I haven't thought through clearly yet the proper division of responsibility between blog discussions and on-line technical summary papers, but you know a lot more about that than I do.

As usual, I have violated one of my own "Charlie's Laws": "Never define a problem by stating your solution." But hey, sometimes I just can't help myself.

Hope this gives you some useful food for thought.

I believe you have the potential to make a big difference. Go for it! If later you decide it's not for you, you always have plan B — the deeper integration of advanced math and theoretical physics.

Best regards,

Charlie Clingen

This plan appeals to me, not only for noble reasons, but also for ignoble ones. I enjoy grandstanding and being at the center of attention. Unfortunately, pursuing this plan would require a vast amount of energy and dedication on my part — especially the parts that involve dealing with other people. I have a lot of energy when I'm doing what I want to do, but my interests flit around in a somewhat unpredictable way. I want to indulge that tendency a bit more than I've been able to recently. One reason I want to quit working on n-categories is that it started feeling very burdensome: there were more and more half-completed papers I needed to finish, less and less time for free exploration. I feel much happier if I wake up and have a few hours to do whatever I want — whatever excites me at the moment! That's when I come up with good ideas. As soon as I quit working on n-categories, I started having lots of good ideas.

This Week's Finds suits this aspect of my personality — it lets me jump from subject to subject without feeling guilty about it. I like the idea of writing a This Week's Finds whose scope is broader, including environmental issues. But making myself into a dedicated compiler of information — "the oracle, the wizard" — and building up people's expectations that I'll be there whenever they need a judgement rendered — that doesn't sound so good. I'll burn out!

So, I need to tweak Charlie Clingen's plan a bit. I want a role that's a bit less grandiose, a bit less of a full-time job.

Also: when my 5 grad students finish up, I'm going to switch back to having one or two at most. It's great fun having a big team of people to work with. But doing research with grad students is like driving a train: it takes a long time to start them up on something, and you can't stop a project suddenly, or turn on a dime. When I get more seriously into work on ecology, engineering, quantum computation and the like (and maybe algebraic geometry with Jim Dolan), I want to become more maneuverable.

February 16, 2010

A Tibetan Buddhist monk on a road in the Qinghai province of China.
He is on a pilgrimage to Lhasa, 400 miles away. The journey will take about six months.
Photograph by Christopher Brown of NPR.

Here's a fun and enlightening radio show about a journey down the Mekong River:

I'll be pretty near some of these places when I move to Singapore! I'd like to visit Cambodia and Laos. But I don't want to get anywhere near the opium fields of the Golden Triangle, or Myanmar. Those places sound like trouble.

February 20, 2010

It rained last night!

As you can see, I'm trying to get myself into the swing of living in Southeast Asia a little bit before I actually go there. So I found this story interesting:

A condescending story title, but a nice story about an album of Cambodian rock music from the late 60's and early 70's — the good old days, before the Khmer Rouge took over and devastated the country, killing almost a quarter of the population through execution, torture, forced work and starvation. The album is called Electric Cambodia, and it was compiled by a band called Dengue Fever.

This sort of music can grate on Western ears. For starters, the high-pitched, sweet but slightly whiny female vocals in Cambodian music don't sound "cool". It reminds me a bit of how Lisa used to combat neighbors who played loud rock music by playing Chinese opera at full volume! And there can also be something a bit pathetic about Asian attempts to imitate Western music fads. You think "why don't they do their own thing?" But music styles are like highly contagious viruses — they hop unstoppably from culture to culture, mutating in weird ways as they go. And I'm learning it's good to check your pre-established notions of "cool" at the door if you want to explore a wide range of music. And so, I think I could get to like this stuff.

Here's what I'm listening to these days. Click on the artist's name for more information on them, or the album title to hear a bit:

February 21, 2010

Mike Stay has some great pictures on his blog:

February 22, 2010

Simon Willerton arrived in town last Tuesday — he's a pal of mine, a mathematician from Sheffield who works on quantum topology. He's going to be visiting Riverside until the end of April. We had him over for dinner last night, along with a music professor at UCR named Paolo Chagas, who Lisa met recently — he turned out to live just a few

It was cloudy all day yesterday, and last night it finally broke down and rained again! The hills are nice and green.

February 24, 2010

In 1750 there were about 280 part per million of carbon dioxide in the atmosphere. Now it's about 380 ppm.

MIT scientists predict roughly 886 parts per million by 2095 in a business-as-usual scenario - that's the median of a probabilistic simulation described here.

So, by the century's end, the the amount of CO2 in the atmosphere may almost quadruple from its level before the industrial revolution. How much will each doubling of the CO2 raise the Earth's temperature? The long-term answer may be quite different from the short-term answer, because there are feedback effects like the ice albedo effect that take a while to kick in. A short-term estimate is roughly 3°C. Long-term estimates could be more like 6°C:

The MIT study predicts a median of about 5°C of warming by the century's end, with a 90% chance of warming between 3.5 to 7.4 degrees. They call for rapid and massive action to avoid this.

February 25, 2010

In response to my mention of Hansen's 2007 estimate of 6° C 'long term' temperature increase in response to doubled CO2, a correspondent writes:
I've always been skeptical about Hansen's claim of 6 C for long-term sensitivity. This may be appropriate over the glacial-interglacial cycles, in which there was a large ice-albedo feedback from vast continental ice sheets in the Northern Hemisphere. However, since we are in an interglacial, those ice sheets aren't there anymore. Thus, we should expect a much smaller ice-albedo feedback under further warming, even if the Greenland and West Antarctic ice sheets disintegrate.

Hansen was still making his 6 C claim in this paper:

I've never quite followed the chain of logic in that paper completely. He derives a slow-feedback of 6 C from the glacial-interglacial cycle in Section 2.2. But the question is does that generalize to slow feedback starting from a low-ice state? He argues that 35 My in the Cenozoic is a good analog for that, but I don't see him actually derive a number for slow feedback in Section 3.2. In Section 3.3 he argues that "the equilibrium climate sensitivity ... is almost as large between today and an ice-free world as between today an the ice ages", which implies that you should still get a slow 6 C into a future ice-free state. I guess I could sit down and try to verify this from figures, but he doesn't show a calculation as far as I can tell.

This paper was written in 2007 and published some time in 2008. Interestingly, in December 2008 Hansen gave the Bjerknes Lecture at the AGU fall meeting, with slides here. In these slides (see pp. 8-9), Hansen says that the empirical climate sensitivity relevant to today's climate is "nailed" precisely to 3 C, including paleo constraints. In the footnotes he notes that slow feedbacks are sensitive to the climate state. And he doesn't mention 6 C anywhere in the lecture. Many have interpreted this as backing off from the claim that the slow sensitivity relevant to a modern interglacial climate is 6 C. But he doesn't come out and say it.

My question to you is, since you've read his new book, does he say anything about very large slow feedbacks which are relevant to modern climate (i.e., a 6 C climate sensitivity)? i.e., is he still pushing this claim publicly? I would imagine that if so, he would put it in his book somewhere, since he's been pounding the drum about extreme climate risk (e.g., his runaway greenhouse stuff ... which he mentions in his Bjerknes lecture, but using GISS modelE with a climate sensitivity of something like 2.8 C, IIRC).

My answer: no, apparently Hansen is no longer pushing the 6° C figure. On page 45 of his book:
Fortunately, Earth's history allows precise evaluation of climate sensitivity without using climate models. This approach is suggested by the fact that some feedback processes occur much faster than others.


Using Earth's history, we can evaluate Charney's fast-feedback climate sensitivity by comparing the last glacial period, 20,000 years ago, with the recent interglacial period, the late Holocene. [...] We see that the total forcing of about 6.5 watts maintained an equilibrium temperature change of about 5 degrees Celsisus, implying a climate sensitivity of about 0.75 degree Celsius for each watt of forcing. This corresponds to 3 degrees Celsius for the 4-watt forcing of doubled carbon dioxide. The sensitivity is smack in the middle of the range that Charney estimated, 1.5 to 4.5 degrees Celsius.

From page 157:
Earth's temperature changed about 14 degrees Celsius between 50 million years ago and the recent ice ages. Between 50 and 34 million years ago, the period when there were no large ice sheets on Earth, we expect climate sensitivity to be 3 degrees Celsius for doubled carbon dioxide (the empirical climate sensitivity we inferred earlier from glacial-interglacial climate change).
I don't understand where higher sensitivity due to larger ice sheets fits into his story here, and I don't even understand where the fast/slow issue shows up. But the figure of 3 degrees, not 6, is the one that keeps showing up.

Here's an online book about tipping points for climate change:

Since I live in Southern California, where a cycle of drought, bark beetle infestations, fires, and mudslides have been destroying forests for about a decade now, I was particularly interested in the section on broad-scale forest die-back as a threshold response to climate change in the southwestern United States. Here's a quote, with references deleted to make it easier to read:

Currently, climate-induced dieback of woody plants is being recognized as an important vegetation response to climate variation and change, with examples of forest dieback emerging from around the world. (It should also be noted that other recent studies have documented increased tree growth in dry forests, perhaps because of increased water use efficiency.) Recent research shows that water stress appears to be driving increases in background tree mortality rates in western North American forests. In addition, observations of extensive tree die-off — especially from semiarid ecosystems where woody plants are near their physiological limits of water stress tolerance — are being documented globally, for example, in Australia, Africa, west Asia, Europe, South America, and North America. Climate-induced water stress over extended time periods can exceed the physiological tolerance thresholds of individual plants and directly cause mortality through either 1) cavitation of water columns in the xylem conduits ("hydraulic failure") or 2) forcing plants to shut down photosynthesis to conserve water, leading to "carbon starvation". These individual-scale threshold responses to climate stress can trigger tree mortality that propagates to landscape and even regional spatial scales, sometimes amplified by biotic agents (like bark beetles) that can successfully attack and reproduce in weakened tree populations and generate massive insect population outbreaks with positive feedbacks that greatly increase broad-scale forest mortality.


Although tree mortality almost certainly occurred across much of the southwestern United States in response to the 1950s drought (and probably for previous regional-scale droughts as well), few studies exist that allow scientists to test projections about the rapidity and extent of potential vegetation die-off responses to drought. A recent drought beginning in the late 1990s and peaking in the early 2000s affected most of the western United States. This was the most severe drought in the Southwest since the 1950s. Substantial mortality of multiple tree species has been observed throughout the Southwest during this 2000s drought. For example, mortality of the piñon pine spanned major portions of the species' range, with substantial die-off occurring across at least 1,000,000 hectares from 2002 to 2004. For both droughts, much of the forest mortality was associated with bark beetle infestations, but the underlying cause of dieback appears to be water stress associated with the drought conditions.

The precipitation deficit that triggered the recent regional-scale die-off of the piñon pine across the Southwest was not as severe (dry) as the previous regional drought of the 1950s, but the recent 2000s drought was hotter than the 1950s drought by several metrics, including mean, maximum, minimum, and summer (June-July) mean temperature. Although historic data from the 1950s is limited, available data suggest that piñon pine mortality in response to the recent drought has been more extensive, affected greater proportions of more age classes, and occurred at higher elevation and wetter sites than in the 1950s drought. Hence, the warmer temperatures associated with the 2000s drought may have driven greater plant water stress through increased evapotranspirational demand and resulted in more extensive tree die-off. Because global change is projected to result in droughts under warmer conditions (referred to as "global-change type drought") the severe piñon pine dieback from the recent drought may be a harbinger of vegetation response to future global-change type droughts. In addition to the die-off of dominant overstory tree species, high levels of dieback also were observed in other Southwestern U.S. species and life forms in response to the warm regional drought in the 2000s. These include species where bark beetles are unimportant or nonexistent, including one-seed juniper (Juniperus monosperma) — a co-dominant with piñon pine for much of its range; shrubs such as wavy-leaf oak (Quercus undulate) and mountain mahogany (Cercocarpus montanus); and blue grama (Bouteloua gracilis), the dominant herbaceous species in many of these woodland systems. In addition to direct climate-induced mortality, severe protracted drought also can cause substantial reductions in the productivity and soil surface cover of herbaceous plants, which in turn affects numerous other ecological processes. In particular, reductions in herbaceous ground cover can trigger a nonlinear increase in soil erosion once a threshold of decreased herbaceous cover has been crossed, through increased connectivity of bare soil patches. On the other hand, dieback of woody canopies tends to cause an immediate successional shift toward greater cover of understory vegetation if moisture conditions are adequate, which propagates a different set of effects.

For more, try: A quote:
Future drought is projected to occur under warmer temperature conditions as climate change progresses, referred to here as global-change-type drought, yet quantitative assessments of the triggers and potential extent of drought-induced vegetation die-off remain pivotal uncertainties in assessing climate-change impacts. Of particular concern is regional-scale mortality of overstory trees, which rapidly alters ecosystem type, associated ecosystem properties, and land surface conditions for decades. Here, we quantify regional-scale vegetation die-off across southwestern North American woodlands in 2002-2003 in response to drought and associated bark beetle infestations. At an intensively studied site within the region, we quantified that after 15 months of depleted soil water content, >90% of the dominant, overstory tree species (Pinus edulis, a piñon) died. The die-off was reflected in changes in a remotely sensed index of vegetation greenness (Normalized Difference Vegetation Index), not only at the intensively studied site but also across the region, extending over 12,000 km2 or more; aerial and field surveys confirmed the general extent of the die-off. Notably, the recent drought was warmer than the previous subcontinental drought of the 1950s. The limited, available observations suggest that die-off from the recent drought was more extensive than that from the previous drought, extending into wetter sites within the tree species' distribution. Our results quantify a trigger leading to rapid, drought-induced die-off of overstory woody plants at subcontinental scale and highlight the potential for such die-off to be more severe and extensive for future global-change-type drought under warmer conditions.

February 26, 2010

In 2008 the World Meteorological Organization figured out the 10 warmest years since people started keeping accurate records around 1850:

1. 1998
2. 2005
3. 2003
4. 2002
5. 2004
6. 2006
7. 2007
8. 2001
9. 1997
10. 2008

Now NASA reports that 2009 is tied for the second warmest year since 1880!

Oz sent me this link to lots of pictures and animations of the ocean's temperature:

February 27, 2010

Map of computers infected by the Waledac botnet in a recent 18-day period.
Picture from Microsoft.

Microsoft got US courts to help them launch an attack on the Waledac botnet — an army of approximately 20,000 - 30,000 computers that seem to be controlled by an outfit called the Russian Business Network, which specializes in spam, child pornography, malware, phishing and various other forms of cybercrime. Here's a graph of the number of IP addresses infected by the Waledac botnet each day in the last month — or at least the number newly discovered by the guys who keep tabs on them:

A quote from Bright's article:

Botnets — large networks of malware-infected PCs remotely controlled by criminals — are a serious problem on the Internet. The spam, phishing attacks, and malware that these networks send accounts for a massive proportion, in excess of 80 percent, of e-mail traffic. One such network, known as Waledac, has been stopped in its tracks after Microsoft got a court to issue a secret temporary restraining order. The restraining order took 277 domain names used by the criminals to communicate with the botnet offline. Without these domain names, it is hoped that the controllers of the botnet will permanently lose access to the machines running their malware.

The Waledac botnet is presumed to be run by Eastern Europeans and to be made up of hundreds of thousands of compromised machines. It sends hundreds of millions, if not billions, of e-mails each day, as well as distributes malware to help recruit new machines to the network. Microsoft's complaint describes in detail how the botnet is organized, with a complex hierarchical control system. At the root of the system is the command-and-control servers. The botnet uses the 277 domain names to connect to the command and control servers to download new commands. These commands are then distributed through the different tiers of the network using peer-to-peer transmission.

By obtaining the restraining order, this command-and-control system was disrupted; with the domain names offline, the machines in the botnet were no longer able to locate their control servers, rendering them mostly harmless. The court action had to be taken in secret to avoid warning the botnet's operators; with sufficient warning, they might have been able to set up new domain names and new control systems, thereby circumventing Microsoft's efforts. The names have now been offline for three days, presumably sufficient to cause permanent disruption, and the injunction is now public.

Similar action against past botnets has been attempted by security researchers before, but the results were only temporary as new command and control servers were set up. Microsoft's intent is for this action to be more permanent. "Operation b49," as Redmond has called it internally, still has further work to do to ensure that the peer-to-peer communication between computers in the botnet is disrupted.

February 28, 2010

A lot more rain yesterday! But today the sky is utterly clear.

I'm getting ready to start a new blog that talks about what mathematicians and scientists can do about the really big issues of our day: climate change, the mass extinction event, and so on. I want people to understand what I'm saying... so these tips are handy:

For my March 2010 diary, go here.

Through crumbled fingers gone.
Can never be recollected.
- Brian Eno

© 2010 John Baez