Forest carbon prospecting for climate mitigation

Carbon nance projects that protect tropical forests could support both nature conservation and climate mitigation goals. Global demand for nature-based carbon credits is outpacing their supply, due partly to gaps in knowledge needed to inform and prioritize investment decisions. Here, we show that at current carbon market prices the protection of tropical forests can generate investible carbon amounting to 2.4 (±1.1) GtCO2e yr-1 globally. We further show that nancially viable carbon projects could generate return-on-investment amounting to $77.5b y-1 in net present value (Asia-Pacic: $38.4b y-1; Americas: $33.4b y-1; Africa: $5.7b y-1). However, we also nd that ~75% (1.24 billion ha) of forest carbon sites would be nancially unviable for failing to break even over the project lifetime. From a conservation perspective, unless carbon prices increase in the future, it is imperative to implement other conservation interventions, in addition to carbon nance, to safeguard carbon stocks and biodiversity in vulnerable forests. that

condition for certifying all carbon credits, including nature-based credits traded in the voluntary carbon market, under the rules of the United Nations Framework Convention on Climate Change 8 .
We modeled the magnitude of certi able carbon from forest carbon projects and its climate mitigation potential to produce a global investible forest carbon map (Fig. 1).
This analysis presents spatially explicit information on the relative climate mitigation potential of protecting tropical forests, accounting for not only the carbon stock of a forest, but also the risk of losing that forest 8,10 ( Fig. 1; see Methods).
Our estimates of investible carbon are 33-43% lower than those reported in previous studies 2, 3 , which were largely based on aggregated country level data on carbon stocks and deforestation rates. Furthermore, our analysis also incorporates several criteria of the Voluntary Carbon Standard (VCS), such as the requirement to set aside buffer credits 8 , not considered in other studies 2,3 . As such, we are able to compare our estimates of investible carbon with empirical data on veri ed carbon units reported from 28 real world VCS projects (verra.org/), and report a relatively strong correlation between the two datasets (R=0.53, p<0.05, Root Mean Square Error=0.53; Fig. S1; see Methods). Table 1. Global, regional, and country level estimates of investible carbon and return-on-investment (based on net present value).  Table 1).
Perhaps unsurprisingly, much of this investible carbon potential remains unrealized. For example, the total volume of veri ed carbon units produced in these top ve countries from existing VCS projects represents <0.1% (0.98 MtCO 2 e yr -1 ) of the countries' total investible carbon potential estimated from our analysis (1.2 GtCO 2 e yr -1 ) (Tables 1, S1) 11,12 .
Barriers to the establishment of forest carbon projects may include competing interests and priorities from other economic sectors (e.g. agriculture), lack of enabling conditions and policies, governance and institutional constraints, and prohibitively high technical entry bar 11-13 . Many of these barriers may be overcome if actionable information regarding both the nancial risks and return-on-investment of projects is available to incentivize solutions.
Indeed, investible carbon projects may not all be pro table. The nancial viability of a project depends on a range of factors, including operational costs and carbon pricing, as well as political risk, which may vary with location and over time 14 .
We modeled the relative pro tability of these projects to produce a global forest carbon return-oninvestment map (Fig. 2).
We based our analysis on several simplifying assumptions (see Methods for details). Brie y, we applied a cost estimate of $25 ha -1 for project establishment, and $10 ha -1 y -1 for subsequent years for project maintenance. We also assumed a constant carbon price of $5.8 t -1 CO2e for the rst ve years, followed by a 5% price appreciation for the subsequent years over a project timeframe of 30 years 14 . Finally, we applied a risk-adjusted discount rate of 10% in our calculation of net present values (NPV) for the returnon-investment of tropical forest carbon projects.
We nd that the vast majority of nancially viable (i.e. yielding positive NPV) and most pro Globally, ~75% (1.24 billion ha) of the investible forest carbon sites would be nancially unviable for carbon nance for failing to break even over the project lifetime (i.e. yielding negative NPV; Fig. 2).
Importantly, these forests represent forgone climate mitigation at a rate of 0.7 GtCO 2 e yr -1 . From a forest conservation perspective, these ndings suggest that carbon nance will fail to protect the vast majority of investible carbon sites, which are also, by de nition, vulnerable to deforestation (Figs. 2 & S2; Table 1).
However, if global demands for nature-based carbon credits continue to grow 6 , future carbon prices may also increase. We modeled the effects of carbon pricing on the nancial viability of forest carbon sites globally. We nd that carbon pricing at $15 t -1 CO 2 e and $50 t -1 CO 2 e are needed to protect 50% and 80% of investible carbon sites, respectively (Fig. 3). Further carbon price increases above $25 t -1 CO 2 e would only bring marginal forest conservation and climate mitigation bene ts (Fig. 3).
Nevertheless, it is important to note that some nancially viable but less pro table forest carbon sites will struggle to compete with lucrative land uses, particularly in countries such as Brazil and Indonesia, which are the world's major producers of soy and palm oil, respectively 14 . In other countries, such as the Democratic Republic of the Congo, hydrocarbon exploration and logging developments with multiple vested interests may pose additional barriers to carbon projects 16 . Therefore, it is imperative to implement other conservation strategies and interventions, in addition to carbon nance, to safeguard the carbon stocks and biodiversity in these vulnerable forests. Fig. 3. Effect of carbon pricing on the nancial viability of forest carbon sites (i.e. the proportion of investible forest carbon that are nancially viable for carbon nance). Shadings around the lines represent con dence bands based on standard deviation.
Obviously, there is a wide range of environmental, socioeconomic, governance and geopolitical factors that can in uence climate strategies, conservation actions and investment decisions 17 . For example, some carbon projects may include nancially unviable sites that are important for biodiversity conservation, maintaining rural livelihoods or provide other co-bene ts of forest protection that may be highly valued by society but not internalized in our analysis 4 .
Furthermore, the political ecology landscape of existing and new carbon investments within a host country may also in uence the long-term success of forest carbon projects, and ultimately the permanence of carbon credits 9 . For example, the political risk for certi ed carbon credits has recently increased signi cantly in Brazil. In exchange for political support, the Brazilian government laid the foundation for landowners to accelerate deforestation 18 . This political bargaining may have seriously compromised Brazil's ability to meet the Paris target. These political risk considerations are crucial to ensure the long-term viability of carbon investments.
Our analyses draw from a sliver of the best available data to provide a snapshot of the relative investible carbon and return-on-investment for the protection of tropical forests as a natural climate solution. By clarifying some of the opportunities and constraints of tropical forest carbon projects, we help to calibrate expectations, incentivize actions, and expedite public and private sector engagements and capital investments in natural climate solutions to bene t the environment, climate and society.  Hijmans, R. J. & Van Etten, J. raster: Geographic data analysis and modeling. R package version 2.5-8. Vienna, Austria: The R Foundation. Retrieved from https://cran/. R-project. org/package= raster (2016).

Overview of methods
In this study, we modeled the magnitude of certi able carbon from forest carbon projects and its climate mitigation potential to produce a global investible forest carbon map. We then compared this estimated volume to that of veri ed carbon units reported from existing veri ed projects in the Verra database (https://verra.org/). Lastly, we modeled the relative pro tability of these forest carbon sites to produce a global forest carbon return-on-investment map based on their net present value (NPV). All calculations were based on data dated between 2012 and 2017 and at a resolution of 0.00833 degrees (~1 km). To ensure data standardization, we resampled (bilinear) ner-scaled data where necessary, for example, for data sourced from ESA-CCI 19 .
Estimating investible carbon To model and produce a spatially explicit map of investible carbon, we performed two key analyses. The rst was to estimate relative above-and below-ground biomass carbon across tropical forest areas: To achieve this, we adapted methods from Saatchi et al. 20 and applied them to more recent data (2012)(2013)(2014)(2015)(2016). In particular, we based our estimates of aboveground biomass on maps from Avitabile et al. 21 applying a stoichiometric factor of 0.47, which was based on the average value across several reference studies-(e.g. 3,20,22 ).
These maps were constrained to include only tropical forests between ~23.44°N and 23.44°S, and excluding all land cover types that would preclude forests, for example, bare ground, water, agriculture and urban areas 19 .
We then used the aboveground biomass maps from Avitabile et al. 21 to estimate the belowground biomass following the methods of Saatchi et al. 20 , which was derived from the root:shoot biomass ratio equation in Mokany et al. 23 : Belowground biomass = 0.489•Aboveground biomass 0.89 We applied the same stoichiometric factor to estimate belowground biomass carbon, and added the organic carbon density of the topsoil layer (0-30cm) obtained from the European Soil Data Centre 24 . This produced an overall belowground biomass carbon estimate.
Following this, we used a conversion factor of 3.67 3 to estimate the volume of CO 2 e associated with both above-and below-ground biomass carbon.
To the above-and below-ground biomass carbon estimates, we applied the key criteria for certifying carbon credits and Voluntary Emissions Reduction units (VERs) under the rules of the UNFCCC, Kyoto Protocol, and the various voluntary certi cation standards, such as the Veri ed Carbon Standard (VCS) 8,9 . Importantly, our analyses were guided by the requirements stipulated by VCS-the most widely used voluntary greenhouse gas program globally 8 : A key component of the requirements is "additionality" or the amount of forest carbon stocks that would be lost to deforestation without the protection of the proposed project. To estimate additionality, we utilized the average predicted deforestation rates (2029) generated in Hewson et al. 10 annualized over the prediction period (15 years). This produced an estimated annual deforestation rate, which we then multiplied by the above-and below-ground biomass carbon layers to approximate additionality.
While belowground carbon pool estimates are an optional consideration in VCS, we included them in our study, and calculated a conservative 10-year decay estimate 8 .
Additionally, we further excluded areas that would not qualify as certi able within these forest areas 8 . These included recently deforested areas (2010-2017) 25 as well as human settlement areas within these forest 26 .
Lastly, we also accounted for VCS requirement to set aside buffer credits of 20% net change carbon stocks in each area 8 .
Comparing our investible carbon estimates to veri ed carbon units We compared our estimates of investible carbon to the volume of veri ed carbon units reported by existing veri ed avoided deforestation projects. We utilized the Verra Project Database (https://verra.org/), extracting all avoided deforestation projects that: 1) possessed project area data; 2) is entirely within the tropics; 3) has been veri ed. Veri ed projects here included both the statuses "Veri ed, under veri cation" and "Veri cation approve" (Table S1).
To compare these data, we rst extracted the shape les of 28 projects that met these criteria (see Table S1). These shape les were then used to extract the corresponding total volume of estimated investible carbon credits via masking.
We then compared these values to the volume of veri ed carbon units (VCUs) issued across the years (2005-2018) for each project. The number of yearly data points for each project ranged from 1-13, and generated a total of 134 points of comparison. We then assessed the degree of correlation, with Pearson's correlation, and relative accuracy, via Root Mean Square Error (RMSE), of these corresponding data.

Estimating return-on-investment
Based on our map of investible carbon, we modelled the relative pro tability of these forest carbon sites to produce a global forest carbon return-on-investment map. We calculated the NPV of these returns based on several simplifying assumptions following established values from previous studies 14 .
First, we estimated the cost of project establishment at $25 ha -1 . This was based on the a wide range of costs that are key to the development of a project-including but not limited to project design, governance and planning, enforcement, zonation, land tenure and acquisition, surveying and research 14,27,28 .
Second, we estimated an annual maintenance cost of $10 ha -1 , which included aspects such as education and communication, monitoring, sustainable livelihoods, marketing, nance and administration 14,27,28 .
Third, we assumed a constant carbon price of $5.8 t -1 CO 2 e for the rst ve years. This price was based on an average price of carbon for avoided deforestation projects recorded by Forest Trends' Ecosystem Marketplace reports between 2006-2018 6 . After the rst ve years, we calculated a 5% price appreciation for the subsequent years over a project timeframe of 30 years 14 .
Based on these criteria, we calculated the NPV as well as the accumulated pro ts over the next 30 years, based on a 10% risk-adjusted discount rate.
We secondarily calculated NPV based a range of carbon prices, with a maximum of $100 t -1 CO2e matching the cost-effective thresholds from Griscom et al. 2 . Speci cally, we considered the carbon price intervals-$1, $5, $10, $15, $25, $50, $100 t -1 CO 2 e-while maintaining the project establishment and annual maintenance cost, price appreciation, discount rates and timeframe. Based on these criteria and excluding site that unable to breakeven, we then calculated the potential pro table forest areas associated with these carbon prices as a percentage of the total investible forest carbon sites.
These values of investible carbon and return-on-investment (based on NPV) were summarized to global, regional, and country level estimates (see Table 1). While some countries extend beyond tropical latitudes, we only analyze and present data based on their tropical areas. These values were rounded to the nearest 1000 values.
Accounting for uncertainty To incorporate uncertainty across our estimations, we utilized the uncertainties inherent to the source datasets (reported as standard deviations) 21 . We also estimated the uncertainty associated with the price Thoumi, G. Emeralds on the equator: An avoided deforestation carbon markets strategy manual. (2008).  Global forest carbon return-on-investment from nancially viable sites, presented as net present values over a 30-year timeframe. We also present the accumulation of pro ts over time at the global and regional levels (inset).