National Academies Press: OpenBook

Negative Emissions Technologies and Reliable Sequestration: A Research Agenda (2018)

Chapter: 3 Terrestrial Carbon Removal and Sequestration

« Previous: 2 Coastal Blue Carbon
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 61
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 62
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 63
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 64
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 65
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 66
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 67
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 68
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 69
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 70
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 71
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 72
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 73
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 74
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 75
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 76
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 77
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 78
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 79
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 80
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 81
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 82
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 83
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 84
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 85
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 86
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 87
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 88
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 89
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 90
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 91
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 92
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 93
Suggested Citation:"3 Terrestrial Carbon Removal and Sequestration." National Academies of Sciences, Engineering, and Medicine. 2018. Negative Emissions Technologies and Reliable Sequestration: A Research Agenda. Washington, DC: The National Academies Press. doi: 10.17226/25259.
×
Page 94

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

3 Terrestrial Carbon Removal and Sequestration INTRODUCTION Definition of Terrestrial Carbon Sequestration Terrestrial carbon sequestration (CS) is defined here as the increase in the amount and maintenance over time of organic carbon in biological stocks, driven by plant assimilation of CO2 from the atmosphere. Biological carbon stocks are largely controlled by actively cycling processes, i.e., assimilation (carbon uptake from the atmosphere) and respiration (carbon loss to the atmosphere). The carbon stocks of interest are those that can accumulate and persist over multi-decal timescales, namely woody biomass and coarse woody debris and soil organic matter. More ephemeral carbon stocks, including herbaceous biomass and plant litter with short residence times (<1 yr), are generally ignored in the context of carbon sequestration because these do not represent a persistent removal of CO2 from the atmosphere. Harvested woody biomass used for long-lived wood products may also accumulate and persist, but is also subject to release of stored carbon as product use is discontinued unless sequestered in a landfill. While the committee recognizes that technologies to increase carbon uptake by forestry and agriculture may also result in increased emissions due to decomposition and disturbance of recalcitrant carbon in soils, virtually all the data reported are for net carbon uptake, determined from stock changes over time. Hence decomposition and disturbance impacts on carbon losses are included in the net values. As a general principle, the overall balance of these carbon stocks is driven by the difference between carbon inputs (via plant CO2 assimilation) and carbon losses to the atmosphere (via decomposition/microbial respiration as well as from fire/combustion). Thus, increasing the standing stocks of carbon can be done by increasing the rate of input of carbon, decreasing outputs of carbon or both. For woody biomass, this implies growing more trees with more biomass per unit area, and maintaining that biomass over a longer time span, and/or decreasing loss of woody biomass through tree mortality, fire, and harvesting operations. For soil organic carbon, this means increasing the rate of input of plant-derived detritus to the soil and/or reducing the rate at which the organic compounds added or already in the soil are decomposed and mineralized to CO2. Achieving and maintaining increases in biological carbon storage requires active management practices that manipulate system carbon balances, for which a variety of existing as well as potential future options exist. Rationale for Terrestrial Carbon Sequestration Global terrestrial biotic carbon stocks include ca. 600 Gt carbon in biomass and ca. 1500 Gt carbon in soil to a depth of 1 m (ca. 2600 Gt C to 2m); annual fluxes from and to the atmosphere and terrestrial ecosystems, on the order of 60 Gt C yr-1, are roughly balanced but with an average net residual C uptake by terrestrial ecosystems of 1-2 Gt C (Le Quéré et al., 2016). However, historically the conversion of native ecosystems to managed land, particularly cropland, has been a large net source of CO2 flux to the atmosphere with a significant depletion of biomass and soil carbon stocks, particularly over the past ca. 200-300 years as human population and land transformation exploded. Recent estimates of the anthropogenic-induced losses of terrestrial carbon stocks are on the order of 145 Gt C from woody biomass and soils from 1850 to 2015 (Houghton and Nassikas, 2017), 133 Gt C from soil carbon stocks over the last 12,000 years (Sanderman et al., 2017), and 379 Gt C of biomass over the past 10,000 years (Pan et al., 2013). These estimates of historical losses indicate a hypothetical though highly impractical upper bound for restoring terrestrial carbon stocks through adjustments to land management. PREPUBLICATION COPY 61

62 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda The rationale for pursuing terrestrial carbon sequestration as a CO2 removal strategy is at least three-fold. First, the ‘technology’ for carbon capture and storage already exists, via the uptake of CO2 by plants and its storage in longer-lived biotic carbon pools, providing opportunity for storage that is energetically and economically competitive with other carbon removal and mitigation options. Secondly, there is a significant storage capacity based on the magnitude of historical carbon stocks compared with much lower contemporary stocks. To a large extent, terrestrial carbon sequestration approaches can be seen as a reversal of previous ecosystem degradation—changing land use and management to favor greater biomass and soil carbon stocks. Finally, increasing biomass and soil carbon stocks may confer additional ecosystem service benefits, including watershed protection, increased biodiversity and improved soil health and fertility. In contrast, ecosystem service benefits, including biodiversity, may be reduced if the carbon sink creating activities change disturbance regimes or other landscape characteristics that affect species. These are discussed in more detail below. APPROACHES FOR TERRESTRIAL SEQUESTRATION The Committee’s assessment of terrestrial carbon sequestration options centers around the predominant types of managed ecosystems: forestland, cropland and grassland, for each of which multiple management ‘levers’ exist to modify carbon stocks. Other intensively managed land such as urban/peri- urban landscapes contain biotic carbon stocks in trees and greenspaces (parks, lawns, etc.), for which some of the sequestration practices described under the forestland and grassland category could apply. Wetlands are of particular importance in that they contain the highest carbon stocks per unit area of any ecosystem. Where they have been drained and converted to crop, grazing or forest production, approaches to recover carbon sequestration capacity will be discussed under each of the three main land use types. Coastal ecosystems (“coastal blue carbon”) which are on the edge of terrestrial and ocean ecosystems are covered in Chapter 2. Bioenergy with carbon capture and storage (BECCS) is likewise covered in Chapter 4, with references to this chapter regarding land requirements for all approaches. In describing technical options, the Committee distinguishes between ‘conventional’ technologies or practices to increase carbon stocks via CO2 removals, versus what the Committee terms ‘frontier’ technologies. Conventional practices refer to management practices that to a limited degree are already in use and have an existing body of applied research knowledge. Increasing carbon stocks via these practices mainly requires a much greater degree of participation by land managers (land area applied to) and refinement in their applicability and support services to facilitate broader adoption. In contrast, frontier technologies are approaches that are still in a basic research phase and not yet tested for widespread deployment (e.g., perennial grain crops) or involve practices (e.g. ‘enhanced wood preservation’) that are outside of the frame of current land management objectives or practices. Conventional Forest Practices Forestry practices that can act to remove CO2 and reduce emissions fall into three categories of action: 1) avoiding conversion of forest land to other land uses (deforestation) that store less carbon and have lower rates of removal, 2) converting non-forest land to forest (afforestation/reforestation) thereby increasing carbon removal and stocks, and 3) modifying forest management practices to either increase carbon stocks, or increase net removals from the atmosphere (balance of gains and losses of CO2), or both. Modifying forest management involves an array of practices, such as accelerated regeneration of non-stocked forest after disturbance which increases carbon removal in the near term, restoring degraded forests to healthier and sustainable conditions that maintain removal capacity, and extending the rotation length (age of forest at harvest) which maintains removal capacity, avoids emissions associated with wood harvest, and directs more biomass into long-lived wood products that store harvested carbon. Altering management and wood product design to foster preservation of more long-lived wood products PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 63 that store the harvested biomass carbon is a fourth category that is treated later in this chapter as a frontier technology. Assessing the carbon consequences of forestry activities is complicated by the many different direct and indirect effects on the carbon cycle. Forest ecosystems are composed of three main carbon pools that respond differently to harvesting, management, and other disturbances: live biomass (above- and below-ground); standing and down dead wood; and soil organic matter including surface litter, humus, and mineral soil layers. These carbon pools respond to disturbances over periods of years to decades such that accounting for impacts is an ongoing process. Besides ecosystem impacts, harvested wood products have multiple effects on the carbon cycle including their function as temporary storage of removed carbon while in use or disposal, substitution of wood for other construction materials that require substantial quantities of fossil energy to produce (avoided emissions), and use of wood for biofuel which may reduce net emissions relative to burning fossil fuels (discussed in BECCS chapter 4). Although avoiding deforestation typically has the most significant and immediate impact on a per-hectare basis by reducing emissions and sink capacity, it is not explored in depth in this report since it is mainly an emissions reduction or avoidance activity rather than an activity that increases CO2 removal. Avoiding deforestation is a potentially significant activity for the U.S. and globally, with avoided emissions ranging from 56 to 116 Mg C/ha/yr in the U.S. and 96-103 Mg C/ha/yr globally (EPA, 2005; Griscom et al., 2017). Afforestation/Reforestation Afforestation/reforestation has been extensively studied and implemented in the U.S. and around the world, so may be considered immediately deployable for carbon removal purposes. Afforestation/reforestation involves planting trees or facilitating natural regeneration of trees on land that has been in a nonforest use condition for some length of time. The Committee considers regeneration of a harvested forest as a forest management practice even though tree cover may have been temporarily reduced to a level typical of a non-forest use. Foresters have planted new forests or reforested non-forest lands for many decades and have good knowledge of which species are likely to be successfully established, and which areas are suitable for planting or assisted natural regeneration (Sample, 2017). In regions where tree planting is associated with timber production, such as the U.S. Southeast and Pacific Northwest, provenance trials, tree breeding programs, and genetic improvements have progressed to identify trees that have good survival rates, grow fast, and have properties that are most suitable for wood products (Burley, 1980). However, to date, carbon removal rates have not been the target of research about growing stock improvements; instead, improvements in planting stock have been targeted to increase timber production (above-ground biomass) while the effects on below-ground biomass and soil C remain largely unknown (Noormets et al., 2015). As a result, there is a need to consider selections of trees that would increase whole-tree biomass as a means to increase their function as carbon removal organisms (e.g., Muchero et al., 2013). Data on tree growth and carbon removal rates are readily available, as are guidelines for successful planting and natural regeneration. Table 3.1 illustrates some of the more common and representative carbon removal rates that span a wide range of values because of heterogeneous climates, sites, and forest types. Rates of ecosystem carbon increase with afforestation/reforestation in the U.S. range from 0.7 to 6.4 Mg/ha/yr for a period of 50-100 years or more (Table 3.1). The carbon removal rate is variable over this time period, generally following a growth curve having a diminishing rate of increase as the forest reaches maturity (Figure 3.1). However, if the activity is initiated over a period of years on different parts of the landscape, which would be likely given infrastructure and funding constraints, the idealized pattern of incremental carbon storage may be more linear except for the early years (Figure 3.2). Estimates of the potential carbon sink from afforestation/reforestation include changes in biomass and soil carbon without consideration of impacts of future harvesting or natural disturbances. Disturbances would release stored carbon from the ecosystem. In the case of harvesting, some of the carbon lost from the ecosystem would PREPUBLICATION COPY

64 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda TABLE 3.1. Published estimates of biomass carbon sequestration rates for various forestry activities. Global Estimates Activity Net stock increase Reference (Mg C ha-1yr-1) Afforestation/Reforestation 2.8-5.5 Griscom et al., 2017 Afforestation/Reforestation 3.4 Smith et al. 2016 Improved forest 0.2-1.2 Griscom et al., 2017 management U.S. Estimates Activity Net stock increase Reference (Mg C ha-1yr-1) Afforestation/Reforestation 0.7-6.4 Birdsey 1996 Improved forest 1.4-2.5 Denef et al., 2011 management FIGURE 3.1. Changes in carbon stock from afforestation. Example is for loblolly-shortleaf pine in the Southeast U.S. Data from Smith et al., 2006. be retained in wood products or landfills. Increases in harvested wood products may also reduce emissions through substitution of wood for other types of construction materials that require more energy to produce. Emissions reductions from increasing harvested wood in place of other materials are not estimated here even though potentially significant, since these activities do not represent increases in carbon removal. Improved Forest Management Improved forest management has been extensively studied in developed countries with respect to impacts on carbon removal, feasibility, and cost; therefore, this technology may be considered as immediately deployable. Forest management involves a very wide array of practices that have been tailored for regional and local conditions based on experience and extensive silvicultural experiments, PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 65 FIGURE 3.2. The impacts of management actions on carbon storage of a hypothetical Douglas fir forest at different spatial scales (adapted from McKinley et al., 2011. many of which have been ongoing for decades (Malmsheimer et al., 2011; McKinley et al., 2011). In general, promising forest management practices to increase carbon removal include: • Accelerating regeneration in areas that have had major disturbances • Restoring forests that have been converted to “unsustainable” forest conditions—this includes both increasing carbon stocks by returning a forest to its original vegetation type that is better adapted to site and climate, and reducing carbon stocks of overstocked stands to levels that are less likely to have intense wildfires that convert the forest to the much lower carbon density of non-forest • Extending harvest rotations to grown larger trees and sustain carbon removal rates (not to mention avoid emissions associated with harvest) • Maintaining healthy forests by treating areas affected by insects and diseases or preventing conditions that foster outbreaks • Thinning and other silvicultural treatments that promote overall higher stand growth compared with untreated conditions Changing forest management practices such as extending timber harvest rotations and improving stocking and productivity could store an additional 0.2 to 2.5 Mg C/ha/yr for several decades based on U.S. and global estimates (Table 3.1). These estimates include changes in biomass and soil carbon but not changes in the stock of harvested wood products. Emissions reductions from increasing the use of harvested wood in place of other materials are not estimated here even though potentially significant, since these activities do not represent increases in carbon removal. The range of potential benefits from improved forest management is very wide because of the need to account for different land use and management histories as well as the many different climates, site conditions, and forest types. Generally, the per-hectare values are significantly less than observed for afforestation/reforestation because the growing stock of trees in an existing forest in the absence of management will still take up significant quantities of CO2; therefore, the additional uptake of CO2 is less than if comparing forest with no forest (McKinley et al., 2011). PREPUBLICATION COPY

66 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda Frontier Forest Practices Increase Harvested Wood Use and Preservation According to many studies, the main climate benefit of increasing the use of harvested wood products is emissions reductions from the substitution of wood products for materials such as concrete and steel that require more fossil fuels for production (e.g., Hashimoto et al., 2002; Perez-Garcia et al., 2005; Sathre and O'Connor, 2010). However, increasing the preservation of harvested wood by improving preservative treatment methods (Song et al., 2018) or advanced landfilling could be a significant CO2 removal approach with additional economic benefits. Zeng, 2008 proposed harvesting live trees and other biomass from managed forests and burying the logs in trenches or otherwise storing them to prevent the carbon from being released. Improving the preservation of wood products from existing harvest operations, and potentially increasing harvest with high levels of product preservation, could be viable approaches to increasing carbon removal. Wood removals from the world’s 4 billion ha of forests average around 3 billion m3 yr-1 or 0.65% of the growing stock, of which about half is for timber products and half for fuel (FAO, 2015a). During harvest, a significant portion of the live biomass is left in the forest as logging slash (not including tree roots), ranging from roughly 30 to 40% of the pre-harvest biomass at the national scale, with much higher variability at more local scales (Oswalt et al., 2014; Winjum et al., 1998). Some of this logging slash ends up being used for fuelwood and charcoal. Much of industrial wood removals comes from approximately half of the world’s forests designated for timber production or multiple use (FAO, 2015a). Much of the biomass removed from forests for timber products is emitted during primary processing into products, with losses ranging from roughly 20 to 60% depending on conversion efficiency (Bergman and Bowe, 2008; Ingerson, 2009; Kline, 2005; Liski et al., 2001). After the end of their useful life, wood products are typically deposited in landfills that are often designed for relatively rapid decomposition, or subject to other fates such as deposit in dumps that emit their stored carbon (Skog, 2008). According to several estimates, about 0.5 to 0.7 Gt CO2 yr-1 of the harvested carbon is sequestered in use or landfills after accounting for the inputs from current harvests and inherited losses from past harvests (Hashimoto et al., 2002; Miner and Perez-Garcia, 2007; Pan et al., 2011; Winjum et al., 1998). Conventional Cropland and Grassland Practices The vast majority of agricultural lands in the US (and globally) are not managed optimally for increased soil carbon storage. Most annual croplands in temperate climates have bare-fallow conditions outside of the main crop growing season and intensive tillage practices are still widespread. Some annual cropland is on marginal lands that are subject to continuing soil degradation. Many pastures and rangelands employ unimproved grazing systems and suboptimal forage management. However, there are numerous conservation management practices available that can increase carbon stocks in soils and that are successfully practiced by progressive farmers and ranchers. In many cases these practices have been well studied, with many long-term field experiments and comparative field observations. Table 3.2 lists several categories of management practices, classified according to their main mode of action in either increasing carbon inputs to soils and/or reducing carbon losses from soils. Following implementation of improved practices, soil carbon accrual rates can continue over several decades but attenuating over time as soil carbon contents tend towards a new equilibrium state with no further carbon gains unless additional carbon-accruing management practices are adopted. Furthermore, due to the dynamics of mineral-organic matter interactions that largely control the residence time of carbon in soil (Lehmann and Kleber, 2015), soils with initially low carbon concentrations have a greater propensity to gain carbon compared to soils with already high carbon concentrations. Thus there is PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 67 TABLE 3.2. Examples of conventional agricultural management actions that can increase organic carbon storage and promote a net removal of CO2 from the atmosphere (adapted from Paustian, 2014). Management Practice Increased carbon Reduced carbon inputs losses Increased productivity and residue retention X Cover crops X No-tillage and other conservation tillage X X Manure and compost addition X Conversion to perennial grasses and legumes X X Agroforestry X X Rewetting organic (i.e., peat and muck) soils X Improved grazing land management X X effectively a ‘saturation limitation’, which varies as a function of soil texture and mineralogy (Stewart et al., 2007), which hinders further carbon accrual for mineral soils with very high organic matter contents. Improved Annual Cropping Systems On annual croplands, farmers may adopt a number of cropping choices that increase inputs of carbon into soils: replacing winter bare-fallow with seasonal cover crops, planting crops that produce large amounts of residues, promoting more continuous cropping (reduced summer fallow frequency) in semi-arid environments, and increasing the proportion of perennial grass/legume forage crops within crop rotations. Such cropping choices can maximize the time during which live vegetated cover is maintained on the soil and increase the amount of root-derived carbon added to the soil (Rasse et al., 2005). In the past few years, interest has grown in the use of cover crops and they have been strongly promoted by USDAs Natural Resource Conservation Service. Cover crop adoption is rising; however, adoption rates in the US are still low (<5% of cropland area; USDA, 2014). This reflects unfamiliarity for growers, barriers due to additional costs, restrictions rated to crop insurance and possibly immature technology in some areas. Systems to increase cropping frequency and reduce summer-fallow in semi-arid cropland have been successful in increasing productivity as well as soil carbon stocks (Peterson et al., 1998) although large areas dominated by alternate year wheat-summer fallow practices remain. Wider adoption of crop rotations that incorporate two to three years of grass or legume hay with annual crops (common in mid 20th century Corn Belt agriculture) is limited by the higher prices for main commodity crops (e.g., corn, soybean) which encourages continuous grain mono-cropping. Hence, for a variety of reasons, best cropping practices for sequestering carbon are currently not widely used on US annual cropland, which means ample room to increase adoption rates if soil carbon sequestration becomes a more prominent policy goal. Tillage is used by farmers to manage crop residues and prepare a seed bed for crops, and is the main source of soil disturbance in croplands. Intensive tillage tends to accelerate decomposition rates of soil organic matter (Paustian et al., 2000). Advances in tillage implement technology and agronomic practice have allowed farmers in recent decades to reduce tillage frequency and intensity, sometimes ceasing tillage altogether with a practice known as ‘no-till’ (NT). Reduced tillage systems, particularly no-till, can increase the mean residence time and slow decay of soil organic matter (SOM; Six and Paustian, 2014), promoting greater soil carbon storage (Table 3.3). Many field studies show increases in soil organic carbon following adoption of reduced till and NT, with variations due to soil texture and climate. However, there are instances in which no-tillage does not increase soil carbon relative to conventional tillage—particularly in wet, cool climates where productivity under NT may be reduced (Ogle et al., 2012) and in soils with already high topsoil carbon contents, where organic matter stabilization efficiency may be lower than for residues mixed deeper into the soil with tillage (Angers and Eriksen-Hamel, 2008; Ogle et al., 2012). No-till and reduced tillage are not widely adopted on most annual cropland in the US, although with significant regional variation. PREPUBLICATION COPY

68 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda In summary, a combination of an extended vegetated period, increased residue (particularly root- derived) inputs and minimal tillage systems could be broadly implemented as best practices for increasing soil carbon stocks on annual cropland. Rates vary significantly as a function of climate, soil and land use history, but rates of 0.2-0.5 t C/ha/yr over time periods of 20-40 years are representative, as observed in numerous meta-analyses of long-term field experiment data both globally and in the U.S. (Table 3.3). Organic Matter Amendment Organic matter additions such as animal manures and composts can increase soil carbon contents, both by virtue of the added carbon in the amendment itself and through improving soil physical attributes and nutrient availability. Cropland soils receiving substantial organic amendments invariably show increases in soil carbon concentrations derived from the amendment itself. However, this does not equate necessarily with a CO2 removal from the atmosphere, but simply a transfer of carbon from another location (Leifeld et al., 2013). To the extent that the amendments improve soil performance and thus increase in situ plant productivity and residue carbon inputs, then the amendment can in fact stimulate real increased CO2 removals. Hence a full life cycle assessment (LCA) approach in which the boundaries of the assessment extend outside the farm to include the GHG and net carbon emissions with the production and baseline utilization of the amendment is needed for an accurate accounting of net CO2 impacts. Since nearly all raw animal manure is land applied (Wu et al., 2013), increasing rates of manure addition at one location would necessitate an equal reduction elsewhere. Hence counting the manure carbon added as part of the overall carbon balance is problematic. For other waste streams, such as municipally sourced compost, where the alternative use may be landfilling, an LCA approach that accounts for net emissions associated with a landfilling alternative vs production and land application of compost could be evaluated as a CO2 emission strategy. One study looking at compost applications to grasslands in California (DeLonge et al., 2013; Ryals and Silver, 2013), estimated a net CO2 removal and GHG reduction of 23 t CO2 equivalent/ha for the first 3 years following compost application, due both to soil carbon increase (discounting added compost carbon) and net avoided GHG emissions from diverting lagoon-stored manure and landfilled green waste to compost. However, similar studies are lacking for other areas in the US. Conversion of Annual Cropland to Perennial Vegetation Probably the most effective means of increasing soil carbon stocks on annual cropland is to convert to perennial vegetation, either for grazing and forage production, afforestation (see previous section), dedicated energy crops (e.g., switchgrass, miscanthus) or as conservation set-aside. Perennial grasses in particular allocate a large fraction of their carbon assimilates to belowground production, which enter the soil organic matter pool through root exudation, sloughing and turnover. Over time, in the absence of long-term soil degradation, SOC stocks can approach or equal those in precultivation native grassland systems. Perennial grasses purposed as dedicated energy crops (see Chapter 4) and established on former cropland typically increase soil carbon stocks which can contribute to the net GHG balance of those systems (Field et al., 2018; Liebig et al., 2008). Conversion to perennial woody vegetation (e.g. forest) increases stocks of woody biomass carbon but may also increase soil carbon stocks. In many cases accrual rates are similar to those for conversion to grassland vegetation (Guo and Gifford, 2002; Post and Kwon, 2000), although some studies suggest soil carbon gains occur under deciduous species whereas soil carbon gains under coniferous woody vegetation is minimal (Laganière et al., 2010; Morris et al., 2007). A number of metaanalyses have documented soil carbon gains on cropland converted to perennial vegetation (Table 3.4). PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 69 TABLE 3.3. Examples of published summaries and meta-analyses for soil carbon sequestration rates with conservation practice adoption on annual cropland. Delta SOC denotes mean change in soil organic carbon stocks (metric tonnes C/ha/yr) with standard error (SE) (where reported) or in some cases values are reported as a representative range in mean annual change rates. Practice(s) adopted ∆ SOC (SE) Region # of field Source (t ha-1 y-1) comparisons Crop rotation Cover crop 0.32 (0.08) Global 139 Poeplau and Don, 2014 Cover crops 0.36 US 31 Eagle et al., 2012 Improved rotations 0.14-0.18‡ US 78 Eagle et al., 2012 Improved rotations 0.1-0.21¶ Global 13 Ogle et al., 2005c (Temperate-dry) Improved rotations 0.17-0.34¶ Global 13 Ogle et al., 2005 (Temperate-moist) Conservation tillage No-till 0.48 (0.13) Global 276 West and Post, 2002 No-till 0.15-0.80¶ Global 160 Ogle et al., 2005 No-till 0.33 US 282 Eagle et al., 2012 No-till 0.30 (0.05) Southeast US 60 Franzluebbers, 2010 No-till + cover crops 0.55 (0.06) Southeast US 87 Franzluebbers, 2010 No-till 0.48 (0.59) Northcentral US 19 Johnson et al., 2005 No-till 0.27 (0.19) Northwest US 40 Liebig et al., 2005 ‡ Range in mean values is for rotation improvements from elimination of bare (summer) fallow and adding perennials forages (for 1-3 years) to annual crop rotations. ¶ These studies reported annual changes in SOC stock (0-30 cm depth) as % changes relative to baseline soil carbon stocks. To convert to an amount (t/ha/yr) for a representative range of SOC change rates, the Committee used 15 tC/ha and 60 tC/ha (0-30 cm depth) as high and low stock values, respectively. These correspond to low and high values for carbon stocks on permanent cropland amount major soil classes in the IPCC soil carbon inventory methodology (IPCC, 2006). TABLE 3.4. Examples of published summaries and meta-analyses for soil carbon sequestration rates following conversion of annual cropland to perennial grassland or forest vegetation. Delta SOC denotes mean change in soil organic carbon stocks (metric tonnes C/ha/yr) with standard error (SE) (where reported) or in some cases values are reported as a representative range in mean annual change rates. Practice(s) adopted ∆ SOC (SE) Region # of field Source (t C ha/y) comparisons Annual cropland to grassland 0.9 (0.1) Global 161 Conant et al., 2017 0.12-1.1¶ Global 58 Ogle et al., 2005 0.28-0.56 Global 76 Guo and Gifford, 2002 0.33 Global 46 Post and Kwon, 2000 Annual cropland to forest 0.27-0.54 Global 38 Guo and Gifford, 2002 0.34 Global 30 Post and Kwon, 2000 0.16-1¶ Global 189 Laganière et al., 2010 ¶ These studies reported annual changes in SOC stock (0-30 cm depth) as % changes relative to baseline soil carbon stocks. To convert to an amount (t/ha/yr) for a representative range of SOC change rates, we used 15 tC/ha and 60 tC/ha (0-30 cm depth) as high and low stock values, respectively. These correspond to low and high values for carbon stocks on permanent cropland amount major soil classes in the IPCC soil carbon inventory methodology (IPCC, 2006). PREPUBLICATION COPY

70 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda TABLE 3.5. Examples of carbon accrual rates for biomass and soil carbon from recent meta-analyses and multisite assessments of temperate zone agroforestry practices. Delta carbon denotes mean change in soil organic carbon or biomass stocks (metric tonnes C/ha/yr) or in some cases values are reported as a representative range in mean annual change rates. NR denotes not reported. Practice(s) adopted Δ C stock yrs Region # of field Source (t ha/hr) comparisons Alleycropping (wood 0.65 6-41 France 13 Cardinael et al., 2017 carbon) Alleycropping (soil carbon) 0.24 6-41 France 6 Cardinael et al., 2017 Agroforestry (soil carbon) 0.43-1.88 NR US NR Eagle et al., 2012 Agroforestry Agroforestry involves the incorporation of trees into agricultural systems, either in combination with annual crops or grazed pastures (the later often referred to as silvopastoral systems). There are a myriad of different practices in which trees may be interspersed with crops (e.g. alley-cropping) or herbaceous forage, or used as border or buffer plantings (e.g., living fences, windbreaks, forest buffers) or used in a time sequence or rotation with annual crops (e.g., improved tree fallows). Regardless of form, the inclusion of perennial woody species in combination with annual crops or pastures typically results in increases in both soil carbon stocks and woody biomass stocks (Table 3.5). Improved Grazing Land Management Grasslands contain some of the highest soil carbon stocks of any managed ecosystems. With the exception of some managed pastures, grazing lands are infrequently or never tilled and the perennial grasses that dominate in most rangelands and pasture allocate a substantial portion of their photosynthetically-fixed carbon below ground, thus supporting comparatively large soil carbon stocks. Grasslands in the US can be loosely categorized into pastures, which are typically in more mesic (i.e., moister) environments, with higher productivity, often with selected (sown) species and more intensive soil management such as fertilization, liming and irrigation. Pastures in some cases may be rotated with annual crops and typically have soils that have been historically tilled at some time in the past. In contrast, rangelands are mainly found in semi-arid and arid climates, are usually dominated by native grassland vegetation, often have never been tilled and usually have no (or minimal) management interventions in terms of fertilizer or irrigation use—grazing management is the primary management ‘lever’ for rangelands. Grazing management influences grassland productivity and carbon storage (Table 3.6) in both pastures and rangelands, via the amount, timing and duration of vegetation removal by grazing animals (Milchunas and Lauenroth, 1993). With excessive grazing pressure (overgrazing), plant productivity and hence carbon uptake is reduced and soil carbon stocks decrease (Dlamini et al., 2016). Hence reducing stocking rates and grazing intensity can allow vegetation productivity to recover and lead to increased carbon stocks. However, other than eliminating overgrazing there is considerable debate regarding the impact on soil carbon of more intensively managed grazing systems (i.e., rotational grazing, adaptive multi-paddock (AMP), mob grazing), compared with less intensively managed, continuous grazing systems. It’s been suggested that management-intensive grazing, utilizing short duration heavy grazing followed by long grazing-free rest periods, can increase soil carbon storage (e.g. Chaplot et al., 2016; Wang et al., 2015), although many experiments do not show a significant difference between light to moderate continuous grazing vs management-intensive grazing systems (e.g., Briske et al., 2008). However, interactions determining vegetation productivity and soil carbon responses for different grazing systems are complex (McSherry and Ritchie, 2013). In general, grazing systems that maintain plant cover and maximize plant vigor and productivity are most conducive for building and maintaining soil carbon PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 71 TABLE 3.6. Examples of soil carbon sequestration rates from recent meta-analyses and field studies of improved grassland management practices. Delta SOC denotes mean change in soil organic carbon stocks (metric tonnes C/ha/yr) with standard error (SE) (where reported) or in some cases values are reported as a representative range in mean annual change rates. Practice(s) adopted Δ SOC (SE) Region # of field Source (t C ha/yr) comparisons Improved fertility 0.57 (0.08) Global 108 Conant et al., 2017 Legume inter-seeding 0.68 (0.22) Global 13 Conant et al., 2017 Improved grazing 0.3 (0.14) Global 89 Conant et al., 2017 Improved grazing -0.8 to +1.3 US 13 Eagle et al., 2012 Adaptive multi-paddock 0.48 US (Texas) 6 Teague et al., 2011 stocks (Eyles et al., 2015). Additional field-based research and improved models are needed to determine the best grassland- and region-specific systems for promoting soil carbon sequestration and their capacity for net GHG reductions and CO2 removal (Conant et al., 2017; Eyles et al., 2015). Rewetting of Organic Soils Organic soils (i.e., peat and muck soils), which develop under wetland vegetation and are saturated with water to the soil surface for a substantial part of the year, have extremely high organic matter content (>12% of soil mass as organic carbon, for a minimum 20 cm horizon depth; FAO, 1998), which can extend to several meters in depth. When converted to agricultural use (or forest plantations) following drainage, liming and fertilization, they can be extremely productive soils, but with rapid rates of organic matter decomposition and carbon loss rates as high as 20 t C ha/yr (Armentano and Menges, 1986). Consequently, restoring the wetland hydrology and perennial vegetation can reverse the processes driving soil carbon losses and greatly reduce CO2 losses compared to drained organic soils and in many cases can reestablish the soil as a net carbon sink, although increased CH4 emissions following rewetting can decrease the overall net sink. In a global review, Wilson et al., 2016 reported that rewetting managed organic soils led to net annual CO2 removals for most soil classes (usually <1 t C ha/yr), varying as a function of thermal regime, site productivity, water table height and time since restoration. Data is most lacking for impacts of restoration on tropical organic soils. Frontier Cropland and Grassland Practices The management interventions described above are well known practices that have been the subject of previous (and ongoing) research and that are, to varying degrees, being practiced now in the US and elsewhere. In most cases, causal mechanisms and relative magnitude of carbon stock responses are known, although important basic and applied science questions still remain. In contrast, there are several approaches that we characterize here as ‘Frontier technologies’ in that they are in an early stage of development, with many unanswered questions regarding rates and capacities for carbon removals, and with little or no deployment of these technologies outside of field research situations. Two of the technologies deal with increasing the residence time and stability of organic carbon in soils and the other two involve development of plants with greatly enhanced capacity to add carbon into the soil. PREPUBLICATION COPY

72 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda Biochar Amendment Pyrolysis (heating in absence of O2) of plant material is an exothermic (i.e., energy-yielding) process that produces volatile compounds and oils (which can be purposed for bioenergy and biomaterials) as well as solid, ‘charred’ organic residues with high aromaticity and reduced O:C and H:C ratios, that are much more resistant to microbial decomposition than the original plant residues (e.g., straw, wood, shells). The longevity of biochar applied to soil can vary substantially depending on the pyrolysis temperature and duration, biomass type, and climate and soil conditions. Highly condensed, low H:C chars derived from wood are highly recalcitrant and can remain in soils over century time scales (Lehman et al., 2015). Naturally occurring pyrogenic carbon (black carbon, charcoal) makes up a substantial portion of the total soil organic carbon in many fire-prone grassland and forest ecosystems (Skjemstad et al., 2002) and additions of charred biomass by Neolithic farmers have resulted in soils with greatly increased organic carbon stocks (e.g. Terra Preta soils of the Amazon; Glaser and Birk, 2012). Pyrolysis is a bioenergy pathway (predominantly for the production of liquid biofuels) that could be part of a bioenergy production strategy (see BECCS Chapter 4) that aims to both displace fossil-based fuels as well as produce a biochar coproduct that can be added to soils for long-term carbon storage. Alternatively, purposed biochar production, with the pyrolysis liquids and volatiles combusted for process heat and/or energy generation, can produce biochar as the main product for soil application to sequester carbon and improve agronomic performance. The extent to which biochar production and addition to soil represents a net CO2 removal depends on a full life-cycle consideration of all GHG emissions associated with the biomass sourcing, its production and use and fossil fuel offset values (Roberts et al., 2010; Woolf et al., 2010). In addition, indirect impacts of soil application of biochar on other biogenic GHG emissions and plant carbon uptake factor into assessments of its potential to reduce net GHGs. Recent metaanalyses suggest that biochar applications may reduce N2O emissions from soil on the order of 10% (Verhoeven et al., 2017), although the response varies between soils, with some biochar additions increasing N2O flux. Sánchez-García et al., 2014 suggest that biochar decreases N2O emissions from the denitrification (anaerobic) pathway but increases emissions from the nitrification (aerobic) pathway, possibly accounting for some of the differences between soil types and soil conditions in biochar effects on N2O flux. Biochar applications also appear to decrease CH4 emissions from flooded (e.g. rice) soils and in acid soils (Jeffery et al., 2016) whereas application to non-flooded soils, with neutral or alkaline pH may reduce rates of CH4 oxidation (a CH4 sink). Finally, biochar application often increases plant productivity and hence plant carbon inputs to soil, particularly in highly weathered and acidic soils, averaging around 10% increase in recent metanalyses (Jeffery et al., 2011; Verhoeven et al., 2017). However, plant responses vary by soil type, biochar attributes and management systems and thus further research is needed to better target conditions under which biochars can best increase plant carbon uptake, reduce non-CO2 GHG emissions and contribute to net CO2 removals from the atmosphere. Deep Soil Inversion Placing organic matter rich surface soil into subsoil layers, through a one-time deep inversion tillage, may substantially increase the overall soil organic matter residence time. For most soils, microbial activities decrease sharply with depth, due to less aeration, cooler temperatures, and sparser and more dispersed organic matter inputs. In addition, the placement of low organic matter content subsoil at the top of the soil profile where majority of root system reside will increase soil carbon sink generated by root system and photosynthesis process. Alcantara et al., 2016 found that deep plowing and burial (to >50 cm depth) of organic matter-rich surface soil horizons, coupled with subsequent enhanced carbon accumulation in surface-exposed and carbon-depleted subsoil horizons, yielded an average carbon accumulation rate of ca. 3.6 t CO2 ha/yr over a 40 time period at several cropland sites in Germany. Testing of deep inversion tillage for carbon storage has recently started in a few other locations (e.g. New Zealand; M. Beare pers. comm.), but not to the Committee’s knowledge in the US. A large portion of US PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 73 cropland in humid regions, particularly in the Central and North-Central regions, might be potential areas for implementation. High Carbon Input Crop Phenotypes The most direct way to increase soil carbon storage is to increase the rate of organic carbon addition to soils in the form of plant residues, particularly belowground. Here we discuss two approaches: 1) Increasing the relative amount of carbon allocated to roots for the annual crops that dominate current agricultural food production systems and 2) developing new perennial food crop species, which have inherently higher belowground carbon allocation and could mimic the high soil carbon storage capacity found in perennial pasture grasses. Breeding for Annual Crops with Increased Carbon Input One the main reasons that annual crops are less effective at increasing SOC stocks compared to perennial grasses is the lesser amount of dry matter allocation belowground to roots and rhizosphere deposition. Most SOC is derived from roots (as exudates and through root death and turnover; Rasse et al., 2005) and thus an option to increase SOC stocks would be to develop crops which allocate more dry matter belowground and/or that have deeper root systems where the decomposing organic carbon compounds would have a longer mean residence time (Kell, 2012). Increasing carbon allocation to roots but not at the cost of reducing aboveground yields will be a key issue for crop breeders and there are reasons that this is possible. Many root characteristics (e.g. architecture, depth distribution, size) are strongly controlled by genetic traits that can be selected for (Hochholdinger et al., 2004; York et al., 2015). Where available nutrients such as P limit yields (common in tropical soils) enhanced root growth can increase nutrient acquisition and increase total assimilation and yield (Lynch, 1995). Finally, sink size inhibition often limits total plant carbon assimilation and thus plant breeding for an increased root carbon sink does not necessarily reduce aboveground productivity and yields (Jansson et al., 2010). Preliminary estimates suggest that widespread adoption of annual crops with enhanced root-phenotypes could increase carbon stocks in US soils by 500-800 Mt CO2eq/yr over several decades (see Table 3.8). Perennializing Grain and Oilseed Crops The dominant commodity crops in the world are grains (e.g. maize, wheat, rice, sorghum, millet) and oilseeds (soybean, pulses, sunflower) which are annuals that have to be replanted each year and which have been bred to maximize dry matter allocation to the harvested seed. Compared to perennial grasses and forbs, current varieties for these annual crops allocate much less carbon to roots and unharvested residues and are less effective at building and maintaining high soil carbon stocks. Thus, developing seed crops with a perennial growth habit and greater belowground carbon partitioning could radically increase the potential for soil carbon gains, if perennialized crops could substitute for a substantial portion of production currently from annual crops. Research to develop perennial analogs for major annual crops has only begun over the past few decades (Glover et al., 2010) and at a limited scale to-date. Approaches include both creating perennial hybrids between related annual and perennial species (e.g., rice (Zhang et al., 2017a); wheat (Hayes et al., 2012); sorghum (Cox et al., 2018)) and domestication of naturally large-seeded wild perennials to further increase seed yield and quality (e.g., intermediate wheatgrass (Culman et al., 2013); perennial sunflower (Vilela et al., 2018)). The major challenge is developing perennial grains with sufficiently high yields to be viable from an economic and food security perspective, with the most promising results to date being for perennial rice (Zhang et al., 2017a). It has been argued that evolutionary tradeoffs between annual and perennial life histories (i.e., high allocation of resources to the seed in annuals vs to belowground structures in perennials) inherently limit the potential for perennial crop replacement of annuals (Smaje, 2015). However, the extended vegetated duration of perennials (hence greater carbon assimilation) and PREPUBLICATION COPY

74 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda positive feedbacks between greater root development and nutrient acquisition in some soils do not preclude the potential for high-yielding grain production from perennial crops (Crews and Dehaan, 2015; Jansson et al., 2010). IMPACT POTENTIAL Carbon Dioxide Removal Capacity For all managed terrestrial carbon sinks, potential carbon removal rates are conditioned on the biological/ecological capacity of the technologies employed as well as economic and social acceptance factors. The majority of published estimates of carbon sequestration rates can be considered technical potentials in which achievable rates per unit area are multiplied by estimates of the maximum feasible land areas for implementation. Such rates and capacities represent an upper limit, that do not fully reflect economic constraints, including land availability for competing uses, or other social or policy constraints. A lower economic potential for carbon sequestration reflects the fact that changing land use/management practices to more carbon friendly ones likely requires an economic incentive (although some amount of carbon stock gains may be achievable at negative costs). Combined ecological and economic analyses have been used to estimate marginal supply curves for carbon storage (e.g., McCarl and Schneider, 2001; Murray et al., 2005; Smith et al., 2008), where each additional increment of storage comes at a higher cost per unit carbon stored. However, relatively few formal analyses have been made of economic potentials, given the complexity and high degree of uncertainty in future market conditions. Finally, the actual potential storage achievable would factor in additional constraints related to social acceptance and policy implementation. Estimates for the major land uses in the US and globally are given below. Conventional Forest Practices Afforestation/Reforestation For the U.S., comprehensive analyses have shown that the potential carbon removal from afforestation/reforestation ranges from 0.0 to 0.45 Gt CO2 annually, over a period of approximately 50 to 100 years (Table 3.7). The soil component of this increase in carbon removal may be significant according to a recent study that estimated a range of 0.05 to 0.08 Gt CO2 yr-1 (Nave et al. 2018). At the global scale, the range of published estimates for afforestation/reforestation is large, from 2.7 to 17.9 Gt CO2 annually based on global analyses using top-down approaches for the high end of the range, and bottom-up approaches for the low end (Table 3.7). The wide range reported in studies represents different assumptions and modeling approaches, and variable prices or incentives for implementing activities. At the low end of the range, the assumed price of carbon is low and secondary impacts are few— for example, that there is sufficient marginal agricultural land available and landowners would be willing to participate in mitigation. At the high end, the price of carbon would be as high as $100 per ton of CO2, and 10’s of millions of hectares would be incentivized to convert from crop or grass production to forest. Food production would be significantly impacted, so that the total land area that could be reforested or afforested is constrained by the higher value of alternative land uses (i.e., food price impacts from cropland conversion). The net CO2 reduction may be reduced over time by leakage (i.e., timber harvest elsewhere to meet continued economic demand for forest products), and concerns about permanence (subsequent clearing or natural disturbances). PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 75 TABLE 3.7. Total global and U.S. estimates of carbon removal potential for forestry activities. Study/Citation Estimate Scope Gt CO2eq/y Global Estimates Nabuurs et al., 2007 4.0 Afforestation/Reforestation Griscom et al., 2017 2.7-17.9 Afforestation/Reforestation Smith et al., 2016 4.0-12.1 Afforestation/Reforestation Nabuurs et al., 2007 5.8 Improved forest management Griscom et al., 2017 1.1-9.2 Improved forest management US Estimates Nabuurs et al., 2007 0.445 Afforestation/Reforestation McKinley et al., 2011 0.001-0.225 Afforestation/Reforestation Jackson and Baker, 0.15-0.4 Afforestation/Reforestation 2010 Nabuurs et al., 2007 1.6 Improved forest management McKinley et al., 2011 0.029-0.105 Improved forest management Improved Forest Management The potential of improved forest management ranges from 0.03 to 1.6 GtCO2/yr for the U.S., and from 1.1 to 9.2 GtCO2/yr for the world based on published studies (Table 3.7). Improving forest management on existing forest land has few or no constraints from competition with other land uses compared with reforestation/afforestation, since this activity involves no change in land use, although the cost may be higher because of the large area that should be treated to achieve similar results, and forest management would need to involve a much larger percentage of forest landowners than ever have participated in any incentive program. As with afforestation/reforestation, both leakage and risk of reversal from disturbances could reduce potential carbon removal gains over time. Near-term Forest carbon removal potentials For a recent and practical reference point, adding up the forestry activities proposed by countries in their intended nationally determined contributions (INDCs) to the Paris Climate Agreement yields a total expected global carbon removal benefit from afforestation/reforestation and improved forest management (excluding avoided deforestation) of approximately 1.0 Gt CO2. This is comparable to the low end of the estimates reported above, since the INDCs were purposely modest, representing what different governments considered achievable toward a target of limiting climate change to +2.0 deg. C. A modest program of afforestation/reforestation near the lower end of published ranges could be 0.15 Gt CO2 annually for the U.S. and 1.0 Gt CO2 annually for the world, taking into consideration availability of land and secondary impacts (described later in this chapter). Considering land constraints, a practical upper limit for carbon removal from afforestation/reforestation would be about 0.4 Gt CO2 annually for the U.S. and not likely more than 6.0 Gt CO2 annually for the world (See Box 3.1). Similarly, the near-term carbon removal from improved forest management is likely to be near the lower end of these ranges, about 0.1 GtCO2/yr for the U.S., and 1.5 GtCO2/yr for the world. The upper limit of a more aggressive program to improve forest management could achieve a rate of carbon removal of 0.2 Gt CO2 annually for the U.S., and 3.0 Gt CO2 annually for the world. Frontier Forestry Practices Increased Use and Preservation of Harvested Wood Products If most (up to 80%) of discarded wood products and associated wood wastes from current harvest and manufacturing were placed in a landfill designed for slow decomposition, then this would create an additional sink of 0.1 to 0.3 Gt CO2/yr in the US and 0.2 to 0.8 GT CO2/y worldwide which could be PREPUBLICATION COPY

76 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda extended indefinitely as long as such landfills continued to be constructed. Preservation of currently harvested wood plus increasing harvest of secondary forests to use all available growth (sustainable harvest) has the capacity to remove a total of 0.1 to 0.7 Gt CO2 annually in the U.S. and 0.8 to 9.3 Gt CO2 annually worldwide. This could be accomplished without involving protected or intact forests, or affecting food supply or biological diversity. Zeng et al., 2013 estimated that the global potential of green-tree burial was between 1.0 and 3.0 Gt CO2 yr-1, with the lower end of this range representing roughly doubling the current global harvest and affecting about 800,000 hectares of forest land. They excluded agricultural land, protected areas, inaccessible forests, and wood used for other purposes such as timber and paper. To date, this proposed approach has not been tested though the technology is simple and easily applied. Conventional Cropland and Grassland Practices Over the past twenty years there have been several estimates of the soil carbon sequestration potential globally and for the US (Table 3.8). Nearly all represent a technical or ‘biophysical potential’ assuming nearly complete practice adoption. As such, these represent upper-bound estimates of the carbon sequestration potential for the practices considered. However, the estimates do account for limitations relating to land availability, such as the limited land use conversions to high carbon storage practices like permanent grassland setaside that could occur without compromising food and fiber production. Most published estimates use aggregate data stratified by broadly defined land use and climate categories (in some cases soil type) together with representative per ha rates for different land use/management practices determined from long-term field experiments or other (e.g. chronosequence) measurements. BOX 3.1 Summary of Terrestrial Carbon Removal Estimates Land availability and the degree of adoption of carbon removal practices are major constraints on land-based NETs. Achieving the uppermost estimates for conventional practices (afforestation/reforestation, improved forest management, improved cropland and grassland management) shown in Tables 3.7 and 3.8 are highly unlikely because constraints on availability of land for conversion from current uses, leakage and risk of reversal from disturbances and cost and behavioral constraints to achieving full adoption on all available land. Weighing these constraints, the Committee here provides estimates for ‘practically achievable’ CO2 removal rates that are feasible, with current technologies, without involving use of protected forest areas or compromising food supply or biodiversity from large-scale land conversions. Estimates are included for additional carbon removal capacity from frontier technologies that are not constrained by land availability (e.g., carbon burial approaches, substitution of high carbon input phenotypes within current land use area distributions). Implementation Forest Agriculture Total U.S. -------------- Gt CO2/yr -------------- Practically achievable 0.25 0.25 0.5 Practically achievable + frontier technology 0.35 0.8 1.15 Global Practically achievable 2.5 3 5.5 Practically achievable + frontier technology 3.3 8 11.3 PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 77 TABLE 3.8. Published estimates of global and US soil carbon sequestration potential on managed non- forested lands. All values reflect technical or ‘biophysical’ potential estimates that are not constrained by carbon price or policy design. Study/Citation Estimate Scope Global Estimates Gt CO2eq/yr Paustian et al., 1998 1.5-3.3 Improved cropland management, setaside, restoration of degraded land Lal and Bruce, 1999 1.7-2.2 Improved cropland management, restoration of degraded land¶ IPCC, 2000 3 Improved cropland & grassland management, setaside, agroforestry, restored peat soils Lal, 2004 1.5-4.4 Improved cropland & grassland management, setaside, agroforestry, restored degraded lands Smith et al., 2008 5-5.4 Improved cropland & grassland management, setaside, agroforestry, restored degraded lands, restored peat soils§ Sommer and Bossio, 2.5-5.1 Improved cropland & grassland management, setaside, 2014 agroforestry, restored degraded lands Paustian et al., 2016b 2-5 Improved cropland & grassland management, setaside, agroforestry, restored degraded lands, restored peat soils Paustian et al., 2016b 4-8 Potential from the practices above, plus unconventional technologies including high root C input crop phenotypes and biochar amendments Griscom et al., 2017 3-5 Conservation ag, agroforestry, improved grazing, avoided grassland conversion, biochar Fuss et al., 2018 2.3-5.3 Improved cropland & grassland management, setaside, agroforestry, restored degraded lands, restored peat soils US Estimates Mt CO2eq/yr Lal et al., 1998 275-639 Land conversion and setasides, restoration of degraded land, improved management on cropland Sperow et al., 2003 305 Improved cropland management, setaside of marginal (highly erodible) cropland to grassland Sperow, 2016 240 Improved cropland management, setaside of marginal (highly erodible) cropland to grassland Chambers et al., 2016 250 Improved cropland and grassland management, setaside of marginal (highly erodible) cropland to grassland † Paustian et al., 2016a 500-800 Deployment of enhanced root phenotypes for major annual crops (assumes 2X root carbon input and downward shift in root distribution equivalent to native prairie grasses) ‡ ¶ An additional 1-1.5 GtCO2eq emission reduction was projected from biofuel CO2 offsets § This study also included an estimate of ‘economic potential’: about 2.5 GtCO2eq/y was achievable for < $50 tonne CO2 † Based on estimate for widespread adoption of USDA/NRCS conservation practices on all private lands ‡ Excluded non-irrigated semi-arid cropland with major water limitation on production Frontier Cropland and Grassland Practices Detailed analyses of the global sink potential of the frontier practices described earlier (i.e., high carbon input phenotypes, deep soil inversion) have not been published, with the exception of for biochar amendment. Recent estimates of the global carbon sink capacity of large-scale deployment of biochar application to soil suggest Gt per year potential, e.g. 6.6 Gt CO2eq/yr by Woolf et al. 2010, 2.6 Gt CO2eq/yrfrom Smith, 2016 and 0.5-2.6 Gt CO2eq/yr by Fuss et al., 2018. Since biochar applications mainly target existing current cropland where large addition rates (up to 50 tonne C/ha; Smith 2016) are agronomically feasible, there is no land constraint for application, although there is for biomass feedstock sourcing (see Land Requirements section below). Key uncertainties remaining include net life cycle GHG PREPUBLICATION COPY

78 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda impacts, long-term impacts on crop productivity and economic feasibility of large-scale biochar deployment. Other Radiative Impacts Forest The impact on climate from reducing CO2 by afforestation/reforestation may be partially offset by albedo changes (the proportion of radiation reflected by the land surface). Generally, an increase in forest cover reduces surface reflectivity causing more surface warming, whereas a harvest increases surface reflectivity causing a cooling effect, implying that the effect of increasing harvest or forest land clearing would have a less negative impact on climate than indicated by only accounting for GHG changes. Albedo effects are typically larger in the years immediately following harvest, last longer for land use changes, and are most significant in response to changes in conifer coverage above the snow line (Cherubini et al., 2012; Holtsmark, 2015). These effects on climate forcing are usually less influential than the effects of changes in GHGs but may be highly variable (Anderson-Teixeira et al., 2012). In general, afforestation/reforestation in boreal zones will have a warming effect that is larger than the cooling effect of reducing GHGs, and the opposite effects in the tropics. In temperate zones, the effects are highly variable in space, depending on vegetation type, timing of snow cover, slope and aspect, and other factors. Recent work has extended the analysis of radiative effects to include non-radiative processes such as air turbulence which may have larger and locally more variable effects than albedo changes (Bright et al., 2017). Cropland/Grassland In designing and implementing CO2 removal strategies involving agricultural land management, impacts on other greenhouse gases, particularly nitrous oxide (N2O) and methane (CH4), are of paramount importance. Agricultural soils are the single largest source of N2O, driven by the large additions of N to promote plant growth. Flooded cropland soils (rice) are a major CH4 source. Soil-based CO2 removal strategies should avoid requiring additional N fertilizer, which has large embodied fossil carbon emissions for its manufacture, in addition to impacts on N2O emissions (van Groenigen et al., 2017). Thus any additional N requirements to support increased plant production and carbon inputs should be sourced from more efficient use of N already being applied to agricultural soils as well as from biological N fixation. Management options that offer synergisms with both increased soil carbon stocks, reduced N2O (and CH4) emissions should be prioritized. Agricultural practices to increase CO2 removal likely have minimal impact on other (e.g. aerosols, albedo) radiative forcing factors. For example, conversion of winter-fallow bare soils to lighter surface of senescent winter vegetated cover, as well as reduced tillage systems that retain surface residues, increases albedo in most cases, providing a small cooling effect (Davin et al., 2014). Most other agricultural carbon removal practices will have negligible impacts on albedo. Current Commercial Status The current commercial status of carbon removal via terrestrial land use options is still limited but growing. There currently are only a few compliance cap-and-trade markets for GHG reductions currently in operations (e.g. EU, California) and, in most of those, land-use-based offsets (i.e., allowable emission reduction coming from non-capped entities) play a minor role, if included at all. Hence the PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 79 demand for, and thus the monetary value of, terrestrial carbon storage is still low. More direct financing of carbon sequestration projects involving agriculture and/or forestry exists in the voluntary carbon/emission reduction markets. In 2016 the total volume of forestry and land use project carbon reductions was 13.1 MtCO2e, at an average price of $5.10 per tonne CO2, for an aggregate value of $67 million. Over 95% of the CO2 reduction activities focused on forest biomass carbon, of which REDD+ (Reduced Deforestation and Forest Degradation Plus) projects dominated with 75% of total CO2 reductions (Hamrick and Gallant, 2017). In the US, an indirect measure of the commercial value, to the land owner, are the subsidies paid by government for implementation of conservation practices that also promote carbon sequestration. In many cases these payments address multiple environmental outcomes and are often in the form of cost shares or loans, as opposed to direct payments for tonnes of carbon sequestered as in voluntary or cap-and-trade markets. Chambers et al., 2016 estimated that conservation payments that promoted soil carbon storage averaged around $60M/yr from 2005 to 2014, with an estimate carbon storage increase of 13 to 43 Mt carbon on US cropland, which equates to a carbon price of $5-17/t CO2e. Land Requirements and Competition for Land among Alternative NETs Although scientific uncertainty about the amount of carbon sequestered locally from improved land management practices is low, uncertainty about far-field impacts caused by the long-term effects of climate change and the economic feedback is high. An important consideration is that land requirements and competition for land cuts across several of the NETs in this report—particularly, afforestation/reforestation and BECCS; soil carbon sequestration and biochar use land, but do not compete for land as the land can still be used for the same purpose. Each of these technologies were evaluated separately, but their maximum potential cannot all be realized simultaneously. Here we examine the land requirements for these different activities and compare them with current land uses and potentially available land for additional carbon removal. Table 3.9 summarizes the practically achievable ranges of carbon removal from forestry and BECCS, including an estimate of the land area required for each. TABLE 3.9. Range of practical annual carbon dioxide removal fluxes and associated land requirements for estimates presented in this report. See chapter 4 (BECCS chapter) for source of estimates for BECCS. Activity categories that do not require land use shifts (e.g., agriculture, improved forest management) are not included in the table. Activity category carbon carbon removal- Area-low Area-high removal-low high (Mha) (Mha) (GtCO2/y) (GtCO2/y) U.S. Afforestation/reforestationa 0.15 0.4 3-4 16-20 Global afforestation/reforestationb 1.0 6.0 70-90 350-500 U.S. BECCSc 0.52 1.5 0 78 Global BECCSc 3.5-5.2 10.0-15.0 0 380-700 aBased on Jackson and Baker, 2010. Would require a CO price of $15/t for carbon removal-low and $50/t for carbon removal- 2 high. bBased on Griscom et al., 2017 and Smith et al., 2016. cRefer to chapter 4 for sources of estimates. At the low end, biomass is sourced from existing land uses so no new land dedicated to BECCS is required. PREPUBLICATION COPY

80 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda Forestry Even relatively low amounts of carbon removal and carbon sequestration from afforestation/reforestation (0.15 GtCO2/y) in the U.S. would require converting 3-4 million hectares of non-forested land to never-harvested forest. Improved forest management would require changes on 11- 19 million ha of existing forest for a modest increase in carbon removal of 0.1 GtCO2/y. Global land requirements for afforestation/reforestation to achieve 1.0 GtCO2/y, and improved forest management to achieve 1.5 GtCO2/y, are 70-90 million hectares and more than 1,000 million hectares of existing forest land, respectively. These levels of activity should be achievable at low carbon prices (i.e. $10-$50/Mg CO2), but safeguards would be necessary to reduce negative impacts and insure permanence, and implementation would require a very high participation rate by landowners. Although higher levels of carbon removal are technically achievable from afforestation/reforestation, attaining 0.4 GtCO2/y in the US would require much larger areas of land (up to 20 million hectares) and a high carbon price (i.e. >$50/Mg CO2). Globally, the area of land needed to support a 6.0 GtCO2/y carbon removal from afforestation/reforestation would require up to 500 million hectares of land. These higher levels of afforestation/reforestation would have significant negative impacts, including reduced food production and increased food prices. Higher levels of improved forest management in the U.S. to achieve 0.2 GtCO2/y would require 22-38 million hectares of existing forest land, and globally, 3.0 GtCO2/y would require more than 2,500 million hectares of existing forest land. These management changes could result in leakage by shifting some harvesting elsewhere, and would require participation by most private land owners, which would be difficult to achieve. Agriculture The majority of management practices considered for carbon removal into agricultural soils do not involve changes in land use and hence would largely be operative on the existing land base devoted to agricultural production. To the extent that many practices improve soil health and productivity, carbon removal activities could contribute towards meeting increased food and fiber demands without increasing the area of land under agriculture. The main activity that could contribute to displaced agricultural production and potential land use conversions elsewhere would be setting aside cropland into perennial grassland (or forest, see above), for conservation purposes. To the extent that degraded or otherwise marginal agricultural lands were targeted for set-asides, leakage effects (i.e., displacement of agricultural production resulting in land use conversion and soil carbon losses elsewhere) would be minimal. Currently, roughly 9.8 million ha of cropland is in set-asides under the US Conservation Reserve Program (CRP). At its height during the late 1990’s, US CRP area exceeded 13 million ha (Mercier, 2011) and that level of land set-aside resulted in estimated leakage of around 20% of the sequestered carbon (Wu, 2000), although other authors suggest potentially larger leakage (Murray et al., 2007). BECCS As discussed in the BECCS chapter, the Committee calculated the low-end estimate for this approach assuming that there are sufficient sources of biomass from existing land uses (e.g., wood and other unutilized organic waste) such that no land use changes would be necessary. Dedicated land requirements for the high-end estimate for the U.S. are based on reported economically feasible energy crops that could produce up to 0.65 Gt CO2e annually (DOE, 2016), and assuming an average productivity of 18 t CO2e per ha, this level of production would require 36 million hectares of land. Scaling this estimate up to our high end of 1.5 Gt CO2e indicates an area requirement of 78 million PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 81 hectares, equivalent to almost 20% of the current agricultural land base (Table 3.9). Globally, Smith et al., 2016 estimate the land area required to deliver 12 Gt CO2eq/yr at approximately 380-700 Mha in 2100 for high productivity energy crops such as willow and poplar short-rotation coppice, and miscanthus. Therefore, the large-scale implementation of BECCS is expected to compete with terrestrial carbon capture and storage initiatives, as well as with food production (e.g. Smith et al., 2010) or the delivery of other ecosystem services (e.g. Bustamante et al., 2014). Aggregated Carbon Removal Land Requirements and Availability Of “Marginal Land” The total area of land managed for agriculture and forestry in the U.S. is 791 Mha and globally, 7130 Mha (Table 3.10). Some of this land may be considered marginal by landowners, which can be used as an indicator of the amount of land that could be shifted to another use without having significant impacts on production of necessary services, particularly food. There is no generally accepted definition of marginal land (it is an economic decision that varies over time, or based on subsistence needs), but an indicator in the U.S. is the ca. 10 Mha of farmland currently enrolled in the CRP program. A similar global estimate is more difficult to determine, but one recent study estimated that the global pool of marginal land is about 1300 Mha, though this land supports about one-third of the world’s population and so only a fraction of this total amount would be available for afforestation/reforestation and BECCS (Kang et al., 2013). Another global analysis estimated that there are 760 Mha of land available and suitable for afforestation in IPCC non-Annex 1 countries which mostly involve countries in tropical and sub-tropical biomes (Zorner et al., 2008). Because new land dedicated to BECCS would likely be planted with fast-growing species like miscanthus (on more productive sites), the area required per unit of carbon removal would be much less than that required for afforestation/reforestation. A study by Humpenöder et al., 2014 highlighted the trajectories and relative land requirements for afforestation and BECCS (Figure 3.3) using a land-based modeling approach that simulated a similar level of carbon removal as the high-end calculation in this report. They suggest that the area required for BECCS is much lower than that required for afforestation (Table 3.9) based on strongly increasing yields per hectare for herbaceous biofuel, up to 25-30 t C ha/yr, compared with afforestation yields that are estimated at a much lower rate for natural regeneration of forests, about 2-6 t C ha/yr (Table 3.1 this report). However, as noted in the BECCS chapter, the temporal nature of bioenergy production should also be considered. TABLE 3.10. Managed land area by selected land use categories in 2015 (Mha).a Category Forest Cropland Grassland Total b U.S. 293 163 325 781 c World 2429 1426 3275 7130 a U.S definitions according to EPA 2017. World definitions according to FAOSTAT database. b Data from EPA, 2017. c Data from FAOSTAT database accessed 3-11-2018 PREPUBLICATION COPY

82 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda FIGURE 3.3. Simulated time series of global land use for business as usual (BAU), afforestation (AFF), BECCS, and AFF + BECCS. Both increased afforestation and BECCS require additional land that significantly impacts land used for food and pasture. By the end of the century, global area required for these activities is more than five times larger in case of afforestation (∼2800 million ha) compared to BECCS (∼500 million ha). SOURCE: Adapted from Humpenöder et al., 2014. In summary, at the low end of the Committee’s estimates for carbon flux and capacity for afforestation/reforestation and BECCS, there is unlikely to be any significant competition for land among these two NETs (because BECCS biomass could be sourced entirely from existing land uses), nor would there be competition for land currently required for food production. Land requirements for afforestation/reforestation both in the U.S. and globally could be met by tapping into land considered “marginal” for food production, although globally, some of the required marginal land is used to support population needs for food and fiber. At the high end shown in Table 3.9, the total area required for afforestation/reforestation and BECCS combined would likely exceed the availability of marginal land and considering the additional impact on land required for food production, this level of carbon removal is not likely to be realized. ESTIMATED COSTS OF IMPLEMENTING TERRESTRIAL CARBON SEQUESTRATION AT SCALE The direct costs of establishing new forests and performing management activities in different regions are well known based on experience, and there have been several studies that reveal how landowners would respond to various carbon price levels. However, scaling direct costs to high levels of activity is a challenge. Estimates of indirect costs and impacts on carbon associated with collateral effects such as offsetting land use change and commodity production are available, but knowledge of these effects is limited by ability of macro-economic models to simulate complex responses. The costs of different forestry practices are highly variable by practice and region, and their estimates when scaling up to high level of carbon removal differ wildly depending on the method used to estimate feasibility (Alig, 2010; Figure 3.4). Engineering methods to estimate costs result in higher costs at the low end of total tons sequestered, and lower costs at higher levels of carbon removal, compared with econometric or optimization approaches that account for market adjustments. PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 83 Cost estimates for implementing conventional soil carbon sequestering practices (e.g., planting cover crops, changing tillage practices and crop rotations) are readily available as state- and regionally-specific farm budgets (e.g., https://www.ers.usda.gov/data-products/commodity-costs-and-returns/). However, more general information on future costs and projected willingness to adopt carbon sequestering practices as a function of carbon prices are mainly available from academic studies. While varying for different geographic areas, farming systems and practices (Alexander et al., 2015), many estimates suggest significant adoption rates of improved cropping practices could occur at values <$50 tCO2eq (Tang et al., 2016). In a global analysis, Smith et al., 2008 estimated economically-achievable GHG reductions (>90% of which were from soil carbon sequestration) of 1.5, 2.2 and 2.6 GtCO2eq/yr at carbon prices of 0-20, 0- 50 and 0-100 USD/tCO2eq, respectively. Economic feasibility and marginal cost curves for non- conventional (frontier) carbon sequestering technologies described earlier are not known. An economic optimization model used to estimate forestry, agriculture, and bioenergy supply functions for emissions reductions in the US (Figure 3.5) indicated that at low carbon prices, forestry and agriculture opportunities were about equal in terms of economic potential carbon removal and higher than bioenergy. At higher prices, agriculture had the largest impact on carbon removal followed by afforestation and forest management. FIGURE 3.4. Comparison of marginal cost curves for forest carbon sequestration in the United State by Lubowski et al., 2006 with optimization models (Adams et al., 1993; Callaway and McCarl, 1996) and bottom-up engineering cost methods (Richards et al., 1993). “This study” in the graphic refers to Lubowski et al., 2006. Adapted from Alig, 2010. PREPUBLICATION COPY

84 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda $15/tCO2e $30/tCO2e $50/tCO2e 0 -100 Million tCO2e (cumulative) Forest management -200 Afforestation Agriculture -300 Bioenergy -400 -500 FIGURE 3.5 Mitigation potential of land-based NETs in the United States for three carbon price scenarios ($15, $30 and $50 per tonne CO2e). Negative number indicates removal of CO2 from the atmosphere. Estimates from the FASOM model reported by Baker et al., 2010 and Jackson and Baker, 2010. SECONDARY IMPACTS Impacts of changes in land management on biodiversity, water, and other land attributes may be positive or negative depending on types of changes in land cover and site-specific land characteristics. For example, timber harvesting changes forest structure, composition and productivity, which in turn affects many other properties and services of forests such as wildlife habitat, biodiversity, and runoff (Venier et al., 2014). These impacts may be profound and long-lasting by establishing spatial patterns that broadly affect forest and landscape ecology, similar to the effects of many natural disturbances. Some studies show that managing for carbon in particular forests can decrease biodiversity, because habitat is reduced for species that depend on disturbance to create habitat (Lawler et al., 2014; Martin et al., 2015). Impacts of harvesting roundwood on forest ecology are significant and should be considered in formulating policies as well as assessed through observations, although although additional studies are needed to determine how to assess such impacts at different scales. Turner, 2010 highlighted key concepts of how disturbances and recovery have profound and long-lasting impacts at stand and landscape scale as forests recover. Fundamental ecosystem processes and functions are affected. For example, productivity and mortality are strongly related to time since disturbance, and disturbances create spatial heterogeneity which can be essential for some wildlife species. Co-Benefits of Soil Carbon Sequestration (SCS) Practices In addition to CO2 removal, increasing soil carbon contents provides benefits in terms of soil health and provision of other ecosystem services, including enhanced carbon and nutrient pools, water storage, improved soil structure, aggregation and water and air infiltration as well as reductions in soil PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 85 erosion and enhanced soil biodiversity (Al-Kaisi et al., 2014; Lefèvre et al., 2017). In a study evaluating ecosystem services in conservation planning for targeted benefits vs co-benefits, it was found that including ecosystem services within a conservation plan may be most cost-effective when they are represented as substitutable co-benefits/costs, rather than as targeted benefits (Chan et al., 2011). Well- managed soils that enhance soil carbon also promote soil biodiversity which enhances the function and metabolic capacity of soils and plays a crucial role in increasing food production and soil resilience to climate change. Increasing soil organic matter contents will facilitate i) nutrient storage in SOM; ii) nutrient recycling from organic to plant-available mineral forms; and iii) physical and chemical processes that control nutrient sorption and availability. Managed soils represent a highly dynamic system, and it is this very dynamism that makes soils function and supply ecosystem services (Lefèvre et al., 2017). RESEARCH AGENDA Basis for Research Budget Estimates The proposed budget for terrestrial basic and applied research components for a national research agenda are in line with the current budget for various soil and plant research established by NIFA and AFRI-USDA research programs. The AFRI research program for various components allocates $500K to $750K/year per research project for a period of 3 to 5 years for research programs such as Plant Health and Production, Bioenergy, and Animal Health. Generally, in the estimates quoted for different research activities, the Committee allocated $300K per scientist year for post-doc level research, plus funds for administrative and technician support, equipment, travel, publications costs, and other expenses specific to the research described. The cost of each component of the research agenda is summarized in Table 3.11. Basic Research High Carbon Input Crop Phenotypes Research is needed to develop crop varieties with altered root morphology and biomass (e.g., more roots, deeper root distributions, more recalcitrant to decomposition) to add and maintain high root carbon inputs to soil, while maintaining high aboveground yields. Specific research areas include crop breeding and selection for annual crops with larger and deeper roots, perennialization of major grain and oil seed crops, advanced root phenotyping technologies, new crop performance trials, soil and ecosystem responses to new crop varieties. Such a research program can build further on the recently initiated ROOTS (Rhizosphere Observations Optimizing Terrestrial Sequestration) program established by ARPA- E1. The cost for this research is approximately $40-50 million/yr over an initial 20 year period and could be done through USDA, NSF and/or DOE. For comparison, current annual R&D funding for conventional crops improvement and genetics in the US is approximately $1.5 billion from the public sector and $1.8 billion from the private sector (Fuglie and Toole, 2014). 1 Projects in the ARPA-E ROOTS program seek to develop advanced technologies and crop cultivars that enable a 50% increase in soil carbon accumulation while reducing N2O emissions by 50% and increasing water productivity by 25%. For more information see: https://arpa-e.energy.gov/?q=arpa-e-programs/roots. PREPUBLICATION COPY

86 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda TABLE 3.11. Costs and components of a terrestrial NET research agenda Recommended Research Estimated Time Frame Justification Research (yr) Budget ($/y) High carbon Input Crop $40-50M 20 years The DOE/ARPA-E ROOTS program is Phenotypes currently funded for $35M total, allocated to 10 multi-year projects. Thus proposed funding is a 4-5X increase over this level. Soil dynamics at depth $3-4M 5 years Funding for initiation of 4-6 projects per Basic Research year Harvested Wood Preservation $2.4 M 3 years Funding for initiation of 3 multi-year projects at $800K each, involving representative locations Biochar Studies $3M 5-10 Funding for 3-5 projects per year to assess biochar amendment impacts for different management systems and soil types Monitoring of forest stock >5M ≥3 System development - $1M /yr for 3 Development and Measurement/Monitoring enhancement projects years Continuous operation - $4M/yr to staff a small office to analyze data, coordinate field checks, develop reports Improving international forest monitoring and reporting - 10 to 20 times the amount needed in the U.S. A national on farm monitoring $5M Ongoing Augmentation of USDA’s existing NRI system system Data-model platform for $5M 5 Initial development focuses on systems predicting and quantifying integration, including of existing data agricultural soil carbon removal sources and models. and storage Forest demonstration projects: $4.5M 3 Demonstration projects to improve increasing collection, disposal, disposal and collection of wood products and preservation of harvested after use (three 3-year projects at wood; and forest restoration $500K/y each); 3 multi-year projects for preserving harvested wood in different Demonstration environments (500K/yr each); and 3 multi-year projects to demonstrate carbon benefits of forest restoration in different geographic regions ($500K/yr each). Experimental network $6-9M ≥12 10-15 sites at a cost of $600K/yr per site improving agricultural soil carbon processes. PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 87 Social sciences research on $1 M 3 Extension and outreach educational improving landowner responses programs for transferring research to incentives and equity among findings and technologies to farmers and landowner classes. practitioners. Funding for initiation of 3 multi-year projects Deployment Research on GHG and social $1M 3 Funding for initiation of 1 multi-year impacts of reducing traditional project uses of biomass for fuel Scaling up agriculture $2M 3 Support for initiation of 4-5 regional sequestering activities projects per year to identify solutions to overcome barriers to adoption Soil Carbon Dynamics at Depth Research is still limited on the controls on decomposition and stabilization of soil organic matter in subsoil, i.e., below 30 cm. Several promising technologies to increase soil carbon stocks (e.g., prairie restoration, deep soil inversion and carbon burial, enhanced root phenotypes) are based on increasing carbon additions to subsoil layers. Studies are needed to determine how stabilization and residence times of organic residues vary as a function of depth gradients in aeration, microbial community composition and as a function of soil physiochemical properties (e.g., soil texture, soil mineralogy) and plant (root) residue composition. The cost for this research is approximately $3-4 million/yr for 5 years and could be conducted by USDA and NSF. Harvested Wood Preservation Research is needed in two areas: (1) Landfill designs for achieving the lowest possible rate of wood decomposition. In the past, reducing wood decomposition has not been a stated goal of landfill designs; rather, they have been designed to contain buried waste, collect contaminated precipitation that percolates through the waste (leachate), and collect and control gas emissions. There is a large body of needed landfill design research at several representative sites that forms a foundation for focused attention on minimizing wood decomposition (USFS, EPA and NSF; Cost: $2.4 million/yr for 3 projects lasting 3years). (2) Integrated assessment of net greenhouse balance, costs, and required land, including the implications for worldwide consumption of wood products and their lifecycle emissions. Any change in consumption of wood products will have widespread effects on emissions related to other sectors of the economy, for example, construction or material transportation. (Existing research capacity and cost: refer to integrated assessment section in BECCS chapter.) Biochar Amendment Studies While much is now known on the residence time in soil for different types of biochar and how biochar characteristics vary as a function of the feedstock and type of pyrolysis process used, additional research is needed to assess secondary impacts of different biochars on crop performance, nutrient cycling and retention and N2O and CH4 emissions from soils, all of which affect the net GHG consequences of biochar amendments. In addition, research to produce full life cycle analyses that include consideration of alterative fates/uses of feedstocks from which the biochar is produced (Paustian et al., 2016a) are needed for a comprehensive assessment of net carbon removal potentials from biochar amendments. PREPUBLICATION COPY

88 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda Development & Measurement and Monitoring Monitoring of Forest Carbon Stock Enhancement Projects On private and public forestlands, the USFS should develop a plan to monitor recommended carbon stock enhancing activities, conduct statistical sampling of total ecosystem carbon stored in a subset of projects, and develop local “climate impact factors” that also account for biophysical effects. It is time consuming and expensive to directly measure the effects of many small projects on net GHG emissions; therefore, approaches that could achieve accurate average estimates for aggregates of projects (based on remote sensing and validated expansion factors) are needed to reduce transaction costs. Additional needs are to monitor leakage (which is a global phenomenon) and to use the monitoring system to attribute observed changes in carbon removal to management activities vs. increasing CO2 or climate change. Monitoring leakage could require a new LiDAR satellite dedicated to mapping global forestry activities. Knowledge and monitoring of lateral transfers of carbon from land to inland waters are lacking and are not currently detected by remote sensing or operational field inventories. There is a substantial existing capacity in remote sensing and field monitoring on which to build the additional needs described here. For example, the USFS FIA program is funded at approximately $70 million per year, collects continuous field data on status and trends of U.S. forests, and collaborates with NASA on developing methods to integrate remote sensing data with field data. Internationally, the status of monitoring is highly variable, with many countries lacking field measurements and capacity to implement monitoring programs. However, there is a significant international aid effort to improve capacity in forest monitoring at the country scale, as well as advancing research on global monitoring capability using satellites. The research cost for the U.S. only is approximately $1.0 million/yr for 3 years for system development and continuous operation would cost approximately $4.0 million per year to staff a small office to analyze data, coordinate field checks, and develop reports. A significant contribution to improving international forest monitoring and reporting, including detection of leakage worldwide, would require about 10 to 20 times the amount needed in the U.S. A National On-Farm Soil Monitoring System A national on-farm soil monitoring system should be fully implemented by USDA on existing National Resource Inventory (NRI) points on cropland and grassland. Full buildout to approximately 5000-7000 NRI locations with soil sampling and analysis carried out at intervals of 5-7 year (on an annual rotating basis similar to the FIA system) is recommended. Similar systems already exist in many countries, including in the EU, Australia, NZ, and China (van Wesemael et al., 2011). This system would provide an ongoing data stream to improve national-scale soil carbon inventory systems and reduce uncertainties. Such a monitoring system has the potential for broader utility for tracking and assessment of soil health and to support long-term sustainability of soils for the U.S. agricultural economy, which has a GDP of over $135 billion from farm output alone. The cost for this system is $5 million/yr as an augmentation of USDA’s existing NRI system. Model-Data Platform for Quantifying and Forecasting Carbon Removals by Soils For implementation of soil carbon sequestering practices at Gt/yr scales, improved platforms for quantifying soil carbon storage accurately and cost effectively are needed (Figure 3.6). Such systems should integrate data from a repository of existing and new experimental field sites to inform process- PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 89 based land use/ecosystem models that can be independently validated using on-farm soil monitoring networks. Spatial data layers of model drivers (e.g., weather, soil maps, topography), together with FIGURE 3.6. Conceptual design for a data-model platform for quantifying carbon removal to soils. comprehensive remotely sensed activity data on management practices (e.g., crop species, presence/absence of cover crops, tillage system, irrigation, vegetation productivity) provide the main data inputs to the model. Potential crowd-sourcing of farm management information (e.g., nutrient management) from the land users themselves can provide data that cannot be readily obtained from remote sensing. Such a system is scalable to estimate national-scale outcomes of CO2 removals over time as a function of land use policies (e.g. for national GHG inventory reporting) as well as to provide information on field- and local-scale dynamics that can inform policies based on carbon markets and/or sustainable product supply chains supported by agricultural industries. Many component parts of such a system (e.g. field experiment networks, remote sensing data) already exist and could be leveraged. Funding of $5M/yr for a 3 year development period followed by annual operating costs of similar magnitude would support system integration, data-model fusion, model development, decision support systems, visualization and communication. Demonstration Forest Demonstration Projects Conventional forestry practices have been implemented for decades worldwide, so demonstrations are readily available. One area needing new demonstrations is the frontier technology of increasing preservation of harvested wood products, particularly, how to improve the disposal and collection processes of wood products after their useful life. This research is linked to the basic research needed to improve landfill design for HWP preservation. The cost for this research done by USFS and EPA is $1.5 million/yr for 3 projects of duration 3 years. Associated with the basic research of improving landfill design to improve preservation of harvested wood, it is necessary to implement several demonstration projects representing different environmental conditions to facilitate scaling up this new technology. The cost for this research done by USFS and EPA is $1.5 million/yr for 3 projects of duration 3 years. Another promising activity needing demonstrations of potential impacts on carbon removal is forest restoration, which has been described for fire-prone areas as increasing carbon removal in the long term as a result of short-term reduction in carbon stocks of over-stocked stands; for areas converted from natural vegetation to unsustainably managed forests that prematurely lose ability to sequester carbon, and PREPUBLICATION COPY

90 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda for degraded forests that would respond to interventions to improve regeneration and stocking. The cost for this research done by USFS and partners is $1.5 million/yr for 3 projects of duration 3 years. Agricultural Systems Field Experiment Network Field experiments to rigorously evaluate region-specific best management practices for soil carbon sequestration (and net GHG reductions), in comparison to conventional practices, should be established in a coordinated network across the major agricultural regions soil types of the US, involving land-grant universities and other research institutions with the relevant expertise in ecosystem carbon and GHG dynamics. Sites selected should complement existing experimental sites in USDA Agricultural Research Service’s (ARS) GraceNET and Long-term Agricultural Experiment networks and relevant sites within NSF’s LTER and NEON programs. The new network should have a coordinated set of measurement protocols and methods, including whole system carbon balance (i.e., eddy covariance) and other GHG (e.g., N2O and CH4) flux measurements as well as precision soil carbon stock and stock change measurements. Data sharing and archiving to support modeling and meta-analysis should be priority and field sites should include a significant Extension, Outreach, Capacity building, technology transfer, and demonstration to provide information to producers, extension agents, crop consultants, agency personnel, etc. Sites should be designed for the field experiments and demonstrations to run for a minimum of 12 years. This research should be done by Land-grant universities and USDA at about 10-15 sites at a cost of $600K/yr per site for a total of $6-9 million/yr. Deployment Forest Carbon Project Deployment Although scientific uncertainty about the amount of carbon sequestered locally from improved forestry practices is low, uncertainty about far-field impacts caused by the long-term effects of climate change and the economic feedback is high. To reduce this uncertainty, the following research is needed: (1) Improve economic models that estimate leakage—reduced harvest in one location causing deforestation in another - and the impact of converting cropland to forest on food prices. Biological research is also needed to develop understanding of the impacts of significant impacts of carbon removal on biodiversity and development of safeguards to ensure a sustainable biosphere. This research and its cost is discussed in the research agenda for improvements to integrated assessment modeling in the BECCS chapter. (2) Improvements in life cycle assessment methods to evaluate substitution of wood products for other materials that could expand wood use, especially if these displace structural materials such as steel and cement, which require large carbon emissions to produce. Substituting wood for other material typically results in a net reduction of GHGs; but, tools to quantify the net reductions in GHGs are not readily available or validated. The research needs for LCA analysis and the costs are discussed in the BECCS chapter. (3) Social sciences research on how landowners respond to incentives and the utilization of Extension Service system at the land-grant universities to reach more landowners and practitioners who own and work with small parcels of forest as a means to improve the equity of participation in assistance programs that tend to favor larger landholders. Only a small percentage of forest landowners respond to incentive programs or price signals by changing land management practices. Very little research has been done or is ongoing regarding these topics. (USDA and NSF; $1.0 million/yr for 3 projects over 3 years). (4) Research on GHG and social impacts of reducing traditional uses of biomass for fuel, which involves households and small entities using wood biomass for heating and cooking. Globally, PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 91 increasing durable wood products or increasing commercial use of biofuel could reduce consumption of biomass for traditional fuel use, which would have rippling effects on types of energy supplied. (USDA and NSF; $1.0 million/yr for 3 years). (5) Social science research on responses to economic and behavioral incentives to reduce meat consumption and food waste. This research and its cost are included in the research agenda for improvements to integrated assessment modeling in Chapter 4 (BECCS). Scaling up Agricultural Carbon Sequestering Activities Current adoption of existing best management practices for soil C sequestration are low (e.g., cover crop are used on <5% of US cropland; Wade et al., 2015) and barriers to scale-up including economic and behavioral changes (e.g., value proposition, risk management, motivation), information needs and technology transfer are still poorly understood. Economic and behavioral research along with continuation and expansion of pilot emission reduction and carbon removal projects, such as through USDA/Natural Resource Conservation Service’s Conservation Innovation Grants, can provide needed empirical knowledge on which barriers are most limiting and how to design the most effective policies and education programs to promote agricultural carbon removal activities at scale. Regional Life-Cycle Assessments Any program to offer incentives to landowners for adopting carbon removal and storage should be preceded by a regional assessment of life-cycle emissions, costs, co-benefits and negative impacts, including estimates of leakage and the degree of permanence, which can be undertaken by federal land management agencies (e.g. USDA Forest Service and USDA Agricultural Research Service) and/or university researchers, with findings reviewed by an independent scientific board (Cost - refer to BECCS chapter, section on life-cycle assessment). Implementation of Terrestrial Carbon Sequestration Policies that overcome potential barriers for developing robust research and adoption of carbon removal and soil carbon sequestration for terrestrial NETs (cropland and forestland) may include several different mechanisms including: 1) government subsidies to land owner to adopt carbon sequestering practices, similar to existing conservation payments in the Farm Bill, for example, 2) carbon offset markets, in which land-based ‘carbon projects’ market emission reductions/carbon sequestration to major GHG emitters participating in either voluntary or mandatory emission reductions (i.e., cap-and-trade) and 3) demand-side programs in which carbon removal activities are undertaken in response to demand for land-based consumer products that have a low carbon footprint. All three approaches currently exist. For government-based incentives (e.g., payments for practice adoption), the US has a well-developed infrastructure of federal and state land management agencies and extension and outreach specialists to engage in implementation. For market-based systems, the voluntary greenhouse gas registries (e.g. VCS, ACR, CAR) and state agencies (e.g. CA Air Resources Board) have experience and governance structures that provide a starting point. New efforts on the part of companies to incentivize producers as part of sustainable or ‘low carbon’ supply chain initiatives are at present much more fragmented, with potential for double-counting and low transparency in the carbon removal activities actually undertaken. Hence institution building and development of governance structures to better engage the private sector in pursing demand-side carbon removal activities is a future need. PREPUBLICATION COPY

92 Negative Emissions Technologies and Reliable Sequestration: A Research Agenda Land use and conservation practices that promote or encourage SCS for agricultural land and forest should be coupled with incentives and levels of regulations that encourage farmers, land managers, and landowners to adopt such practices. The existing USDA conservation programs, such as CSP, EQIP, and CRP provide cost sharing for implementing such practices are good examples. However, there is a need to develop a set of incentives within the Farm Bill by linking conservation adoption and level of compliance to the level of improvement in soil carbon. The permanence of SCS and associated soil and ecosystem services co-benefits can be enhanced via financial incentives through government programs or as price premiums driven by industry seeking to develop more environmentally-friendly supply chains. Ultimately, the widespread adoption and long-term maintenance of working lands that place a high priority on SCS and reduced GHG emissions requires that they are as (or more) profitable to the farmer/rancher/forester as the conventional systems that they replace. Barriers to the Implementation of the Research Agenda For the recommended research agenda to achieve its goal and have lasting impact, certain barriers to adoption of carbon sequestering management practices need to be overcome. Actions to lower these barriers include capacity building of expertise and robust educational programs through Extension and Outreach and other government and private entities to promote the carbon removal and SCS concept, research findings and technology transfer to end users, economic incentives to compensate for added costs or potential yield loss while transitioning to new management practices and development of agronomists and specialists that can train and work with end users, to name few. However, a focus on the delivery and promotion of new technology alone is unlikely to overcome these barriers to adoption. These efforts should be coupled with social science research to better understand the drivers need to elicit the fundamental changes in land management practices necessary to achieve terrestrial carbon removal at scale. The research agenda should include an integrated platform for cropland, grassland, and forestry projects such as with the NIFA Coordinated Agricultural Projects (CAP), where research, extension and education are required components of funded proposals. In particular, the applied components of the research agenda (e.g., on-farm network and demonstration) would benefit from a multidisciplinary approach and collaboration between federal, state, and private agencies to achieve the goals of the research program. Development of teaching curricula for K-12 is another gap in developing greater awareness about the practices that enhance carbon removal and soil carbon sequestration, including training opportunities for teachers, students and public at large. Such efforts may include hands-on demonstration of proven technologies, such no-till, cover crop, residue management and agroforestry practices. Universities and USDA scientists can collaborate to develop and facilitate education and technology transfer through existing infrastructure and programs within land-grant universities and other public and private institutions. SUMMARY Terrestrial ecosystems can play a significant role in carbon dioxide removal and sequestration, through practices that increase the amount of organic carbon stored in living plants, dead plant parts and the soil. Please note that Biomass Energy with Carbon Capture and Storage (BECCS), which also involves carbon removal via plant photosynthesis, is covered in Chapter 4 of this report. Two suites of land management practices involving agriculture and forest lands are sufficiently mature, both scientifically and in practice, to materially increase carbon storage if widely deployed in the US and globally. Options exist for inexpensive remote monitoring and verification of the practices that lead to carbon capture and storage, coupled with data and model-based quantification of net carbon storage, to reduce the need for expensive on-site direct measurements of carbon on lands that adopt these practices. PREPUBLICATION COPY

Terrestrial Carbon Removal and Sequestration 93 Increasing the area of forested land through afforestation/reforestation creates a carbon sink ranging from 1.5 to 6.4 t C/ha/yr for a period of 50-100 years or more. Changing forest management practices such as extending timber harvest rotations and improving stocking and productivity by restoring degraded forests could store an additional 0.2 to 2.5 t C/ha/yr for several decades. Development of new technologies to increase the use of long-lived wood products as well as carbon burial schemes for harvested wood are options for increasing carbon removal potential from forests. For cropland and grassland, many long-term experiments document increases of soil carbon in the range of 0.2-0.5 t C/ha/yr (over time periods of 2-4 decades) from adopting conservation practices including more diverse crop rotations, use of cover crop, reduced tillage and improved grazing system. Other changes in land use and management such as agroforestry or re-establishment of wetland and perennial vegetation on marginal lands can achieve increases in SOC stocks of more than 1 t C/ha/yr. Additional frontier soil NETs include deep soil inversion and burial of carbon rich topsoil carbon burial, biochar amendment and development of enhanced root carbon input crop phenotypes. We estimate ‘practically achievable’ carbon removal amounts. In other words, implementation of existing practices at rates that do not require significant land use conversions that would jeopardize food security and biodiversity of intact native ecosystems in the US of 0.25 Gt CO2/yr from forest land and 0.25 Gt CO2/yr in agricultural soils, and corresponding global estimates of 2.5 and 3 Gt CO2/yr, from forests and agricultural soils, respectively. Much of this CO2 removal would be achievable at costs less than $50 per tonne. If frontier NETs prove practical and economical, rates of carbon removal for both forests and agricultural soils could roughly double in size. A broad-based mix of R&D needs, including both basic research, development of improved measurement and monitoring technology, regionally-representative demonstration projects and research on overcoming barriers to deployment and scale-up are outlined and cost estimates given. PREPUBLICATION COPY

Next: 4 Bioenergy with Carbon Capture and Sequestration »
Negative Emissions Technologies and Reliable Sequestration: A Research Agenda Get This Book
×
Buy Prepub | $94.00 Buy Paperback | $85.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

To achieve goals for climate and economic growth, “negative emissions technologies” (NETs) that remove and sequester carbon dioxide from the air will need to play a significant role in mitigating climate change. Unlike carbon capture and storage technologies that remove carbon dioxide emissions directly from large point sources such as coal power plants, NETs remove carbon dioxide directly from the atmosphere or enhance natural carbon sinks. Storing the carbon dioxide from NETs has the same impact on the atmosphere and climate as simultaneously preventing an equal amount of carbon dioxide from being emitted. Recent analyses found that deploying NETs may be less expensive and less disruptive than reducing some emissions, such as a substantial portion of agricultural and land-use emissions and some transportation emissions.

In 2015, the National Academies published Climate Intervention: Carbon Dioxide Removal and Reliable Sequestration, which described and initially assessed NETs and sequestration technologies. This report acknowledged the relative paucity of research on NETs and recommended development of a research agenda that covers all aspects of NETs from fundamental science to full-scale deployment. To address this need, Negative Emissions Technologies and Reliable Sequestration: A Research Agenda assesses the benefits, risks, and “sustainable scale potential” for NETs and sequestration. This report also defines the essential components of a research and development program, including its estimated costs and potential impact.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!