The best kittens, technology, and video games blog in the world.

Thursday, November 05, 2009

Inevitability of geoengineering

Cat and fire by Michel Filion from flickr (CC-BY)
I want to write a quick post showing some arguments about inevitability of large scale geoengineering which I find extremely convincing. I'm sure I won't convince that many people, most have made up their minds already.

First, look at list of countries by CO2 emissions. In 2006 global emissions were 28.43 billion tons per 6.55 billion people, or 4.34 ton/person. World population is still growing and is estimated to stabilize at levels around 9.5 billion for 2050-2100, which is the time in question.

Now let's see what would happen if everyone got to current European standards of living, and European CO2 emissions, which are 7.84 tons/person. This would result in CO2 emissions of 74.5 billion tons, or 2.6x times current levels. This number is already quite optimistic - it makes a huge assumption that all countries with higher CO2 emissions like US, Japan, Canada, Australia, Russia etc. scale them down quite considerably.

Now a brief discussion is in order on why numbers like that are relevant. First, I assume that there won't be a massive genocide, or nuclear warfare, or pandemic, or anything else that could drastically reduce population levels. The last time there was a significant reduction in population levels was 1340s. None of the world wars, or any other events even stopped population growth - in fact people tend to reproduce a lot more during times of crisis, than during times of peace and prosperity, so it would need to be the biggest war, pandemic, or other kind of failure in history to significantly reduce population levels. This assumption is probably the safest.

The second assumption is that the rest of the world will grow quite quickly. World GDP/capita is currently $9,774 PPP. EU GDP per capita is $30,513 PPP. If world economy grows at per capita rate of 2.8%, then by 2050 world average GDP will reach current EU levels. This coincidentally is almost exactly annual per capita growth rate of the world economy of the last 30 years. Unlike with population, where predictions are quite solid, here there are more reasons why growth could be significantly slower, or significantly faster than predicted. We just had a completely unexpected global recession for a fairly obvious example. And not so long ago we had massive completely unexpected boom in China and then India, so it can go both unexpectedly badly and unexpectedly well. Still, 2.8% is a pretty decent first estimate.

I want to write a bit more on the subject of economy. Many people are sceptical about the idea that the "poor" countries will ever attain levels of economic development seen in rich countries. But the rich-poor divide is illusory. As you can see yourself with Gapminder, the divide is long gone and most countries are in the middle. World average life expectancy today is higher than in most richest countries in 1950s years ago. World average life expectancy in all but one countries is higher than world average 100 years ago - all that in spite of wars, malnutrition, malaria, HIV, and lack of clean water in many of them! Life in countries considered by many to be "permanently poor" is vastly better than it was in Europe of your grandfathers, and rapidly improving. To think it will somehow suddenly stop is to ignore history.

Development can happen extremely rapidly - world's largest CO2 emitter has been China for a few years now, per capita emissions of which increased by factor of x2.2 between 1990-2006. Likewise for India, growth was x1.6 during that time. Even for big countries not typically associated with rapid growth you have high growth - x4 for Vietnam, x2.4 for Thailand, x2.3 for Malaysia, x1.9 for Indonesia, x1.5 for Pakistan, x1.4 for Brazil and Nigeria and so on (and these are all per capita rates, their total emissions increased far more). Developing countries are catching up fast.

Let's look at past performance of world CO2 emissions. Between 1992-2007, world's emissions increased by 38% - mostly driven by rapidly developing countries like China and India. Even developed countries weren't terribly successful - US and Japan had significant increases in emissions, and EU's stayed pretty much even, and nominally decreased mostly due to early-90s' collapse of heavy industry in former communist member countries. Even if developed countries got their act together and implemented reductions that were agreed on in Kyoto Protocol, they would still be swamped by just Chinese increases, resulting in net emission increases.

And the outlook for the next couple of decades doesn't look terribly promising. Intrade markets, thin as they are (and forgive me lack of links, but their website is virtually unlinkable), think it's extremely unlikely that China and India will agree to any CO2 limits anytime soon, and have pretty low opinion on what reductions developed countries will agree to - chance of agreeing to 10% reduction by 2025 seems to be about even, and there's quite a big difference between agreement and implementation (especially due to possibility of accounting shenanigans of land use changes and such), and such reductions would be swamped by developing countries increase. And in any case the 74.5 billion tons number already assumes that developed countries average will get reduced to current EU levels.

If you hope on peak oil to make 74.5 billion tons/year kind of number impossible, not only peak oil doesn't seem to be happening, coal production grows insanely fast, and estimates of "peak coal" talk about mid 22nd century.

Alternatives to 74.5 billion tons a year

So with business-as-usual, we'll be emitting enough CO2 to get the entire planet stir-fried. What would be alternatives?

First, either outright genocide or indirect genocide by limiting economic growth of developing nations - remember that poverty kills hundreds of millions of innocent people and causes vast suffering. Not only is it unlikely to happen, it would be probably the worst scenario of all.

Second, learn to live in a warmer world. This isn't necessarily going to be as bad. It's hard to imagine anyone who would like to have their children die of malaria and diarrhea just to satisfy some rich smug liberals. People who measure costs of limiting CO2 emissions usually focus on limits in States, or EU - but this is nonsense! Big increases don't come from SUVs, but from bringing poor people to more civilized standards of living, and limiting that would be disastrous. Costs of just taking global warming would be much less than the genocide that would ensue to keep poor people in their place. I think this scenario isn't entirely implausible - there were quite a few cases in history where international cooperation completely collapsed, and in externalities cases like CO2 emissions, this is a possible result.

The third solution, and we're getting somewhere now, would be to sever the link between GDP and CO2 emissions. This is surprisingly hard. Yes, it's trivial to move from Hummer to Prius, but the problem is people who cannot afford either yet, but will want some means of transportation eventually. And they will want electricity - coming from coal most likely. The alternatives look pretty bleak - nuclear power, and large-scale hydroelectricity, both provide only 3% of total primary energy use now (or about 15% of electricity each). All fancy types of renewable energy together don't even add up to 1%. A surprisingly large amount of energy is "traditional biomass", or wood and agricultural waste used for heating and cooking (but then, estimates of it seem to vary wildly). Other than that, right now we're entirely dependent on fossil fuels.

A popular idea of high standards of living without corresponding increase in energy use seems pretty unlikely to me. Yes, we might be able to break the correspondence at some point, but energy use per capita in poor countries is so low, it's simply bound to drastically increase.

Much more promising path is energy generation without fossil fuels. Our best chance so far was with nuclear power. Unfortunately we suddenly stopped what looked like extremely rapid exponential growth in late 1980s after Chernobyl disaster - estimated 4000 people will die due to increased radiation exposure (much higher estimates based on linear no-threshold model can be safely ignored, as LNT model is known to be empirically completely wrong).
By ridiculously naive estimations, if this growth continued exponentially as it seemed to have up to Chernobyl, nuclear energy would provide around 20% of total energy production today, or half of all oil could be eliminated. Now this kind of extrapolation is ridiculously naive, but you should give it some thought.

So could we switch to nuclear at ridiculously rapid pace? We would need over 50 times as much nuclear capacity to generate enough energy to cover increased energy use by 2050, what seems ridiculously optimistic by any standards, not to mention nuclear proliferation issues. It looks somewhat less ridiculous when expressed as 10% a year growth (or doubling every 7 years like pre-Chernobyl), but it's just not going to happen.
There aren't that many renewable options. Current global energy consumption is 15 TW, and it will be about 39TW by 2050 by extrapolation. The only large scale renewable energy source we use, hydroelectricity, is nowhere close to our needs, and neither is geothermal really, if we look at what small fraction of it we can practically extract thanks to high local concentrations. We're left with just two then - wind, and solar. We would need to use 5% of all wind energy of Earth to fill the energy gap.

In 2008 wind power had 121GW of installed peak "capacity", and produced 260 TWh of energy (what corresponds to 30GW of actual production). It would require order of 1000x increase in wind power, or 19% a year growth, to make it fill the energy gap. And the actual growth is actually fast enough at 29% in 2008. But there are clear diminishing returns here - best locations will be taken, and massive subsidies cannot really scale to TW levels.

The other alternative is solar power. Just Earth-based solar is so plentiful that using even 0.02% of solar power now, or 0.045% of it in 2050, we can replace all of world's energy use. And it doesn't stop there, with space-based solar we can get numbers many many orders of magnitude larger. It would sort of need a space elevator, but we might get there by 2200.

Anyway, while solar is undoubtedly our long term future, it doesn't look that bright in the short term. There's only 15GW of solar thermal installed, and mere 0.6GW of solar thermal. And these are capacity numbers. We're talking x10,000 increase to cover the energy gap, or 25% a year increase, or x200,000, 35% per year for solar thermal. Now these numbers are actually what we see now, but as I said for wind, these are all based on massive subsidies, and there's no evidence that they're scalable to TW range at all.

By the way all this is based on assumptions that all energy is basically equal, and all greenhouse gas emissions is energy production. Neither of these are true - electricity is relatively easy to replace with renewables, at least if we ignore their intermittent character, but transportation much less so. And even with 100% solar electricity-based economy, there would still be some significant sources of greenhouse gases like methane from agriculture - so whatever we estimate here is bound to be overly optimistic.

With all this said, thanks to better energy efficiency, and widespread use of renewables, it's quite likely that humanity energy use by 2050 will be much lower than baseline number of 74.5 billion tons a year. On the other hand expecting it to be vastly lower than current emissions is just extremely unlikely.

And here comes the final solution - without genocides, without unexpected revolution in either nuclear power or renewables, we will have much more CO2 in atmosphere in 2050 than now. And due to massive lag it has, it will keep warming Earth for a very long time after we finally get our emissions to decrease. So if we cannot agree to live in warmer world, we have to geo-engineer to survive the next century without massive and irreversible environmental change.

Nature did some real life large scale geoengineering testing for us, and it works without major side effects (you probably haven't even noticed the experiment in 1991). Economic estimations put it in ridiculously cheap range - some schemes are estimated to cost just 375 million dollar a year, or just 17,000 Toyota Priuses (seems like a highly relevant unit of measurement) - for comparison replacing every American car with a Prius would cost 5 trillion USD, and it would still do very little towards limiting global CO2 emissions. And stratospheric sulphur isn't the only solution, just the one we have best data on (by the way it's not about polluting more, only about moving less than 1% of existing pollution to higher layers of atmosphere).

So the alternatives are:
  • Outright genocide
  • Indirect genocide by forced poverty
  • Accepting the warming
  • Increase in nuclear or renewable energy far beyond anything that can be realistically expected in this timeframe
  • Geoengineering
Guess which one is the least bad?

Now I just need to blog on abortion, and I win blogging controversy bingo.


Anonymous said...

The latter part of your statement "Nature did some real life large scale geoengineering testing for us, and it works without major side effects..." cannot be substantiated. On the contrary, there is evidence that putting sulphur particles in the upper atmosphere could well have significant side effects. I suggest you read the Royal Society report on Geoengineering to be better informed on the subject. See also this blog that reports on the Royal Society report - Note also that none of the Solar Radiation Management techniques address the issue of ocean acidification resulting from high levels of CO2 in the atmosphere.

taw said...

Well, I cannot really read it as their server is down. I might try again later.

But in any case - ocean acidification is purely hypothetical issue, and I'm not really sure why anybody would care about it, relative to let's say children dying of malaria right now, because they're too poor to buy bed nets.

And remember how it works. Peak warming happens much later than peak CO2 concentrations, peak CO2 concentrations happens much later than peak CO2 emissions. The world will not start cooling just because we reduce emissions.

I argue in this post that peak emissions are unlikely to occur until significantly after 2050 (due to economic development of poor countries, which means people won't be dying of lack of bed nets and clean water like they do now), but even in the unlikely case the world manages to peak in let's say 2020, peak concentrations will happen far later, and peak warming even more in the future. So we might need geoengineering no matter what.

Also there's a small possibility of major temperature-mediated positive feedback loops for warming. If raising temperatures melt ice and decrease planetary albedo, or massive amounts of methane get released - all such problems can be averted by solar radiation management.

Quickshot said...

Sadly the ocean acidification isn't theoretical, the PH is already starting to get threateningly low. Enough so that it's hurting the corals among others. If it continues to far we'll even kill the fish eventually and large classes of the oceans primary producers.
While this isn't exactly a lethal problem for us, it is a massive extinction event all of its own and would do quite some damage to oceans productivity for humanity.

As for feedbacks, well the Arctic is already thinning dangerously, which is a rather large surface area. So that's something to worry about, expecially considering that we probably won't manage to sufficiently limit emission. Even if we take considerably more optimistic estimates then used by Taw, we'll probably end at much higher concentrations and it is almost for sure we'll atleast trigger one major positive feedback loop due to this, namely loss of the Arctic sea ice in summer. With a secondary concern being that if the Greenland ice sheet starts losing altitude, it will melt faster as it continues to descend in to ever warmer air, from it's previously mountain like altitudes.

Which probably leaves us to the same basic conclusion even using fairly optimistic estimates, that we're probably not avoiding some use of Geoengineering.

taw said...

Quickshot: By "theoretical issue" I don't mean it's not happening, or that some ridiculously high amounts of acidification wouldn't cause significant damage.

I mean there's very little evidence that levels of acidification we can realistically expect are likely to reduce primary production significantly. Changes in species composition will happen, but we've been doing that since Lower Paleolithic, so it's not like it will be unusual.

Here's relevant Wikipedia:

> However, some studies have found different response to ocean acidification, with coccolithophore calcification and photosynthesis both increasing under elevated atmospheric pCO2, an equal decline in primary production and calcification in response to elevated CO2 or the direction of the response varying between species.

Also, how fast is ocean pH response to CO2 concentration? I would guess very fast, so peak acidification time will be pretty much the same as peak concentration time, and much earlier than peak warming time. Therefore, it's an easier problem.

And even back in 2004 32% of world fish production was aquaculture, with sustained long term 8% annual growth rates, while wild fisheries catch has been flat. By naive extrapolation they'll get equal around 2014, and aquaculture will match current total fish production by 2020, so I'm really not all that worried about reduced wild fish catches by 2050.

Quickshot said...

If I remember correctly response time is actually not all that fast, so peak acidification should happen well after peak CO2, but I imagine well before peak temperature still.

And yeah, some species will jump in when other species fail, but it most likely won't be the setup of life we knew for the oceans. But yeah, won't be the end of the world, just not very nice.