Nature based solutions to climate change?

Nature based solutions are likely to deliver c20-25% of the decarbonization in a realistic roadmap to net zero. Reforestation is low-cost (c$50/ton), technically ready, practical and helps nature. Key challenges are improving the quality of nature-based CO2 removals and accelerating momentum. We see opportunities for companies that tackle these challenges. Our top conclusions into nature-based solutions to climate change follow below.


(1) Deforestation and land use changes have emitted over 1 trillion tons of CO2 since pre-industrial times, as 5bn acres have been deforested (data here). That is 25% of total net anthropogenic emissions. Today, deforestation continues at a pace of 12M hectares per year (30M acres per year), emitting 6GTpa of CO2e, 12% of the global total, equivalent to all the world’s passenger cars (data here). There is clearly no route to net zero without some renewed balance with nature in the 21st century (note here). And more optimistically, photosynthesis fixes 700GTpa of CO2, compared to anthropogenic CO2 emissions of 50GTpa, so small changes in the natural carbon cycle can have large impacts.

(2) What are CO2 removals? Land that has been deforested will store minimal CO2. Mature forests store an additional 200 tons of CO2-equivalents per acre. Sometimes over 400 tons of CO2-equivalents per acre. Hence over a 40+ year growing cycle, trees take up an average of 5 tons of CO2 per acre per year (data here). The growth rates are sigmoidal (note here). Then ultimately mature forests sequester minimal amounts of CO2 as mature trees grow 40-50% more slowly and biomass decomposition matches new accumulation. CO2 is removed when land moves from ‘not forest’ to ‘forest’ and then stays forested.

growth rates forests during reforestation follow a sigmoid function sequestering CO2 slowly then quickly then slowly again

(3) Residuals and convenience. Many industrial processes are complex and competitive. Look at the charts for how the world makes carbon fiber (e.g, for wind turbine blades) or polyester (for textiles and solar backsheets), or AI chips, or any of the other metals and materials that explain 40% of all global CO2 emissions; and are needed to build other decarbonization technologies themselves, such as wind, solar and electric vehicles. In our view, there will always be some residual emissions in industrial processes. It will be an order of magnitude more practical to offset these emissions in the forest rather than in the factory. Many industrial companies will be delighted to pay for this service, rather than having to ‘reinvent and retool’ (examples here).

(4) Costs of CO2 removals. Ultimately, companies that offer the best decarbonization product at the lowest price will “win”. Our best model for the costs of reforestation is here, and sees an average CO2 removal cost of $50/ton. We also reach similar numbers when modeling mangrove restoration. $50/ton is feasible. And a long way below high cost technologies, such as DAC credits at $200-1,000/ton (how much do you have to hate nature, to pay $1,000/ton for a DAC credit rather than $50/ton for natural remediation?!). But CO2 removal costs in forests are also variable and can range from $3-120/ton. Economics hinge upon land costs, seedling costs, planting density, yield class, timber product prices. Over time, experience curves will apply in forestry, and many variables can be optimized.

(5) How much land for reforestation? Out of the world’s 40bn acres of land, we think 3bn acres can be reforested (note here, data here). To be realistic, this estimate does not include core croplands, inhospitable climates (e.g., Antarctica), or deserts (although at the upper end of the cost curve, we have explored using waste water in semi-arid petroleum basins, note here). The best land to reforest is drawn from the world’s 5bn acres of degraded-abandoned land and 8bn acres of pasture land. 47 emerging world countries could uplift GDP by 6-60% as they reforest 1.5bn acres via adopting ‘reforest and reinvest’ as an economic development model (note here). For another case study that is superb, nay sublime, look at Finland, which has offset two-thirds of its fossil emissions over the past century (note here). Reforestation potential by country is screened here.

(6) Quality and trust. Historically, carbon offset projects have not had sufficiently high quality, in our opinion. If buyers of carbon credits cannot trust a project, then its future carbon credits will end up heavily discounted. Possibly even valueless. For the highest quality projects, which will attract the highest CO2 prices, there must be no doubt that the project is real, incremental (CO2 removals not avoidance credits), its CO2 must be correctly measured, it should enhance biodiversity (which can even boost CO2 uptake by 15-70%) and the permanence of the CO2 storage should be enhanced by guarantees and buffers.

(7) Where is today’s market? In 2022, Thunder Said Energy contributed $7,700 to nature-based solutions projects, to offset its own CO2 emissions 20x over, studying half-a-dozen projects in detail, using a five-point framework. Today’s market is evolving painfully slowly. It lacks depth and policy support. High-quality credits are most bottlenecked. This makes us wonder whether early adopters will achieve higher returns. Our best note summarizing these findings is linked here.

(8) Controversies? The biggest controversy in today’s carbon markets are over permanence, and our best research note is here, comparing the total CO2 balances of forest products depending on their sources and uses. 1GTpa of timber is still burned for fuel, including biomass and BECCS, which is obviously not permanent. Pulp and paper also has low permanence. Construction materials have high permanence. The oldest wooden structure still standing is Japanโ€™s Horyuji Buddhist temple, constructed in 639AD. There is buried biomass that has lasted 45,000 years (note here). And studies show that 80-95% biomass carbon can remain preserved after 40-100 years in landfill (chart below, data here). There are also environmental controversies, due to forest-climate feedbacks. There are increasing risks of forest fires in fire-prone regions, which need to be managed and ‘buffered’ when booking carbon credits. But forests may also grow 25% faster due to high atmospheric CO2. And enhanced cloud seeding should offset forests’ low albedo.

(9) Other nature-based solutions. Further running room in nature-based CO2 removals includes farming carbon into soils via conservation agriculture (note here), carbon offsets in the ocean (note here), blue carbon (here), savanna management (note here), biochar (note here) and most controversially, ocean iron fertilization (“give me half a tanker of iron and I will give you another ice age”) (note here). Our roadmap to net zero derisks 15GTpa of CO2 removals from forests, 2GTpa from agriculture, and 600MTpa from biochar.

(10) Nature-based CO2 companies. Companies and charities offering CO2 offsets are screened here. Companies measuring forest and soil carbon are screened here. Arborgen is a listed seedling producer. Stora Enso makes sustainable materials from timber products. Cross laminated timber can be used for large construction projects, lowering CO2 by 15-80% compared to steel and cement. Tricoya is a listed mid-cap making a long-lasting engineered timber product. Tigercat is a leader in forestry machinery. More exotically, we have screened biotech and agtech companies increasing the CO2 uptake rates of plant species. Biochar companies are here and aquaculture companies are here. We also think large beneficiaries from nature-based CO2 will include industrial incumbents. Energy incumbents may commercialize zero carbon fuels by offering high-quality CO2 removal credits in conjunction with carbon emitting fuels (note here); and investment firms may offer carbon-neutral funds that dedicate a share of dividend proceeds to abating the look-through CO2 of the funds’ holdings (note here).



Nature-based CO2 removals: a summary?

Overview of nature-based CO2 removal

This data-file is an overview of nature-based CO2 removal projects that we have been supporting at Thunder Said Energy. Our research ‘scores’ different nature-based projects on a 100-point scale, using criteria to check whether they are real, incremental, measurable, permanent and bio-diverse. The average project supported so far scores 70/100 and sells CO2 offsets at $5-50/ton.


In 2022, we spent $8,000 to support five projects, which have most likely ‘credited’ 480 tons of CO2, for an average cost of $16/ton. Projects span across Costa Rica, Nicaragua, Kenya, Uganda, Indonesia and Madagascar.

The average nature-based reforestation initiative that we supported in 2022 scored 70/100 on our framework for assessing nature-based CO2 removal projects, and was priced at $17/ton of CO2.

Two of the projects scored over 80/100. Whereas three of the projects were given lower scores, due to question marks around whether they were fully incremental, fully measurable, or fully bio-diverse.

Overall we were least concerned about whether the projects were real, as most of them were issuing CO2 offsets that had been certified by Verra or Gold Standard, independently audited and with detailed documentation.

Overall we were most concerned about whether the projects were permanent, in turn a good reason to consider complementary solutions such as CCS and DAC projects?

Statistical distributions are also explored in this data-file, as there are clearly going to be ‘uncertainties’ in natural remediation projects: both implementing the projects over 40-year timeframes and quantifying the CO2 benefits.

The statistical distributions of nature-based CO2 removals are not normally distributed. We estimate our own probability distributions in the data-file. More on CO2 measurement in our allometry research.

A Monte Carlo approach can be used to quantify nature-based CO2 removals across a portfolio. Overall, we are 75% confident that the projects we supported in 2022 have offset over 400 tons of CO2, and 90% confident they have offset over 300 tons of CO2.

You can download this data-file for an overview of nature-based CO2 removal projects we have supported to-date. Or see our nature-based CO2 removals category for full details on the underlying projects.

Reforestation: what planting density for seedlings?

planting density for reforestation projects

What is the typical planting density for reforestation projects globally? This matters as it can determine the costs of reforestation. Hence in this data-file we have collated data from 25 different case studies globally, which have tended to plant a median of 670 seedlings per acre (1,650 per hectare). However, the range is broad, from 400 seedlings per acre in low-density Southern US forestry to 4,000 seedlings per acre in mangrove restoration projects.


Planting density depends on forestry practices, because as trees grow, they compete more for light, water and nutrients. Hence starting with a higher stand density will tend to require more thinning to avoid over-crowding and weak trees. Conversely, starting with a lower stand density will under-utilize valuable land in early years.

Planting density depends on forestry objectives. Dense stands may favor early harvesting, which is fine when growing crops for pulp, paper, wood-based fuel or lower-value sawlogs. However less dense stands may favor longer growing cycles, to produce high value timber, locking carbon away in long-lived construction materials.

Planting density depends on species. For example, mangroves are famously dense-growing. European pine and spruce forests can also be very dense (minimal branching). Whereas large broadleaves can be very extensive if grown to 100+ years.

Planting density depends on climate. For example, some rainforest reforestation projects have favored dense planting, to prevent slower growing trees from being outcompeted by fast-growing jungle plants. Conversely, Southern US forestry has some of the lowest planting density, falling by over 80% in the past 50-years, as survival has become better and better, averaging 88%.

Planting density depends on seedling cost. Another study notes that planting density can be optimized, year by year, depending on the costs of seedlings, favoring lower densities (and less subsequent thinning) to mute the impacts of seedling shortages.

Growing economies: reforest and reinvest?

clean economic development

CO2 removal credits could add 6-60% to the GDP of 47 emerging countries as they reforest 1.5bn acres and create a 7.5GTpa CO2 sink, while the resultant cash flows could double these countriesโ€™ investment rates. Reinvesting in wind, solar, electrification avoids higher carbon fuels and deforestation for firewood. Reinvesting in timber value chains maximizes CO2 permanence and value. This 13-page note explores ‘reforest and reinvest’ as a promising framework for clean economic development in the energy transition.

Wood products: typical pricing?

pricing of different wood products

This data-file aggregates the pricing of different wood products, as storing carbon in long-lived materials matters amidst the energy transition. It can also add economic value. While upgrading raw timber into high value materials can uplift realized pricing in reforestation projects by 20-60x, which improves the permanence of nature-based CO2 removal credits.


Stumpage prices reflect the prices of timber at the immediate point of harvesting in a forest. We think stumpage prices are typically around $40/m3, ranging from $20-80/m3, depending on the timber type and location.

Whole logs that have been de-limbed, transported out of the forest, partially dried, and possibly de-barked are around 3-4x more valuable reaching $100-200/m3.

Sawn beams are 3x more valuable again, with pricing recently exceeding $500/m3, after the additional steps of drying, grading, cutting into specific shapes, and other possible treating steps.

Paper comes next, and we think pricing around $875/m3 is often sufficient for a 10% IRR at a new paper mill, while recent paper pricing has run around the $1,000/m3 mark.

Board materials are another 2-4x more valuable gain. Plywood, which is formed by unravelling entire logs into long, continuous sheets, might price above $1,000/m3, while lower grade board materials below $1,000/m3 will be formed by binding and re-constituting wood-chip (Oriented strand board, OSB) or sawdust (medium-density fibreboard, MDF).

Engineered timber products included glulam and cross laminated timber. These can have more variable pricing, but in both cases, we think recent pricing has run above $1,500/m3. We think CLT is an interesting alternative to steel in construction.

Hardwood flooring can be among the most valuable timber products. Again, pricing is variable, but it can be 2x more expensive than engineered timber products.

We also see growing value in mass timber, related chemicalswood in wind, and interesting companies with specialized wood productsequipment or engineered wood technologies.

CO2 removals: CO2OL Panama project?

CO2OL reforestation project

The CO2OL Tropical Mix project has planted 9M trees on 13,000 hectares of degraded pasture land across 45 sites in Panama since 1995; 40% teak, 60% native species; to produce sustainable hardwoods, especially for furniture, on 25-year rotations; while 20-30% of the land is reserved for conservation and bio-diversity. The project achieved a relatively high score of 88/100 on our usual assessment framework. CO2 credits are priced at $38/ton. We contributed $1,900 to the project and offset 50 tons of CO2.


Panama’s CO2OL Project is a reforestation project, established in 1995, and managed by German environmental services company Forliance. As part of our ongoing work into nature based solutions to climate change, and to practice what we preach in offsetting our own CO2, we have scored the project on our usual 100-point scorecard (others linked here).

Real. The CO2OL project is certified by Gold Standard, independently audited, and has high quality imagery and offtakers. Our score was mildly marked down for a complex corporate structure, which has been modified over time.

Incremental. Reforesting former cattle pasture in a country losing 0.4% of its primary forests per year. Minor “leakage” has been considered and deducted from CO2 calculations.

Measurable. The calculation methodology is clear, realistic, includes “give-backs” and baselines, is backed up by satellite/GIS data, has been audited, and de-risked by diversification.

Permanence. After harvest, hardwoods will lock away carbon in furniture, which has good carbon credentials as a wood usage. Score is marked down due to country transparency, long-term future and complex corporate structure.

Bio-diverse. The project is 40% teak, 60% spread across 5-20 native species, creating animal habitat and up to 200 jobs, although we marked down mildly as rotations are short at 25-years.

Full details on the CO2OL Reforestation project, which led us to the scores above are given in the data-file. Despite recent media scandal-mongering into nature-based solutions, we were happy to find a project that scores relatively well on our framework.

Tree seedlings: costs and economics?

Costs of tree seedlings

The US plants over 1.3bn tree seedlings per year. Especially pine. These seedlings are typically 8-10 months old, with heights of 25-30mm, root collars of 5mm, and total mass of 5-10 grams, having been grown by dedicated producers. This data-file captures the costs of tree seedlings, to support afforestation, reforestation or broader forestry. Costs should average 13.5c/seedling. A detailed cost breakdown of tree seedlings is also given in $/m2 and $/seedling terms.


Growing practices vary. We think bare root seedlings, grown in open fields, in warmer climates, have a typical cost of 7c/seedling. Conversely, in colder climates, seedlings maybe grown in containers, in heated greenhouses, which also has the advantage of permitting earlier growing and better transportability, despite higher costs, closer to 15c/seedling. Typical costs of tree seedlings are disaggregated in $/m2 and $/seedlings for both categories.

At 13.5c/seedling prices, and a typical planting density of 500 seedlings per acre, seedling costs might explain $70/acre of the total up-front costs in a reforestation project, which might run around $300-400/acre (ex land acquisition).

Operational leverage is very high in the seedling industry. At least compared with other industries captured in our economic models. Net margins are likely in the low single digit percentages in normal times, as costs are spread across labor, packaging materials, soil materials, seeds, herbicides, fertilizers, irrigation, maintenance G&A; and in greenhouses, also heat and electricity. Labor is the large cost line, but the other lines add up too.

This does raise the question whether a rapidly growing market for reforestation could create meaningful upside. A 1c/kWh (c7%) increase in end pricing, e.g., due to under-supply, might flow through to double the net margins of a seedling producer.

ArborGen stood out as one of the leading pure-play companies producing seedlings, in the US and Brazil. It is listed. And it has historically sold 80% of “Mass Control Pollinated” seedlings, which are selected from the best cultivars, adapted to regional conditions, especially in the Southern US. High-quality seedlings should command a premium.

This data-file also includes our notes from technical papers, with some details into the differences between bare-root and greenhouse growing. Weyerhaeuser is another company up in the works, which is backwards integrated and produces tens of millions of seedlings per year.

Albedo of different landscapes: a challenge for reforestation?

Albedo of different landscapes

Forests are darker than their surroundings? So does their low albedo curb our enthusiasm for nature-based solutions? This data-file aggregates the average albedo of different landscapes, based on technical papers and internet sources. The albedo impact of reforestation seems numerically very small. There is even an intriguing link where forests can increase the formation of clouds, which have the highest average albedo of any reflective category.


Our latest roadmap to net zero assumes that c20% of all decarbonization of the 2050 energy system will come from new forests and other nature-based solutions to climate change. These solutions need to prove they are real, incremental, measurable, long-lasting and have other social and ecological benefits, to score highly on our nature-based project screens.

However, some commentators have criticized that forests are a not a good climate solution, because they are darker than the surrounding landscape. Forests have an average albedo of 13% versus open landscapes at 20%. Thus they absorb more heat. Thus could planting more forests increase warming in addition to the impacts of CO2?

Four observations on this argument are summarized below, and supported by the data in this data-file.

(1) Starting with a definition: Albedo is the percentage of incoming radiation that a surface reflects. This includes both visible light (from violet at 380nm to red at 700nm), which is c43% of the incoming energy reaching the Earth’s surface; and infra-red radiation, which is 49%; and ultra-violet light, which is the remaining 8%. Physics’s theoretical ‘black body‘ has an albedo near 0%, and the brightest whites have an albedo near 100%. So it is true that land uses with a lower albedo will absorb more heat. But how much more heat?

(2) Albedo impacts are a rounding error? Overall, our roadmap to net zero sees 3bn acres of the planet going from ‘not forest’ to ‘forest’ (note here), which is 8% of the world’s total land (40 bn acres in total) and 2.4% of the planet’s total surface (126bn acres in total) (database here). Moreover, where land use is moving from ‘not forest’ to ‘forest’ in our models, it is mainly moving from pastureland or grassland (average albedo of 20%) to forest. We are not reforesting ice sheets (average albedo of 60%) or the Sahara desert (average albedo of 35%). And there may even be some areas where albedo increases, if reforestation occurs on dark and bare soils (albedo of 10%) or when planting mangroves that fill in open bodies of water (albedo of 8%). However, overall, a 7pp heat absorption delta on 2.4% of the world’s land is a rounding error. And it may itself be counteracted…

(3) Clouds and the cooling impacts of transpiration. One of the most interesting links in technical literature is that forests may increase cloud cover, and in turn, clouds have an average albedo of 63%, which is the highest of any surface condition that is averaged in our data-file. The best paper to cross our screen evaluated 10-years of satellite data over large oak and pine forests in Europe, finding 5-15pp higher absolute cloud cover during summer months over the forests (compared with surrounding cropland), and consistently higher cloud cover downwind of the forest. The downwind effect was most pronounced when wind speeds were higher, and absolute cloud cover decreased by 20pp after a large portion of one of these forests was damaged by a storm in 2009 (see Teuling, et al, 2017). Another recent study also found increased cloud cover over forests, and calculated that this would amplify the net carbon benefits of reforestation, especially at mid-latitudes (see Cerasoli, et al, 2020).

(4) Conflicting conclusions. It should also be highlighted that other technical papers have found more mixed links between forest cover and cloud cover. From the technical papers that crossed our screens, there seems to be much less certainty and much more debate than in other areas of climate science that we have reviewed. Even the science of measuring albedo is quite full of controversy, varying over time, with the seasons, with weather conditions. Some commentators argue that in winter, Arctic-latitude forests are much darker than snow, but to this I say, come visit me in Estonia in December-January, and see how little daylight there is to absorb/reflect!! Our summaries of different technical papers are in the data-file, along with average values for the albedo of different landscapes.

Overall, we still think that protecting and restoring nature is climate-positive, as well as being a philosophical virtue in the eyes of any true environmentalist (almost as a defining criterion!). We will continue to screen nature based CO2 removals in our research in 2023, and hope to see a new array of exponentially improved projects coming to the market.

Nature based solutions: CO2 removals in 2022?

Market for nature based carbon offsets

Is the nascent market for nature-based carbon offsets working? We appraised five projects in 2022, and contributed $7,700 to capture 440 tons of CO2, which is 20x our own CO2 footprint. This 11-page note presents our top five conclusions. Todayโ€™s market lacks depth and efficiency. High-quality credits are most bottlenecked. Prices rise further in 2023. A new wave of projects is emerging?

CO2 offsets: Pachama’s AI platform?

Pachama CO2 offset review

Pachama is a nature-based technology company, which has raised $79M, to create a portal where buyers can choose “from rigorously vetted forest restoration and conservation projects”, which in turn are tracked using proprietary AI. This data-file is a Pachama CO2 offset review. We have assessed the portfolio, some challenges and our own experiences, via our usual framework for assessing nature-based CO2 removals.


As of November-2022, the majority of projects available on Pachama’s portal are avoided emissions projects. These are excellent conservation projects, accredited by VERRA, protecting vulnerable eco-systems, and achieving some of the highest biodiversity scores of any projects that have crossed our screens.

However, it remains debatable whether these projects can be considered to be “offsetting CO2”. CO2 credits are not being awarded for pulling additional CO2 out of the sky and storing it in a natural eco-system, as per other CO2 removal projects that we have assessed.

Rather, CO2 offsets are being issued relative to a hypothetical scenario where a protected forest is deforested at a rate of 1-2% per year (varies by project) over the next 20-70 years (chart below).

We think that over time, Pachama would like to seed new forests, and more incremental projects on its platform, but for now there is limited depth in the nature-based CO2 market, and most of the certified CO2 offset projects are REDD (conservation) projects.

In one of the largest CO2 offset projects in the Pachama portfolio today, CO2 offsets are issued relative to a scenario whether a carbon-dense, 26-000 year old peatland is drained and thus caused to release c500MT of CO2. Blue carbon eco-systems can store a lot of carbon, over 1,000 tons/hectare, possibly over 2,000 tons/hectare. But 500MT is a truly enormous number. It is equivalent to the direct annual emissions of the entire global fertilizer industry per our CO2 breakdown. This raises some question marks.

It gets a bit philosophical, but in our view, carbon “offsetting” should be about cancelling out the net impacts of emitting +X tons of unavoidable CO2 into the atmosphere by pulling out -X tons of CO2 from the atmosphere and sequestering it over the long-term. Not by avoiding a further +X tons of emissions. (Morally, you cannot atone for a murder by enumerating the list of people you have not murdered !!).

We want to support conservation of nature, and high-quality organizations in nature-based solutions; and so we allocated $700 to offset 40 tons of CO2 from the Pachama portfolio at the current price of $17.6/ton. However, our overall experience was somewhat disappointing: per the Pachama website, we thought we were buying from “Pachama’s global portfolio of high-quality forest projects” (screenshot above). But after making the purchase, all of our purchase ended up allocated to the single, large peat conservation project, described above.

Further details on our Pachama CO2 offset review are in the data-file. We have also appraised other CO2 removal projects using the same framework.

Copyright: Thunder Said Energy, 2019-2024.