December 13, 2009
John Tierney On Climate Engineering

John Tierney points to discussions by climate scientists about whether to do climate engineering to prevent global warming. Tierney is confident that the nations of the world are not going to agree to huge sacrifices to cut back on carbon dioxide emissions. I tend to agree. So if CO2 really is going to warm the planet other measures will be needed to deal with it. So climate engineering is getting wider consideration.

The National Academy of Sciences and Britain’s Royal Society are preparing reports on climate engineering, and the Obama administration has promised to consider it. But so far there has been virtually no government support for research and development — certainly nothing like the tens of billions of dollars allotted to green energy and other programs whose effects on the climate would not be felt for decades.

For perhaps $100 million, climate engineers could begin field tests within five years, says Ken Caldeira of the Carnegie Institution for Science. Dr. Caldeira is a member of a climate-engineering study group that met last year at the Kavli Institute for Theoretical Physics under the leadership of Steven E. Koonin, who has since become the under secretary for science at the United States Department of Energy. The group has just issued a report, published by the Novim research organization, analyzing the use of aerosol particles to reflect shortwave solar radiation back into space.

Caldeira thinks we should test a cooling method well in advance of when we think we might need it. But other researchers fear side effects (e.g. decreased precipitation of rain and therefore crop failures) make such experiments too risky in advance of an acute need.

One of the methods mentioned is to squirt lots of ocean water into the atmosphere to increase white cloud formation and therefore to reflect more sunlight back into space. I would expect that technique to also increase precipitation (what goes up must come down eventually). While the rains wouldn't necessarily come where most needed I would expect they could in some instances. For example, wind-driven misting machines off the west coast of Canada and the United States could cause more rainfall over North American farm fields.

Since we have several ways to cool the planet which can be implemented quickly I think for now research on them ought to fall short of real experiments that alter our weather. Currently climate scientists are wrestling with the question of why haven't their temperature modeling stations detected any warming for the last 10 years? The last 10 years do not fit their models. Obviously their models are far from complete. The shorter term causes of variation aren't sufficiently well understood and so that makes picking up long term signals harder. What if some natural forcer is going to keep the planet from warming further for another 5-10 years?

The warming pause really points to a problem with climate engineering experiments: How to know whether, say, silicon dioxide or sulfur or water mists in a climate engineering experiment really did result in some cooling? One needs to have a certain degree of certainty about what the climate would be like without the intervention. If a climate engineering experiment had begun in 1999 the experimenters might have incorrectly concluded that their intervention caused the last 10 year warming pause.

The climate is an extremely complex system with lots of positive and negative feedbacks and multiple inputs. We do not have a couple dozen planets to experiment with in order to compare results of experiments and verify results. That's the fundamental problem with all climate research. There's a very limited potential for greater understanding thru experimentation. Yet experimentation is at the core of the scientific method.

Share |      Randall Parker, 2009 December 13 03:33 PM  Climate Engineering

Bruce said at December 13, 2009 6:14 PM:

Before "scientists" do anything stupid, they should read some of the articles sprouting up where people are anylyzing the raw data. The classic (of which there are many just in the just the last weekend) is Jeff Id discovering that GHCN only uses one station for the Antarctica, and it is the warmest station on the whole continent.

"So as we can see, of all the stations available in the antarctic, GHCN has chosen to use a single station on the Antarctic Peninsula to represent an entire continent of the earth for the past 17 years (red circle). But it’s not just any station, it’s a special one. Rothera Point has the single highest trend of any of the adjusted station data."

JAY said at December 13, 2009 7:11 PM:

5 years for a hundred million? Could we get ten years for two hundred million? It's a bargain compared to the trillion the politicians want. And in ten years it should be cold enough to give this foolishness up.

LAG said at December 13, 2009 8:31 PM:

"Obviously their models are far from complete."

That's the most magnificent understatement I've read today. Has everybody forgotten the work done in the 60s by Edward N. Lorenz? Their models WILL NEVER be complete. It is provably impossible. I can't believe this is still a topic of discussion.

Jeff Garzik said at December 13, 2009 10:49 PM:

Geoengineering gambles that fighting pollution with more pollution will produce some idealized climate state -- a wild shot in the dark, modifying one of the most complex systems known to man (our planet).

There is no practical way of testing large scale climate modifications in a safe manner, and the risk of even-worse consequences is very real. The only precedent remotely close to geoengineering is... the pollution we are already throwing into the atmosphere.

Outside of the engineering realm (if you can even call this engineering), it seems highly unlikely that 100% of all nations on Earth will agree to geoengineering, making it a legal impossibility. Therefore, any nations attempting geoengineering would, by fiat, be dictating climate policy for the entire world.

Brett Bellmore said at December 14, 2009 4:41 AM:

"it seems highly unlikely that 100% of all nations on Earth will agree to geoengineering, making it a legal impossibility."

Didn't know there was some law already in place making geo-engineering illegal without unanimous consent of all nations. Wouldn't that make excess CO2 emissions, themselves in effect a form of geo-engineering, also "illegal"? Actually, the chief advantage of geo-engineering is just precisely that it doesn't require the sort of world-wide agreement that CO2 control does; It thus escapes that prisoner's dilemma that makes CO2 control so unlikely.

Bob Badour said at December 14, 2009 10:34 AM:
For example, wind-driven misting machines off the west coast of Canada and the United States could cause more rainfall over North American farm fields.

Not bloody likely. The windward side of the mountains (if any) will get all the rain, and the leeward side will be as dry as ever. You would get lots more rain in Vancouver and Seattle, and the rest would fall somewhere in the Pacific. For example, in Kahalui and Hana but not in Lahaina and Kihei.

Nick G said at December 14, 2009 11:43 AM:

The last 10 years do not fit their models.

Are we sure of that? This seems to say otherwise:

"The United Kingdom’s Met (Meteorological) Office announced that the 2000-2009 decade “has been, by far, the warmest decade on the instrumental record”, and that 2009 is on track to become the fifth warmest year in the past 160 years, continuing the warming trend that has accelerated since the 1970s. 2009 has also been warmer than 2008, due in part to El Niño conditions in the eastern tropical Pacific Ocean."

Randall Parker said at December 14, 2009 5:12 PM:

Nick G,

If temmperatures went up in the 1990s they could flatten out in the following decade and still be considered quite warm.

Read your excerpt carefully. Do they say the 2000s have seen increasing temperatures? The sentences were well crafted to convey concern in spite of a flattening.

Now, if the Earth is already warm enough to gradually melt Greenland and Antarctica then we certainly should be concerned.

What I would like to know: Do we need to cool the Earth from where it is now in order to prevent eventual sea level rises.

Nick G said at December 14, 2009 5:54 PM:


The article says unambiguously that the 2000's were 0.17 °C significantly warmer than the 1990's:

"The 2000 – 2009 decade will be the warmest on record, with its average global surface temperature about 0.96 °F (0.53 °C) above the 20th century average. This will easily surpass the 1990s value of 0.65 °F (0.36 °C)."

Randall Parker said at December 14, 2009 6:05 PM:


You still aren't reading carefully enough. The 1990s were a decade of rising temperatures. So their average will of course be lower than the decade of the 2000s even if the latter has flat temperatures equal to, say, 1999. That doesn't mean the temperatures were rising in the 2000s.

The halt or pause in the temperature rise doesn't disprove the global warming theory. But it doesn't exactly support it at this time either. Now, development of a more complicated model that more accurately models the oceans or Sun might show that the last 10 years of temperatures are consistent with a long term warming. But that more complicated model hasn't been developed yet.

Nick G said at December 14, 2009 6:24 PM:


Hmm, yes I see what you mean.

Well, the question is: is this development a disproof to climate models - IOW, are climate models expected to provide predictions of temperature by year, or 5-year period? Scientists in the article you referenced say no (" "We have to explain to the public that greenhouse gases will not cause temperatures to keep rising from one record temperature to the next, but that they are still subject to natural fluctuations," says Latif. For this reason, he adds, it is dangerous to cite individual weather-related occurrences, such as a drought in Mali or a hurricane, as proof positive that climate change is already fully underway. "Perhaps we suggested too strongly in the past that the development will continue going up along a simple, straight line. In reality, phases of stagnation or even cooling are completely normal," says Latif.").

Are they changing their tune, or is it accurate to describe climate models as not predicting annual temperatures? If it climate models really aren't expected to make annual predictions, then this leveling off doesn't mean anything, even if it looks bad.

What do you think?

th said at December 14, 2009 6:27 PM:

Gee nickg, 2000 to 2009 actually had the lowest rate of temperature rise since the 1979 benchmark that began the quest for AGW substantiation, thats really good news for the polar bears and the salmon runs isn't it? Its kinda like a culture change for some as well as weather change, lets hope for more change!

Bruce said at December 14, 2009 7:18 PM:

One thing to remember NickG is that the climate temperature "hockey team" dropped 7000+_thermometers from their "average" starting in the mid 1960s.

For example, there are now 135 thermometers used in the USA. ALL OF THEM are in cities and at airports. I think what you meant to say was that the temperature at airports and in cities suffering from the Urban Heat Island effect went up.

You can't say the world temperature went up when 80% or more of the stations were dropped from the record.

A conspiracy minded person would suggest that CRU and GISS et al have cherry picked the warmest stations and left them in their climate record.

And using places like Darwin as an example ... they then "adjusted the hell out of the data" so it switched from a light cooling to warming.

The most you can say is that the graphs go up. I doubt the temperature actually went up much at all ... if any.

Bruce said at December 14, 2009 7:28 PM:

Sorry. Forgot the airport reference. It is 134 stations. All at airports. GISS is really just measuring airports. 121 out of 134 USA station in GISS are at airports.

NickG ... do you think airports give a good measurement of the USA average temperature.

Engineer-Poet said at December 14, 2009 8:14 PM:

The "skeptical" view of a continent with only one temperature measurement station (don't we have anything at Little America?) appears to be deliberately ignorant of the existence of satellites with IR and microwave radiometry sensors.  The GRACE experiment measures mass (including ice) directly, accounting for glacial outflow while not being fooled by floating ice.

LarryD said at December 14, 2009 8:33 PM:

The satellite temprature readings are calibrated against the ground data. If the ground data has been fudged, the satellite measurements are bad too. And the satellite data has had coverage "issues" for a while now.

In science, a theory is tested by seeing if it can make predictions that are correct. The climate models have never meet this standard. The technical term for this condition is "refuted".

Randall Parker said at December 14, 2009 8:34 PM:


Disprove the models? It depends on what you mean by disprove. We do not need to compare them to real temperatures to know they are in important ways wrong. They have large simplifications and do not model major elements of the climate. Remember that post I did several weeks ago about the nitrogen cycle being added to a model recently? Were all the models wrong before that? They didn't model a major factor. Surely they were wrong. This isn't by intent or incompetence. The modelers do not know enough to build an accurate model and even if they knew enough they probably wouldn't have a big enough computer to simulate the Earth at sufficiently fine granularity.

The climate scientists use models because they have only 1 real system (Earth) to work with, they can't intervene in it (no experiments let along repeatable ones), and Earth takes decades of watching to see what it is going to do. I do not blame them for this. We need to use the best tools available and they do. But accept that the best tools aren't very good. We need definitive answers. Well, we can't have definitive answers. We can't even have good probabilities. We do not know what we do not know.

Climate models might be accurate at the 100 year range even while being inaccurate at the 10 year range if the factors they do not accurately model tend to bounce around some average and if the assumptions in the models about feedback loops are correct. But since I don't have a time machine I can't check to see if that is so.


It is my understanding that satellites are calibrated using ground stations. So satellites are as good as what they are calibrated to.

Gotta say, while I think climate science is important I do not understand how good it is. I asked a Ph.D. physicist and he told me he'd have to spend a solid year getting up to speed studying climate science before he could pass judgment on how good it is and how much the modelers can project 100 years into the future.

Randall Parker said at December 14, 2009 9:11 PM:


To give you an idea of the limits on today's climate models:

One of the shortcomings of large-scale climate experiments is that they use relatively low-resolution computer models, Diffenbaugh said. He points to the United Nations Intergovernmental Panel on Climate Change (IPCC) as an example. "A typical climate model used in the last IPCC assessment had a 155-mile (250-kilometer) horizontal resolution," he explained. "However, the climate can vary quite a bit across that distance. My group takes the IPCC models and drills down to a 15-mile (25-kilometer) resolution."

Diffenbaugh's model also relies on data collected at intervals of 24 hours or less. "To really understand climate change impacts, we need to look at the sub-daily scale," he said. "In our simulations, we store data every three-to-six hours. That level of detail is not available from most global climate models, because it's very data intensive to store that much information."

In that report Noah Diffenbaugh at Stanford is running a relatively higher resolution model because he's only looking at the United States and only for 30 years. The resolution on these models isn't the only problem with them. They need calibration constants and models for all sorts of physical and biological processes. Well, in order to model those things they need to be well understood.

Bruce said at December 14, 2009 9:14 PM:

EP, the biggest difference between GISS and UAH (one of the satellite temp sites) by continent is Anarctica.

"GISS trend minus UAH trend, degrees C per decade" = .139C per decade for Anarctica.

"Those who want to argue that the surface temperature record should be used in preference to that of satellites need to explain why the three areas in which the two diverge the most are the three areas with the worst surface temperature data coverage. This seems to argue that flaws in the surface temperature record drive the differences between surface and satellite, and not the other way around."

Bruce said at December 14, 2009 9:18 PM:

For Anarctica ... "The first thing that stands out in the comparison of Antarctic datasets is the difference in the signs of the linear trends. The GISTEMP data show a positive trend of 0.048 deg C/decade, while the UAH MSU data show a negative trend, -0.091 deg C/decade."

Scott said at December 15, 2009 11:59 AM:

You want to give the west coast MORE rain? Clearly, you've never lived here… But even if that's your goal, I doubt any Pacific-based system would have any more luck getting rainclouds over the Rockies than Mother Nature does.

Other than that, very interesting post. I would argue that while attempts to specifically engineer the temperature without much more precise and accurate climate models may be unwise, we could take a step back in the process and attempt to counteract our carbon/methane/etc. emissions specifically, rather than the uncertain results of those emissions. The most interesting idea I've heard lately was to place simple wave-powered pumps in the ocean to bring nutrient-rich water from the depths up to where algae can feed on it, grow, and absorb the atmospheric carbon dioxide.

Hauser said at December 15, 2009 3:37 PM:

Go easy on them, guys. They still can't understand that the data has been mishandled. They still haven't learned that they can't trust the data. Until they get to that basic level of understanding, they'll just be "gaga, lala, mama" all day long.

Brett Bellmore said at December 15, 2009 4:08 PM:

Aerating the ocean by similar means might be useful, too, especially in areas getting too much fertilization from run off. You can view that as an excess of fertilizer, or a shortage of oxygen...

Randall Parker said at December 15, 2009 8:35 PM:


I live a couple of miles from the Pacific Ocean. But I do not live in the Seattle sub-tropical rainforest. Where I live it is dry for about 8 months of the year and sometimes longer.

As for getting water past the Rockies: Where does the rain that falls on the Great Plains come from? Probably from farther north. Maybe water sprayed into the air off the southern part of Alaska blows over the Great Plains.

Water sprayed into the air off of San Francisco would probably increase rains in the San Joaquin Valley. That'd help crops.

Bob Badour said at December 16, 2009 5:51 AM:
Where I live it is dry for about 8 months of the year and sometimes longer.

Yeah, because you live further south where the prevailing winds blow the other way. Those prevailing winds will blow the moisture out over the pacific where it might increase rainfall in places like Kahalui and Hana but not in places like Lahaina and Kihei.

Where does the rain that falls on the Great Plains come from?

What little there is of it probably comes from the Gulf of Mexico. Many of the storms that rain down on the Great Plains form over the panhandle of Texas then move north and east. But mostly the plains are pretty dry.

Water sprayed into the air off of San Francisco would probably increase rains in the San Joaquin Valley.

Not bloody likely--for the exact reasons already given. At best, they would help wash Pacifica into the ocean.

Steved Koch said at December 17, 2009 4:37 AM:

You would not necessarily have to spray the water into the air over the Pacific. You could flow the water through a pipeline through a river valley that cuts through the mountains. For example, the Columbia River valley cuts through the Cascades. On the west side of the Cascades, it is really wet (eg: Portland). On the east side it is really dry (The Dalles, Oregon is about 10 to 15 inches of rain/year and is 300 feet above sea level).

Given how dry it is east of the Cascades, it probably would not be necessary to spray the pumped water into the air, you could just fill a salt lake and natural evaporation would get the pumped water into the air.

Randall Parker said at December 18, 2009 11:08 PM:


North Dakota gets 15 inches per year which happens to be about what Santa Barbara gets. But the southeastern part of ND gets over 20 inches per year. Kansas gets 28 inches. Check out ND rain distribution in June. In April thru Sept SE North Dakota gets more rain than Santa Barbara does all year.

Prevailing winds: Actually, winds from landward aren't common in Santa Barbara.

Bob Badour said at December 19, 2009 9:16 AM:

Yes, further south and further east in ND is closer to the Gulf of Mexico and more likely to receive rainfall from a storm that forms over Texas. If you look, you will see the area you are talking about is just on the edge of Tornado Alley, and the super-cells responsible for the tornadoes generally form over the panhandle of Texas, as I mentioned previously.

If the prevailing winds in Santa Barbara come from the ocean, your idea would increase rainfall in Santa Barbara but not on the other side of the mountains.

Post a comment
Name (not anon or anonymous):
Email Address:
Remember info?

Go Read More Posts On FuturePundit
Site Traffic Info
The contents of this site are copyright ©