Filed under: Geoengineering, Global change, Interventions in the carbon/climate crisis, Trees
An interesting paper in Climatic Change: Irrigated afforestation of the Sahara and Australian Outback to end global warming by Leonard Ornstein, Igor Aleinov and David Rind Doi: 10.1007/s10584-009-9626-y. (Mason Inman has a nice write up with some background and comment over at ScienceNow; [update] and corresponding author Len Ornstetin chronicles the idea’s rocky research road on his own site). The central idea is that with enough irrigation you can turn big deserts into big forests: forests big enough to suck up a large part of total carbon dioxide emissions for decades or even centuries. I think that you can take this notion as a serious plan, a thought experiment, a jeu d’esprit, a warning or a jumping off point, depending on predisposition. Aspects of all that in what follows.
Here are the basic numbers: The Sahara is about a billion hectares in area, on which you could fit a trillion eucalyptus trees. Those trees, if working flat out, could each put on twenty kilos of biomass a year. If roughly half that biomass is carbon, that would mean a net annual sink on the order of ten billion tonnes of carbon. That’s about the amount that humans currently emit.
To create such a forest in a century, you would have to plant as many hectares of trees every year as are currently lost to deforestation worldwide. And, even harder, you’d have to provide them with what they need to order to grow. You need a great many things to turn a desert into a forest — soil nutrients, microbiota, possibly pioneer plants, a compelling reason for doing the work, and so on — but the biggest hurdle, pretty obviously, is water. Eucalyptus, the authors say, needs about a metre of rainfall a year. For a billion hectares, that’s 10 trillion tonnes of water. The authors assume, reasonably for all that I know, that if you have smart irrigation getting the water to just where it is needed you can get away with half that amount. Even so, even the vast aquifers beneath the Sahara don’t contain the amount of water required, so it will have to come from desalination plants on the and be pumped it up to where it is needed (the average elevation of the Sahara is about 450m). The size of this undertaking — more than 50 new Niles, flowing in reverse — may explain why the authors feel they need to use that fine old-school term “terraforming” for their undertaking. The power requirement, if I’m reading their figures right (4.04kWh/m^3 fresh water delivered), is a bit to the north of 2.2 terawatts, about 40% of it for desalination by reverse osmosis and about 60% for pumping.
The world’s electricity generators currently provide about 18,000 TWh of energy, which averages out at 2TW of constant supply. So in energy terms the desalination and pumping needed for the Sahara forest would use a bit more electricity than the world currently generates for every other purpose. This unavoidably sounds nutty. But that is at least in part because of the nuttiness of the situation, rather than its proposed solution — the nutty situation in which we burn fossil carbon at tens or hundreds of thousands of times the rate at which it is sequestered over geological time. If humanity insists on putting so much carbon dioxide into the air every year that it would take a brand new forest the size of the Sahara to suck it all up, then that’s where the madness starts. That creating such a forest would have to be a large undertaking — large in terms of the whole world economy — is just a consequence of the initial folly.
And in practice the investment would be smaller. A nice thing about forests is that they can go some way to creating their own weather, and the authors have looked at this effect with some climate modelling work. If a forest with irrigation dampened soil is imposed on the Sahara, rain begins to fall, in some places as much as a metre of it every year. This rainfall doesn’t obviate the need for irrigation, because it is strongly seasonal — basically an extension of the West African monsoon of April to November. But it might significantly reduce the irrigation requirements. Maybe you could get away with just a terawatt…
The Sahel, to the south of the Sahara, also gets damper in those enhanced and extended monsoon rains, which is definitely a plus, I’d guess, and the African Easterly Jet, a feature which is driven in large part by the temperature contrast between the desert and surrounding land, seems to more or less vanish. Since a large number of Atlantic hurricanes get their starts as kinks in the AEJ, that might be a pretty significant change, too. Beyond that, the rest of the world seems pretty much unaffected. In particular, the authors say that their models show no additional warming that might be laid at the door of the change of albedo which comes with replacing light desert with darker trees. (I think this fits with the 2007 Bala et al paper in PNAS, which suggested that warming associated with afforestation would be due to changes in boreal, rather than tropical, forest cover).
There is, however, a fly in the ointment. The Bodélé depression in Northern Chad is only a small part of the Sahara, but it is the world’s greatest source of mineral dust, with the winds drawing some 700,000 tonnes a day off the surface. According to Koren et al in ERL, 2007 40 million tonnes of dust a year travels from the Bodélé to the Amazon rain forest, half the total annual mineral inputs into the forest basin (the dust fertilises the mid Atlantic, too, and it may play a role in abating hurricanes too — Jim Giles wrote a lovely piece on this for Nature some time back). There’s a real chance that this dust is crucial to maintaining the soil fertility of the forest, and even if the Bodélé itself were left unirrigated and unforested, the increase in precipitation all round it, and the wetter atmosphere downwind of it, would probably shut it down as a dust producer. If growing a forest in the Sahara hurts the one we already have in the Amazon it obviously becomes a less attractive proposition (though if we are going to lose the Amazon forest anyway, things might look different…). That said, if you are pumping trillions of tonnes of water across continental scales, thenpaying to air dump a few tens of millions of tonnes of fine-particle mineral fertiliser upwind of where you want it is hardly going to break the bank.
Something the authors don’t look into is that the higher the CO2 level in the atmosphere gets, the easier this all becomes. Higher carbon dioxide levels make plants more water efficient, all other things being equal. All other things are not, necessarily, equal — higher CO2 also makes things hotter, which plants don’t much care for. In a world with some solar radiation management, though (such as aerosols in the stratosphere) all things might indeed be kept equal, or at least temperature might be. Martin Claussen has been working for some time on the idea that the Sahara is a “tipping element” in the climate regime, one that can be pushed from a dry state to a wetter one relatively easily. In a more carbon rich but not-too-hot world the circumstances might be right for it to tip the other way, and it might take rather less than a 50-Nile terraforming project to nudge it over.
In the final analysis, I don’t think I take this paper very seriously as a practical proposition. Doubling global electricity generation for a single project seems far fetched. For such a thing to be put anywhere near the top of one’s list of African infrastructure investments would require that a great many other large and important development initiatives (provision of power, water, roads, cold chains, vastly improved agronomical advice, etc to the vast majority of the population, for starters) would already have had to have been put in place. But it’s kind of nice to imagine a world in which we were wealthy and together enough to have actually taken the pressing need for those changes to heart, and were thus in a position to consider greening a great desert too.
And regardless of practicalities I think there’s real value in taking the analysis further. A big idea like this throws off many fascinating questions that force you to look at the earth, and what we know about it, in new ways (or old ways but with a new twist):
What polycultures would you build the new forest with? (all-eucalyptus-all-the-time is fine for first calculations, but doesn’t sound like anyone’s idea of a proper landscape. Baobabs? Laurels? And what fauna might be good, or bad?)
What genetic engineering — reduced flammability, higher albedo leaves, more refractory soil carbon, who knows what else — might help?
How much bioenergy with carbon capture could be built into the scheme, perhaps initially to power some of the inland the pumping stations?
Can biochar help? (and a million other soil-creation questions)
What are the best silvicultural ways to make the new woodlands pay, as that is something people by and large like their environments to do, and can there be room for some agriculture too?
How could local people best be convinced this was a good idea? And what are the property title reforms that would be prerequisite?
If the AEJ stops, do hurricanes stop too? Or does some other mechanism initiate them, maybe somewhere else? And does the dust really have an effect?
When the Sahara was wetter and less dusty in the past, did the Amazon actually suffer from lack of nutrients? (I think there is actually some research already out there on that — but can’t offhand think where)
How can the transformation be made stunningly beautiful?
What regions and landforms do you want to keep as monuments/heritage sites/national or world parks? There would undoubtedly be a real aesthetic/biodiversity loss in the removal of the desert, not to mention risks to some utterly wonderful buildings.
How to stop the Fremen becoming soft and decadent now that Arrakis has become a land of milk and honey?
and so on.
In particular, it would be nice to see some analysis of halfway houses; where in the Sahel and points north might merely huge, as opposed to planet-sized, afforestation be attempted, and what would be the costs and benefits? It is possible to transform land on very large scales, if not quite this large: 40m hectares of the Brazilian cerrado have been brought into agricultural production over the past fifty years. Can afforestation/silvicultural interventions on such scales ever make sense? And where else might be suitable for such things?
And on the topic of where else: My apologies to any Australian readers for not going into the paper’s analysis of foresting the Outback in addition, or as an alternative, to the Sahara. Basically the arguments are largely the same but the costs and effects are a bit smaller. There’s also a risk of interfering with El Nino that would definitely merit further attention. If anyone wants to blog more on that aspect of the subject send me a link and I’ll post it up here.
Image credits: Eucalyptus trees at the top from Big Lands Brazil, who would like to sell you some…; Bodele from Charlie Bristow, reused with permission; Tree of life from Flickr user Solvo under Creative Commons license
7 Comments so far
Leave a comment