Skip Navigation

Environment Magazine September/October 2008


July-August 2012

ResizeResize Text: Original Large XLarge Untitled Document Subscribe

Forest Carbon Offsets: Challenges in Measuring, Monitoring and Verifying

Forest carbon is recognized as an important component in the carbon emissions and global warming equation. The Intergovernmental Panel on Climate Change estimated in 2007 that about 17 percent of greenhouse gas (GHG) emissions, the largest of which is carbon, are from changes in land use, primarily deforestation.1 Managing forests and their carbon is considered by the IPCC and others as a low-cost method of controlling GHG emissions. Studies suggest that the costs of targeted global reductions in GHG emissions can be reduced by up to 40 percent with the extensive use of forests to store carbon.2 In analyses of proposed congressional legislation to mitigate U.S. emissions, the U.S. Environmental Protection Agency, the U.S. Energy Information Administration, and the U.S. Congressional Budget Office conclude that forest carbon management would reduce the cost to the U.S. economy of meeting goals to reduce GHG emissions by half or in some policy scenarios, by more than half.3

A contributing factor to these cost savings is that no new technology needs to be developed to manage or grow trees. (In contrast, other solutions to the problem of climate change in many cases require costly technological development or rapid deployment of new infrastructure, such as renewable energy for electric power or hydrogen vehicles.) Biological growth naturally sequesters carbon as a plant captures carbon dioxide from the atmosphere, draws the carbon atoms into the plant cell, and releases oxygen back into the atmosphere. Trees are particularly efficient sequesterers of carbon. Perennial trees continue to hold carbon in their cells when dormant, and have long lives; trees function as carbon silos and accumulate increasing volumes of carbon for decades and, indeed, often centuries.

Within this context have been the newer initiatives encouraging efforts for Reducing Emissions from Deforestation and Forest Degradation (REDD). These efforts are largely oriented toward protecting existing natural forest areas for multiple purposes. Thus, although maintaining forest carbon is important, other equally important objectives include the protection of large native forest habitats for their biodiversity, environmental services, and unique historical cultural importance. Although REDD involves a larger set of objectives than simply carbon storage, to some extent the emphasis on REDD has been driven in part by frustrations, discussed in this article, with the ability to precisely monitor forest carbon and to provide payments accordingly. In the REDD context, the precise measurement of the forest stocks and their carbon content is of less importance since, generally, separate payments are not envisaged based on carbon content and content changes. Rather, to a large extent the REDD approach would involve payments to provide incentives to support or introduce institutional and enforcement changes that would protect, maintain and improve unique and already existing forests. Thus, any success in REDD programs, although not directly transferable into precise amounts of carbon that are measurable and verifiable, would almost surely have significant carbon sequestration as a component.

The Measurement and Monitoring Challenge

Although forest carbon sequestration may be easy in concept, it is less easy to implement. Concerns center on whether sequestration can be measured, monitored, reported, and verified if it is allowed as a means of offsetting GHG emissions from utilities, the transportation sector, and other polluters. Under some proposals, the exchange from the polluters to forest landowners would be in the form of financial incentives to compensate landholders for alternative uses of the forest, with payments related to performance as determined by measurement and monitoring. For instance, in the United States, proposals for the offset of GHG emissions by using forest carbon sequestration have been part of so-called cap-and-trade strategies, which allow forest carbon sequestration to offset emissions elsewhere in the economy.4 Cap and trade has not advanced in the United States, although forest offsets are under consideration in some states and in Europe.5 Other approaches are not based on cap and trade but provide financial compensation in the form of development aid. For instance, in the proposed U.S. federal budget for fiscal year 2013, the U.S. Department of State includes $469.5 million in a Global Climate Change fund, which includes the goal of combatting deforestation in developing countries.6 This funding would be a step toward the United States fulfilling one of the promises made at the Fifteenth Conference of the Parties (COP) to the United Nations Framework Convention on Climate Change (UNFCCC), in December, 2009 in Copenhagen.7 (The Copenhagen Accord is not legally binding, but because the agreement was made by heads of state, the accord is thought to carry some weight.8) As another example of the potentially significant role of forest sequestration in commitments made by countries who are party to the agreement, China proposed to increase forest coverage by 40 million hectares and forest stock volume by 1.3 billion cubic meters.9 Other parties also promised to use forest sequestration as part of their commitments.10 However, admittedly, some of the fervor for using forest for carbon capture has diminished.

Although promising to manage national GHG emissions by capturing carbon with trees and even providing compensation to do so sounds easy, there are many problems. Tree growing may not require new technology, but the same is not true for accurate measurement and monitoring of forest carbon. Indeed, today, society is still having difficulty obtaining accurate measurements of the forest areas of many places around the globe, and these should be much easier to measure than forest carbon.11 To be sure, some countries provide state-of-the-art inventories for portions of their forested land, such as that supplied by the Amazonia Deforestation Project (PRODES) in Brazil, in conjunction with the Brazilian national space agency, the National Institute for Space Resources (INPE). But few efforts in other developing countries match this effort. As a result of this gap, there is no global map of forested areas at a reasonable level of accurate detail. Documented discrepancies in the various estimates of forest areas for certain countries and regions attest to the large measurement differences and inconsistencies.12 Although part of the problem is differences in countries' definitions of forest, the overall problem is far more complex and reflects both organizational and technical limitations.13

Indeed, the variation in current estimates attests to the variability in measurements, with recent estimates of net forest carbon contribution to total anthropogenic emissions of GHG as low as 6 percent, compared to the IPCC estimate of 17 percent.14 Nor are ground data widely available in many developing countries to measure, report, and allow verification of forest management practices to qualify for financial incentives. With this concern, the funds in the proposed US FY13 budget for climate change financing in developing countries are also to be allocated to improve capacity “to effectively monitor, report, and verify emissions and removals from land use activities.”15

Technology and Its Deployment

Technologies necessary for global and accurate measurement are available but have not been uniformly deployed around the world.16 Traditionally, foresters have measured forest areas and volumes with inventorying processes that involve sampling plots of forested land and the volume of trees. These approaches continue today as the foundation for measurement.

Remote sensing using instruments on airplanes and earth satellites is increasingly able to cover larger areas and to provide consistent measures of ground conditions, including the area of forest cover, deforestation rates, and forest land conversion to other uses, such as agriculture, across large areas. Ground measures provide field truthing of the remote sensing, thereby enhancing its accuracy, but alone cannot provide the global coverage and consistently calibrated data required for more accurate estimates of the quantitative the details of the forest worldwide and of the forest's role in mitigating GHG emissions.

Although remote sensing is playing a greater role in forest measuring and monitoring in the U.S. inventorying process and in the global forest estimates provided by the by the UN Food and Agricultural Organization, remote sensing by itself still has important limitations particularly with regard to measuring forest carbon. To measure carbon accurately, a dry biomass measurement (specific gravity) is necessary. Tree biomass is related to the tree volume, species, site, and density. Even for the same species, the volume–density relationship varies by site conditions. Furthermore, most of the world's forests consist of a mix of species. Tropical forests, particularly, involve large varieties of species even on small areas. Even with precise forest area and volume measurements, estimates of forest carbon can contain significant errors. And note that forest volumes are not yet measured with great accuracy.17

New remote-sensing approaches appear promising for estimating forest volume. Optical imaging devices on satellites can provide highly accurate estimates of global forest area but not forest volume. The oldest land remote-sensing activity, the U.S. Landsat program, has provided imagery of land use since 1972 and is the workhorse in supplying data for efforts around the world to monitor forest area. The imaging instruments on the existing two Landsat satellites are not functioning properly and the program at present is underfunded, however.18

Much more sophisticated technology is required for measuring forest carbon. Radar technology promises increasingly accurate estimates of volume for many forests by aiming radiation toward Earth and reading the reflected radiation to observe changes in elevation, including forest height. Radar instruments are also able to penetrate clouds and thus have an advantage over optical instruments, particularly in tropical areas. LIDAR (light detection and ranging) technology directs lasers toward the ground and measures the timing of the returned light, providing highly accurate estimates of forests in three dimensions.

With the increasing interest in measuring forest carbon and in anticipation of the possibility of a market involving payments for forest carbon sequestration, numerous private-sector companies have been developing airborne LIDAR services. At present, these companies are active in serving landowners who participate in the voluntary carbon exchanges in the United States. Although LIDAR data show impressive detail of some forests, LIDAR has its limits with very dense forests. Aircraft have limited ranges, however, and the width of the swath captured by the LIDAR instrument is relatively small; thus, a large number of passes is required to cover a large area. A recent effort has developed a system for assessing tropical forest carbon using Landsat to map a large area and then using airborne LIDAR to map a sample of the area.19

Leakage and the Need for Global Measurement and Monitoring

The Brazilian PRODES mapping of part of the Amazon and the recent project combining Landsat and LIDAR to map tropical forests are promising, but neither initiative has as its goal to extend mapping coverage throughout the forests of the world. Here, forest carbon sequestration as part of a global GHG management effort suffers from an even more basic problem than the measure of the carbon on a site. Deforestation is fungible—it can shift from one site to another. Protection of one forest might simply deflect deforestation activities to another site. This occurrence is defined as leakage, because the amount of forest protected or carbon sequestered is overstated because some of the gains are offset (leaked) elsewhere. Leakage can occur in developing countries, as, for example, attempts to protect some forest from squatters, logging, or conversion to agriculture simply result in the loggers and squatters moving to another forest to maintain their livelihoods. Some critics of efforts to reduce GHG emissions from deforestation and degradation (REDD) point out that historical experience with projects to conserve forests, including projects involving payments for the conservation, prevented deforestation on less than 1 percent of the enrolled hectares.20 Another study estimated leakage in Bolivia at up to 50 percent.21 Part of the problem rests with weak regulatory institutions and strong pressures to convert forested land to other purposes.22 Additional leakage in commercial forests could be the result of market forces. If harvests of industrial wood for markets are decreased in one area, the market scarcity of wood should cause the market price to rise, thereby providing incentives to increase harvests elsewhere.

How important is leakage in a developed country like the United States? Leakage in U.S. forests has been estimated as between 10 and 90 percent.23 A leakage rate of 50 percent means that for every ton of forest carbon that is saved through a newly protected forest area, half a ton of carbon is released elsewhere. The problem and empirical presence of substantial leakage suggest that even in the United States one cannot simply establish a protected forest area, reduce logging in one forest, or plant new trees and treat the carbon captured by those activities as net sequestration. Some form of countrywide “wall-to-wall” monitoring is necessary to accurately capture the magnitude of these country internal leakages. Note that the implication of leakage is that one must measure and monitor the forest situation throughout the entire country, not simply the forests designated for offset credit or as part of a country's international voluntary commitment under the UNFCCC.

Furthermore, leakages need not be limited to within a country. This is demonstrated by the global response to the reductions in harvests from the U.S. National Forest System that took place in the early 1990s. As timber harvests declined in the U.S. national forests, total U.S. harvests remained surprisingly stable as harvests from the private forests, particularly in the U.S. South, increased markedly, thus demonstrating internal leakage. Additionally, during that same period overall U.S. industrial wood consumption increased as imports rose from eastern Canada and the Nordic countries, offsetting the declines in harvests in the U.S. national forests.24

Other problems confound the process. Forest changes mean that carbon monitoring would need to be persistent and continuing. Systems need to be devised that would allow offset credit to be given during the growth of the forest, but withdrawn should the forest be reduced or destroyed, whether by design or inadvertently. This issue might be addressed if forest carbon sequestration is compensated by payments, which might be in the form of a rental for the service of maintaining carbon in the tree for a finite period of time.25 In concept, forest carbon offset systems can be devised to adjust credits and their values to the carbon sequestered. However, monitoring today has serious limitations in accuracy, as well as being potentially high cost. Finally, fraud has already emerged as a problem both in the developing world and elsewhere and could become more serious if the values grew substantially.26

Solutions and Technological Promise, but Who Pays?

The preceding considerations suggest that a countrywide carbon forest perspective is appropriate, one that will account for internal leakages. If the leakage is pushed outside the country, it could become the responsibility of the new country where it occurs. The lack of measurement precision for carbon at present probably precludes direct offset credits and/or payments as might be achieved in the carbon exchange markets. 27 However, even if the forest carbon cannot be monitored adequately, incentive payments could be related to some measure of controlling deforestation and establishing new forested areas.28 In this context rewards might go from the international community to countries that have demonstrated positive results, but without an attempt to have a precise and detailed accounting of forest carbon changes. Thus, although an accurate carbon measurement may be beyond our technical capacity today, proxy metrics, such as forest area, may be a sufficient substitute for many purposes until more global accuracy and precision is feasible. Note also that monitoring leakage is important not only for verifying whether commitments have been upheld, but for measuring forest carbon sequestration as part of the overall monitoring of the health of the global climate. In other words, accurate measures and monitoring are necessary both for public policy and for climate science.

On the technology front, using satellites equipped with LIDAR could provide global coverage and, depending on the complexity of the system (such as how many satellites), could provide observations possibly at a lower cost per kilometer than the present cost of airborne LIDAR.29 Given the regular orbit of the satellite around Earth, the coverage could be obtained globally, and more frequently than airborne measures, thus providing up-to-date changes in forests due to logging, fire, storm, or pest disturbance. Whether carbon markets would have sufficient revenues to underwrite the cost of a LIDAR system is far from clear, although economic analysis of congressional proposals based on cap-and-trade approaches estimate tens of billions of dollars annually in revenue. Some of the revenue could be allotted to a monitoring and observing system. In the absence of markets and revenue, however, it is far from clear who would be willing to pay for a global LIDAR monitoring system.

Creative design of incentive arrangements to encourage sequestration may be possible, and we know how to grow trees. We also have the technical means to observe, measure, and monitor some forest areas with a high degree of accuracy and precision. Thus, carbon offsets are feasible in many areas. Nevertheless, the problem of leakage remains daunting, because to truly address it requires a global monitoring capacity. But the financing and deployment of the technology for truly global forest observations—necessary for managing a truly global climate problem—is at present limited. Fortunately, even where forest carbon cannot be monitored with precision, incentive payments can be related to a set of performance standards interconnected to the control deforestation and the establishment of new forested areas.

1. IPCC 2007, in B. Metz, O. Davidson, P. Bosch, R. Dave, and L. Meyer, eds., Climate Change 2007: Mitigation (Cambridge: Cambridge University Press, 2007).

2. See M. B. Tavoni, B. Sohngen, and V. Bosetti, Energy Policy, 35(11):5346-53.

3. U.S. Energy Information Administration, Energy Market and Economic Impacts of HR 2454, Tthe American Clean Energy and Security Act of 2009 (4 August 2009), at; U.S. Environmental Protection Agency, EPA Analysis of The American Clean Energy and Security Act of 2009 (23 June 2009), at; EPA Analysis of The American Power Act in the 111th Congress (14 June 2010), at; Congressional Budget Office, The Use of Offsets to Reduce Greenhouse Gases Issue Brief 3, August (Washington, DC: Congressional Budget Office, 2009).

4. Cap and trade were provisions in the The American Clean Energy and Security Act of 2009, passed by the U.S. House of Representatives, and The American Power Act of 2010, both drafted during the 111th Congress. (Also see note 3.)

5. The summary of the U.S. Department of State's FY2013 budget is at (accessed February 20, 2012).

6. See U.S. Department of State, U.S. Climate Action Report: Executive Summary of UNFCCC Statement, Copenhagen (2009), at (accessed February 20, 2009).

7. Stavins and Stowe assess in detail the Copenhagen meetings. See “What Hath Copenhagen Wrought? A Preliminary Assessment,” Environment, May/June: (2010).

8. See letter from the Director General, Department of Climate Change, National Development and Reform Commission of China, to the Executive Secretary of the UNFCCC, 28 January 2010,

9. See commitments made by Annex I countries in Appendix I of the accord at and those made by Annex II countries in Appendix II,

10. P. Waggoner, “Forest Inventories: Discrepancies and Uncertainties,” RFF Discussion Paper 09-29 (2009),

11. See Waggoner, note 12; see also E. Matthews and A. Grainger, Evaluation of FAO's Global Forest Resources Assessment from the User Perspective (Rome, Italy: FAO Corporate Document Repository, 2002), (accessed February 24, 2009).

12. Comprehensive discussion of the problems in developing countries, where technical capacity is often limited, is in GOFC-GOLD (Global Observation of Forest and Land Cover Dynamics), Reducing Greenhouse Gas Emissions from Deforestation and Degradation in Developing Countries: A Sourcebook of Methods and Procedures for Monitoring, Measuring and Reporting (Edmonton, Canada: GOFC-GOLD Project Office, Natural Resources Canada, 2007); see also M. Macauley, D. Morris, R. Sedjo, K. Farley, and B. Sohngen, “Forest Measurement and Monitoring: Technical Capacity and How Good is Good Enough?,” RFF Report, December (2009),

13. G.R. Van der Werf et al., “CO2 Emissions From Forest Loss,” Nature Geoscience 2 (2009): 1–2.

14. See the U.S. Department of State's FY2013 budget at (accessed February 20, 2012).

15. Fagan and DeFries provide a comprehensive and recent survey. See M. Fagan and R. DeFries, “Measurement and Monitoring of the World's Forests: A Review and Summary of Remote Sensing Technical Capability, 2009-2015” RFF Report (2009).

16. Fagan and DeFries (note 17) and Waggoner (note 12) provide examples.

17. See C, E. Behrens, “Landsat and the Data Continuity Mission,” CRS Report for Congress (R40594), 22 May (2009). Because the Landsat program serves many purposes, including providing imagery to monitor urbanization, development, and other uses of land, a U.S. federal task force in 2007 proposed the establishment of a permanent land imaging program. See U.S. Office of Science and Technology Policy, A Plan for a US National Land imaging Program (Washington, DC: U.S. Office of Science and Technology Policy, 2007).

18. G. P. Asner, “Tropical Forest Carbon Assessment: Integrating Satellite and Airborne Mapping Approaches,” Environmental Research Letters 4 (2009), 0:34009.

19. A. Blackman, “Will REDD Really be Cheap?,” Resources (Washington, DC: Resources for the Future, 2010); J. Robalino, A. Pfaff, G. A. Sanchez-Azofeifa, F. Alpizar, C. Leon, and C. M. Rodriguez, 2008, “Deforestation Impacts of Environmental Services Payments: Costa Rica's PSA Program 2000–2005,” Environment for Development Discussion Paper 08-24 (Washington, DC: Resources for the Future, 2008); K. Andam, P. Ferraro, A. Pfaff, J. Robalino, and A. Sanchez, “Measuring the Effectiveness of Protected Area Networks in Reducing Deforestation,” Proceedings of the National Academy of Sciences 105, no. 42 (2008), 16089–16094.

20. B. Sohngen and S. Brown, “Measuring Leakage from Carbon Projects in Open Economies,” Canadian Journal of Forest Research 34 (2004): 8929–839.

21. K. Chomitz, “At Loggerheads? Agricultural Expansion, Poverty Reduction, and Environment in the Tropics” (Washington, DC: The World Bank, 2007).

22. B. C. Murray, B. A. McCarl, and H. Lee, “Estimating Leakage from Forest Carbon Sequestration Programs,” Land Economics 80, no. 1 (2004): 109–124.

23. R. A. Sedjo, A. C. Wiseman, D. J. Brooks, and K. S. Lyon, “Global Forest Products Trade: The Consequences of Domestic Forest Land use Policy,” RFF Discussion Paper 94-13 (1994).

24. R. A. Sedjo and G. Marland, “Inter-Trading Permanent Emissions Credits and Rented Temporary Carbon Emissions Offsets: Some Issues and Alternatives.” Climate Policy 2, no. 3 (2003): 435–444; S. Alexander, P. Pfaff, S. Kerr, R. Flint Hughes, S. Liu, G. Arturo Sanchez-Azofeifa, D. Schimel, J. Tosi, and V. Watson, “The Kyoto Protocol and Payments for Tropical Forest: An Interdisciplinary Method for Estimating Carbon-Offset Supply and Increasing the Feasibility of a Carbon Market Under the CDM,” Ecological Economics 35 (2000): 203–221.

25. “UN's Forest Protection Scheme at Risk From Organised Crime, Experts Warn,”

26. P. Waggoner, “Could Uncertain Forest Inventories Hinter Carbon Markets?,” RFF Feature, September (2009),; U.S. GAO, Carbon Offsets: The US Voluntary Market is growing but Quality Assurance Poses Challenges for Market Participants, GAO-09-1048 (Washington, DC: U.S. GAO, 2008).

27. M. Grieg-Gran, “The Cost of Avoiding Deforestation: Update of the Report Prepared for the Stern Review of the Economics of Climate Change, UK International Institute for Environment and Development” (Cambridge: 2008).

28. At present, the availability of satellite LIDAR is quite limited. The first map of global forest heights using one uniform method was completed in July 2010 by scientists using data from three satellites operated by the US National Aeronautics and Space Administration (see, accessed February 20, 2012). The scientists based the map on data collected during seven years. One of the satellites carried a LIDAR sensor, but because the original intent of the sensor was to observe ice, cloud, and land elevation, it only directly measured about 2.4 percent of the Earth's forested surfaces. A future NASA satellite being designed to carry a LIDAR instrument to observe ecosystems could provide better data. Ground truthing will continue to play a major role.

Roger Sedjo is a Senior Fellow and Director at the Center for Forest Economics and Policy.

Molly Macauley is a Senior Fellow and Vice-President for Research at Resources for the Future in Washington, DC.

In this Issue

On this Topic

Taylor & Francis

Privacy Policy

© 2018 Taylor & Francis Group · 530 Walnut Street, Suite 850, Philadelphia, PA · 19106