An In-Depth Overview of Key Technical Factors
Go here if you are looking for the easy 2-minute high-level summary instead.
Most geoengineering strategies focus on employing an array of exotic technologies or speculative scientific breakthroughs. However, seriously surveying the proposal geoengineering strategies [link] to reduce and reverse CO2 levels in the atmosphere reveals that almost all of the proposals are risky, both in terms of whether they will succeed and whether they could have negative unforeseen side-effects. Many are also prohibitively expensive, and large projects with significant unknowns usually have massive cost overruns.
However, one strategy turns out to be low cost, low risk, politically feasible, and can be implemented in a distributed manner by countries, corporations, communities, and individuals all over the world using relatively low levels of proven technology.
This strategy is the massive global reforestation of around 3 billion acres of land.
Tree-planting is old news. We’ve heard from environmentalists and hippies that planting a tree is good for the planet. It’s low-tech, it’s boring and dull, it’s not shiny and cool. We often imagine that if we can just come up with a clever new machine, it’ll solve our climate change woes.
The key issue is that when solving extremely large-scale problems, the basic “unit” of your solution must be very simple and low-cost, because all of the cost and complexity comes later, when you’re trying to do that basic unit a billion times. Scalability imposes its own unique technical challenges.
The key idea here is that while planting a single tree, or even a single forest, won’t put a dent in the problem — it turns out that there is a number of trees that WILL solve the problem. It’s very large, but it turns out it’s both do-able and affordable.
That number is 3 billion acres of trees.
Here’s the Math:
First, the basic unit of CO2 is a “ton.” A billion tons is a “gigaton.” Each year, the world emits about 45 billion tons, or 45 giga-tons of CO2.
This amount is rising by about 7.5% a year, but let’s focus on the 45 gigatons for now.
How much CO2 does an acre of forest absorb in a year? It turns out that a 50-year-old oak forest absorbs 15 tons of CO2 in a year.
This means we need 3 billion acres of mature forest to offset our annual CO2 emissions.
Pretty simple, but 1) we need to make sure we have enough land for that and 2) trees need water.
According to Drawdown.org, the total available land available for tree-planting is about 1.08 billion acres. That’s not quite enough. But, Drawdown’s number refers to the amount of land available for afforestation, which is the replanting of trees on land where there used to be trees but where they got cut down (“degraded grassland, cropland, and forest”).
What if we could find more land?
It turns out that on Earth, there are 4.7 billion acres of desert.
Can we reforest desert? It turns out that we can. There are numerous projects where barren lands have been reforested — for example in Spain, Jordan, Israel, and China. The latest example is the Kubuqi Desert in China, which reclaimed 1.4 million acres from the desert through reforestation. Small-scale projects have been successfully reclaiming land from the desert for decades. Now we need to do it at a massive scale.
Upon being planted, trees need a lot of water to get established, and then (if in desert regions) require irrigation for about 20 years. This ongoing irrigation is the main cost of reforestation. Is it affordable? It turns out the answer is yes, even in worst-case cost scenarios.
Because the world is already facing freshwater shortages, we cannot assume that existing freshwater supplies will be available for irrigating our massive reforestation project (existing water needs to be available for people and agriculture). Instead, the only way to guarantee a steady supply of freshwater is through desalination.
Desalination is energy-intensive, so we’ll need to build new energy production capability on top of what we have, and it will need to be low- or zero-carbon sources (or we are just undoing our own work). Solar power happens to be uniquely suited to this application.
Thus, we need to figure out the cost of building desalination plants, and supplying them with energy. Well, it turns out that the cost of producing 1000 liters (a cubic meter) of freshwater is about $1 at current technology/cost levels.
The yearly water requirement for an acre of trees is between 500–1000 cubic meters. Let’s take the higher figure.
This means that watering 3 billion acres of forest will take 3 trillion cubic meters of water a year, and at $1/cubic meter, that’s $3 trillion/year.
Can we afford it?
Yes. The world GDP in 2017 was about $80 trillion.
The one-time cost of initially planting each acre also happens to be around $1000/acre (based on the cost of the latest large Chinese projects), so we would anticipate a one-time cost of $3 trillion, and then subsequent ongoing expenditure for irrigation of $3 trillion/year for 20 years.
This is only about 4% of the world’s GDP, and it’s actually a worst-case estimate.
It makes very conservative assumptions:
All 3 billion acres are assumed to be reclaimed desert (because none of the owners of the 1.08 billion acres want to allow us to plant trees there). Reforesting degraded land is much easier because you may only need to irrigate for about 1–5 years just to get the trees established.
We assume that all water used for irrigation comes from desalination. Desalinated water is literally 10x to 100x more expensive than groundwater or other freshwater sources. If any regions can spare freshwater for irrigation, the cost drops dramatically.
No cost savings are assumed on solar equipment or desalination technology. All of these estimates are made using current technology costs, and given current trends (as well as the huge demand that would be generated due to such a massive program), it’s very likely that costs of equipment would drop considerably.
All forests are new forests. Every acre of forest that we grow on land that was previously forest is cheaper to grow. Every acre of deforestation that we prevent can be counted as an acre — and it costs “zero” to simply “not deforest” an acre. In fact, not-deforesting is absolutely the cheapest way to accumulate “acres of forest” towards our target of 3 billion.
The original core metric of “15 tons/year of CO2 per acre of trees” derives from forestry technology of the 1970s and early 1980s, and take no account of research since then in advance plant-breeding techniques developed through Green Revolution technology for food-crop plants. In fact, as early as 1985, plantations in southern Brazil were able to boost eucalyptus tree yields potentially as high as nearly 40 tons/year of CO2 per acre.
There Are Other Significant Benefits
The Sahara Desert was not always a desert. It used to be green. Many of the deserts of the world became deserts due to the activity of ancient humans and their grazing animals.
Biomes are self-sustaining. If you deforest a region and make it a desert, it stays a desert. If you reforest a region, it will actually sustain itself too: plants cool the atmosphere and bring water. This is why we only need to irrigate our forests for about 20 years — once they reach maturity, they will influence their local biome and bring self-sustaining rains. We would be reversing not just the problem of CO2 emissions from the beginning of our Industrial Revolution, but also undoing part of the ecological changes that our ancient ancestors brought upon countless lands (don’t blame them — they were just trying to survive. But today we know better, and we can do something about it) [link]
The reforestation and reclaiming of desert would also result in a massive increase in agricultural land. The 3 billion acres don’t need to be all contiguous. Forests could be sparsely laid out in alternating agroforestry plots of forest and agriculture: the local climate changes from the nearby forest would make the intervening agriculture plots farmable, thus massively increasing food production worldwide.
It’s safer than most plans if things go wrong
Any large-scale effort will have large unintended effects. The reappearance of forests in places that have been deserts for thousands of years could affect weather and precipitation patterns. While the effect is likely to be benign and gradual as the forest mature (especially compared to the dramatically negative effects of climate change now), this solution is extremely and easily tunable: if a particular forest is found to be causing a problem, we just cut it down. Humans are pretty good at cutting down a forest.
It’s distributed and does not require global consensus
Again, any plan of this scale would normally require reaching consensus from countries around the world, which is notoriously difficult. Many plans can be held up by a single recalcitrant actor. This one isn’t. If 50% or 80% of countries decide to do it, they can move forward without the agreement of others. If some of the land needed to comprise the final 3 billion acres sits within a country that doesn’t want to participate, it can be found elsewhere.
Moreover, governments themselves don’t even need to be convinced to participate. Within any country with sufficiently well-enforced private property rights, corporations, communities, or even wealthy individuals can simply purchase or otherwise set aside land to be reforested. Any acre of land with new trees on it that can be irrigated and protected for at least 20 years will help.
Limits To This Plan
The biggest limit is that in certain areas, the change in albedo due to reforestation will cancel out the cooling benefits of the CO2 that the forests sequester. Essentially, if the land is white enough (ice and snow), changing it to darker forests (green) will result in more energy absorbed from sunlight.
Fortunately, research has been done on this already and the regions where the albedo warming would exceed cooling due to sequestering carbon have been fairly well-established. These are, basically, Canada and most of Russia. So we wouldn’t be able to do any of this in those countries — and unfortunately, Canada and Russia comprise about 300 million acres of otherwise reforestable land.
Second, we can’t simply find a species of fast-growing tree and plant it everywhere. The earliest attempts at desert reforestation in China failed because they tried to plant non-native trees. The trees that tend to do best in any region are often the native species, because millions of years of evolution have led to an existing ecology of bacteria, insects, birds, and other plants that mutually support one another and aid the optimal growth of those trees.
Therefore, every single region needs a reforestation plan specifically tailored to its unique area. Local communities are usually the most knowledgeable about these specifics, and global engagement is vital because it combines a sense of community ownership over the forest with the global mission. Community ownership is important because the forests need to be preserved for decades (actually forever), so the people who live near them must understand their significance and value.
Thirdly, local politics influence the long-term viability of this plan. We don’t just need to reforest, we need to guarantee that once planted, the forest is irrigated and remains undisturbed for at least 20 years. Areas with uncertain civil governments or rule of law make this difficult, so care needs to be exercised in deciding where to replant forests.
There have been numerous existing efforts to encourage large-scale reforestation. The most well-known of these is the Bonn Challenge, a global set of non-binding commitments by countries to regrow forests in their own territory.
Unfortunately, the total amount of forests (i.e. assuming every single country fulfilled its commitments) under the Bonn Challenge was 865 million acres, well short of the 3 billion we would need. If the reforestation commitments made in the Bonn Challenge are all fulfilled, it will go a long way towards helping this plan, but we need to do more — a full 3 billion acres.
There are also numerous non-profit organizations throughout the world dedicated to reforestation efforts, sometimes on the order of hundreds or thousands of acres. These are most valuable because they represent decades of experience across dozens of different biomes where people have successfully restored ecosystems and brought back forests. We need is to collect all of that knowledge and then do this on a truly massive scale.
How Do We Do This?
The biggest advantage of this plan is that we are already doing it. It is low-risk, the technologies involved are well-understood and accessible, and it is fairly affordable.
We just need do a lot more. Not an unattainably large amount more, but a large number: specifically 3 billion acres.
We can get that land out of degraded grasslands, croplands, or denuded forests, and we can get the rest of by reclaiming deserts.
We need the participation of not just governments, but also corporations, communities, and individuals. Anyone who can secure land and ensure that it can be irrigated for at least 20 years and protected from logging or grazing animals can contribute. However, always remember: one big advantage is that we also don’t need the participation of any one particular government, corporation, community, or individual.
We need cost improvements in solar energy generation and desalination efficiency.
The biggest risk is that this plan just isn’t seen as dramatic or fancy enough. But it will work. It is based on sound science: we know how much CO2 trees sequester, we know how to plant trees and water them, and we know how to desalinate water. Those are relatively simple things, and if we can do 3 billion acres worth of it, we will solve the problem.
As a cost-validation pilot project, we have begun reforesting 45 acres of desert on the Big Island in Hawaii, irrigated using desalinated seawater powered entirely by a large solar array.
This facility uses the latest solar and desalination technology, and concretely validates the unit costs of the technology, as well as establishes a worst-case cost ceiling: the Big Island is pretty much the worst and most expensive place to build anything. The corner of the island where we’ve located the facility gets very little natural rainfall (12–15 inches/year), so if it can be done there, it can be done anywhere.
If you want to read more, here are the details.