Uncertainty [MGD Sections]

UNFCCC decisions and requirements
IPCC good practice guidance
Relationship to UNFCCC
GHGI coverage, approaches, methods and tiers
Design decisions relevant to national forest monitoring systems
Land cover, land use and stratification
Forest reference emission levels and forest reference levels
Quality assurance and quality control
Guiding principles – Requirements and design decisions
Estimation methods for REDD+ activities
Integration frameworks for estimating emission and removals
Selecting an integration framework
Activity data x emission/removal factor tools
Fully integrated tools
Practical considerations in choosing an integration tool
Guiding principles – Methods and approaches
Remote sensing observations
Coarse resolution optical data
Medium resolution optical data
High resolution optical data
L-band Synthetic aperture radar
C-band and X-band SAR
LIDAR
Global forest cover change datasets
Ground-based observations
National forest inventories
Auxiliary data
Guiding principles – Remote sensing and ground-based observations
Activity data
Methods for estimating activity data
Maps of forest/non-forest, land use, or forest stratification
Detecting areas of change
Additional map products from remote sensing
Estimating uncertainty of area and change in area
Estimating total emissions/removals and its uncertainty
REDD+ requirements and procedures
Reporting forest reference emission levels and forest reference levels
Technical assessment of forest reference emission levels and forest reference levels
Reporting results of REDD+ activities
Technical analysis of the REDD+ annex to the BUR
Additional advice on REDD+ reporting and verification
Guiding Principles – Reporting and verification of emissions and removals
Financial considerations
Country examples – Tier 3 integration
Use of global forest change map data
Relative efficiencies
Developing and using allometric models to estimate biomass

Record Keeping [MGD Sections]

Integration + Estimation [MGD Sections]

Ground Based Observations [MGD Sections]

2.2.2   GHGI coverage, approaches, methods and tiers Previous topic Parent topic Child topic Next topic

For reasons of transparency and consistency (as requested by COP decisions 12/CP.17 Opens in new window, 11/CP.19 Opens in new window and 13/CP.19 Opens in new window), countries should where possible use the same approaches, methods and data for reporting forestry emissions in national greenhouse gas inventories (GHGI) and REDD+ reports. However there reasons why GHGI and REDD+ may not be straightforward to compare; e.g. GHGI estimates contain national estimates of emissions and removals from land use and land use change, whilst REDD+ estimates are for activities which may be sub-national as an intermediate step, and in some cases may have different data and methods because the REDD+ and GHGI estimates may not yet be fully aligned. It is important to ensure consistency where possible, document any differences, and to understand and communicate implications where differences may occur. The relationship between REDD+ estimates and GHGIs in the context of FREL and FRLs is discussed further in Section 2.3.3.
GPG2003 Opens in new window provides methodologies to estimate changes in five carbon pools (above-ground biomass, below-ground biomass, dead wood, litter, and soil organic matter(1)) and non-CO2 GHG emissions for six categories of land use (Forest Land, Cropland, Grassland, Wetland, Settlements and Other Land), and for changes between land uses. Table 2 and Table 3 show how IPCC defines these pools and land categories.

Table 2: Definitions for carbon pools

Pool
Description
Biomass
Above-ground biomass
All living biomass (expressed in tonnes dry weight) above the soil including stem, stump, branches, bark, seeds, and foliage.
Note: In cases where forest understorey is a relatively small component of the aboveground biomass carbon pool, it is acceptable for the methodologies and associated data used in some tiers to exclude it, provided the tiers are used in a consistent manner throughout the inventory time series.
Below-ground biomass
All living biomass of live roots. Fine roots of less than (suggested) 2mm diameter are often excluded because these often cannot be distinguished empirically from soil organic matter or litter.
Dead organic matter
Dead wood
Includes all non-living woody biomass not contained in the litter, either standing, lying on the ground, or in the soil. Dead wood includes wood lying on the surface, dead roots, and stumps larger than or equal to 10 cm in diameter or any other diameter used by the country.
Litter
Includes all non-living biomass with a diameter less than a minimum diameter chosen by the country (for example 10 cm), lying dead, in various states of decomposition above the mineral or organic soil. This includes the litter, fumic, and humic layers. Live fine roots (of less than the suggested diameter limit for below-ground biomass) are included in litter where they cannot be distinguished from it empirically.
Soils
Soil organic matter
Includes organic carbon in mineral and organic soils (including peat) to a specified depth chosen by the country and applied consistently through the time series. Live fine roots (of less than the suggested diameter limit for below-ground biomass) are included with soil organic matter where they cannot be distinguished from it empirically.
Notes: National circumstances may necessitate slight modifications to the pool definitions. Where modified definitions are used, it is good practice to report upon them clearly, to ensure that modified definitions are used consistently over time, and to demonstrate that pools are neither omitted nor double counted.

Table 3: IPCC top-level land categories for greenhouse gas (GHG) inventory reporting

IPCC
Land Category(3)
Description
Forest Land
This category includes all land with woody vegetation consistent with thresholds used to define Forest Land in the national GHG inventory, sub-divided into managed and unmanaged, and also by ecosystem type as specified in the IPCC Guidelines(4). It also includes systems with vegetation that currently fall below, but are expected to exceed, the threshold of the Forest Land category.
Cropland
This category includes arable and tillage land, and agro-forestry systems where vegetation falls below the thresholds used for the Forest Land category, consistent with the selection of national definitions.
Grassland
This category includes rangelands and pasture land that is not considered as Cropland. It also includes systems with vegetation that fall below the threshold used in the Forest Land category and which are not expected to exceed, without human intervention, the threshold used in the Forest Land category. The category also includes all Grassland from wild lands to recreational areas as well as agricultural and silvopastural systems, subdivided into managed and unmanaged consistent with national definitions.
Wetlands
This category includes land that is covered or saturated by water for all or part of the year (e.g., peatland) and that does not fall into the Forest Land, Cropland, Grassland or Settlements categories. The category can be subdivided into managed and unmanaged according to national definitions. It includes reservoirs as a managed sub-division and natural rivers and lakes as unmanaged sub-divisions.
Settlements
This category includes all developed land, including transportation infrastructure and human settlements of any size, unless they are already included under other categories. This should be consistent with the selection of national definitions.
Other land
This category includes bare soil, rock, ice, and all unmanaged land areas that do not fall into any of the other five categories. It allows the total of identified land areas to match the national area, where data are available.
IPCC provides methods to estimate emissions for land remaining in a given category, and for land converted from one category to another. Table 4 shows the possible conversions and the codes used conventionally for them. Land is conventionally assumed to remain in a land converted category for 20 years after the transition that took it to a new land use. This assumption can be relaxed at Tier 3 (Box 6)(5).

Table 4: Land use conversion and definitions according to IPCC good practice

Land Remaining Categories
Land Converted Categories
FF = Forest Land Remaining Forest Land
LF = Land Converted to Forest Land
CC = Cropland Remaining Cropland
LC = Land Converted to Cropland
GG = Grassland Remaining Grassland
LG = Land Converted to Grassland
WW = Wetlands Remaining Wetlands
LW = Land Converted to Wetlands
SS = Settlements Remaining Settlements
LS = Land Converted to Settlements
OO = Other Land Remaining Other Land
LO = Land Converted to Other Land
Deforestation is estimated as the sum of emissions and removals associated with conversions from forest to other land uses. Removals are possible because of growth of biomass in the post-deforestation land use (i.e. cropland, grassland) following conversion. Neither GPG2003 Opens in new window nor the 2006GL Opens in new window identifies forest degradation, conservation of forest carbon stocks, and sustainable management of forests by name, but these can be estimated as the effect on emissions and removals of human interventions on land continuing to be used as forests(6). Enhancement of forest carbon stocks may occur within existing forests and also include the effect of conversion from other land uses to forest. How to make these estimates, including cross referencing the methods described by IPCC is described in Chapter 3, Section 3.1. Box 5 summarizes how REDD+ activities and IPCC land use categories relate to each other.
IPCC describes three approaches to providing activity data involving land area(7). Approach 1 is not spatially explicit(8) and simply uses net areas associated with land use. Approach 2 provides the matrix of changes between land uses. Approach 3 is geographically explicit and allows tracking of land use changes over time and is suited to situations where land use is dynamic, with multiple changes in cover or use over time. Remote sensing data are likely to be used to greatest advantage with Approaches 2 and 3.
Spatial stratification based on type or extent of human activities or type of forest should improve the quality of the results whatever the tier. For example, forests may be subdivided by using data on ecosystem type, climate, elevation, disturbance history, and/or management practice. More information on stratification is provided in Section 2.3.2. IPCC methods are applied at the level of the different carbon pools within the strata and the emissions and removals summed.
IPCC describes methods at three levels of detail, called tiers. Box 6 summarizes the definition of Tiers, based on the description in the GPG2003. Tier 1 is also called the default method, and the IPCC guidelines aim to provide the information needed for any country to implement Tier 1, including emission and removal factors and guidance on how to acquire activity data. Tier 2 usually uses the same mathematical structure as Tier 1 with countries providing data specific to their national circumstances. This would typically require field work to estimate the values required if they do not exist. Tier 3 methods are generally more complex, normally involving modelling and higher resolution land use and land-use change data. IPCC expects that higher Tier (meaning Tier 2 or Tier 3) methods will be applied for key categories (Section 2.2.3), unless the data collection to do this would significantly jeopardize resources required for other key categories(9).
Experience of developing national GHG emissions estimates suggests that even a system that is Tier 3 overall will use Tier 1 or Tier 2 emissions/removals factors for some components. For example, all operating Tier 3 systems calculate carbon dioxide and methane emissions from fire using models, but typically use emissions/removals factors to estimate the nitrous oxide emissions associated with wildfires and slash burning (Kurz et al., 2009). Some Tier 3 systems use Tier 1 or 2 methods for on-going emissions of soil carbon following deforestation. For national GHG reporting, a combination of tiers, most often Tier 1 and Tier 2 may be used, and any combination of Tiers and Approaches, as described above. For REDD+, Approach 3 could provide the spatially explicit information needed to track activities and drivers, and to support estimation of GHG emissions or removals. Increased availability of remotely sensed data makes this more practicable. This may have consequences for national GHGI development, so consistency between the two can be established.
The selection of the appropriate Tier and Approach to use for GHG estimation and for other purposes depends on country circumstances including system development and operational budgets, infrastructure and capacity as well as intended use of outputs from the system. A summary of the key factors to consider is provided in the form of a decision-tree in Figure 2. Cost-effectiveness is discussed in Chapter 4 and Appendix A.
Figure 2: Key factors relevant to system design, tier and approach selection in GHG estimation
Figure2.svg
Considerations at the decision points in the tree are as follows:
Decision Point 1: Is the land sector a key emissions source for your country?
Whether the land sector is a key category will depend on the proportion of emissions that the land sector emits (see key category analysis, Section 2.2.3). It is possible to test if the land sector is going to be a key sector using Tier 1 methods, in the absence of national data (see GPG2003 Opens in new window).
Decision Point 2: Will any possible reductions be used for mitigation targets or results based payments?
A more advanced system than Tier 1 is likely to be required to support mitigation targets for results based payments.
Decision Point 3:Do you need a more advanced system for other reasons?
There are reasons other than UNFCCC reporting to develop a MRV system (e.g. monitoring and reporting on forest resource assessment or more broadly national environmental performance). If the land sector is not a key category in the national greenhouse gas inventory and you do not need an MRV system for other reporting purposes then apply Tier 1.
Decision Point 4: Do you want the system to report national estimates and support projects?
Sub-national and project level reporting should demonstrate consistency with national estimates and document how data acquisitions and calculations are conducted in support of each other.
Decision Point 5: Do you want the system to be broader than emissions?
Some examples of broader requirements (other than those specified in Note 3) include: consideration of including wider land sector activities; environmental and social safeguards; land use planning etc.
Decision Point 6: Do you want to do scenario analysis?
Scenario analysis can be useful in understanding and predicting impacts of various mitigation actions on future results based payments.
IPCC distinguishes between two methods for estimating emissions and removals of CO2 associated with annual rates of change in all carbon pools(10). These are the gain-loss method (which estimates annual emissions and/or removals separately and directly), and the stock change(11) method (which estimates net annual emissions or removals from the difference in total carbon stocks at two points in time divided by the number of intervening years). Considerations for selecting and applying these methods are discussed below. The carbon stock estimates for the stock change method are commonly estimated from repeated field measurements of forest variables as part of an NFI (Chapter 4, Section 4.2.1) or equivalent survey data. Remote-sensing data may be useful in improving the efficiency of sampling in an NFI by assisting in stratification(12) and by providing auxiliary data during estimation.
IPCC notes that the stock change method provides good results where there are relatively large increases or decreases in estimated biomass, or where there are statistically rigorous NFIs (13). Since countries may not possess an NFI(14), and NFIs by themselves do not track or map REDD+ activities, the advice in the MGD focuses more on the gain-loss method. The gain-loss method requires ground data which can come from an NFI as discussed in Chapter 4, Section 4.2.1.
The gain-loss method estimates annual net emissions or removals of CO2 as the sum of gains and losses in carbon pools occurring on areas of land subject to human activities. This may be achieved by the use of emissions/removals factors and activity data or by the use of more sophisticated representative models and integrated systems as discussed briefly at the end of this section and in more detail in Chapter 3. Most of what follows relates to use of emissions/removals factors and activity whereby changes in the carbon pools are estimated as the product of an area of land and an emission or removal factor that describes the rate of gain or loss in each carbon pool per unit of land area.
To estimate emissions and removals using this method, countries need activity data, i.e. information about the extent of REDD+ activities(15). Remote-sensing is likely to provide the main source.
Activity data combined with emission and removal factors and other parameters, usually expressed per unit area, are used to estimate emissions or removals. Activity data generally correspond to strata based on forest type and condition, management practice or disturbance history. Stratification may require auxiliary data and is useful in increasing accuracy and in linking to appropriate emission and removal factors.
For conversions from forest to other land uses which are summed to estimate total deforestation, the gain-loss method multiplies areas of land-use change by the difference in carbon stocks per unit area between forest and the new land use. For Forest Land remaining Forest Land, the gain-loss method estimates the annual change in above-ground biomass carbon as the difference between the annual increment in carbon stocks due to growth and the annual decrease in stocks due to losses from processes such as commercial harvest, fuel wood removal(16), and other disturbances such as fire and pest infestation (Chapter 3.2 in GPG2003 Opens in new window; Cienciala et al., 2008). Collation of data on gains and losses may be useful in management and policy scenario analysis. The balance of gains and losses (i.e. net change) can also be estimated from sample plots representative of strata subject to the processes involved.
The choice between using a gain-loss or stock change method at the appropriate Tier(17) will depend on expert judgment, taking the status of national inventory systems and forest characteristics into account. Figure 3 summarizes these choices recognising that, even if not used directly for estimating emissions and removals associated with REDD+ activities, an NFI, where it exists, can provide potentially useful data for use with the gain-loss method, so that the approaches are in a sense complementary. This is discussed further in Chapter 4, Section 4.2.2.
Figure 3: Method selection for estimating CO2 emissions and removals based on available data
Figure3.svg
Considerations at the decision points in the tree are as follows:
Decision Point 1: Does you country have a National Forest Inventory (NFI)?
An NFI is a periodically updated sample-based system covering all forests within a country to provide information on the state of a country’s forest resources. Where NFI data have been collected on a consistent basis for more than one point in time this data can be used to directly estimate carbon stock change between two points in time and can be used to estimate emission and removal factors.
Decision Point 2: Are you planning to track or map REDD+ activities or drivers using spatially explicit data?
Mapping using spatially explicit data is useful for understanding the relationship between REDD+ activities and drivers, e.g. for policy analysis.
Decision Point 3: Does the NFI data capture REDD+ activities and carbon pools at the required precision?
Existing NFI sampling designs are unlikely to be optimized to estimate REDD+ activities such as deforestation or forest degradation, or carbon pools within the areas subject to land use change, leading to increases in uncertainties in estimating emissions and removals. Key category analysis will assist in assessing if the NFI data is capturing REDD+ activities and carbon pools at the required precision.
Decision Point 4: Is it likely to be cost effective to augment sampling?
Adding to the sampling may be required where required precision is not achieved. Although an NFI for an entire country might be desirable, it is often logistically complex and expensive in large countries, especially those with large areas of non-commercial forest. Increasing the sample size could be regarded as cost effective if it saved resources relative to alternative approaches, or did not involve disproportionate additional expenditure given the benefit anticipated.
Decision Point 5: Do you want to establish an NFI for other forest resource management purposes?
The broader national benefits to be realised from an NFI should be considered in the assessment of cost effectiveness and other broader decision making.
Decision Point 6: Is joint sampling to identify REDD+ activities and carbon pool data likely to be cost effective?
A step could be regarded as cost effective if it saved resources relative to alternative approaches, or did not involve disproportionate additional expenditure given the benefit anticipated.
The gain-loss method can be implemented using default emission/removal factor data from IPCC guidelines and guidance (Tier 1), or nationally relevant data from sampling, forest inventories or research sites (Tiers 2 or 3). Emissions/removals factors do not necessarily represent any specific point on the ground, but are applied to various strata. Emissions/removals factors can be applied at a single point in time (for example, biomass loss during a deforestation event) or over longer periods to represent ongoing gain or loss of carbon (e.g. ongoing loss of soil carbon, or gain of carbon by regrowth of forests). Emissions/removals factors should be representative of the spatial and temporal scale at which they are applied. Use of emissions/removals factors may represent an interim step towards Tier 3 systems, which are more complex but, properly implemented, offer advantages of better representation of the relationships between pools, and greater spatial detail (Chapter 3, Section 3.2).

Box 5: Land use and REDD+ activities

The Cancun Agreements identify five REDD+ activities, namely (a) reducing emissions from deforestation; (b) reducing emissions from forest degradation; (c) conservation of forest carbon stocks; (d) sustainable management of forests; (e) enhancement of forest carbon stocks. The IPCC 2003 GPG refers to five land uses, namely forest land, cropland, grassland, wetlands, settlements and other land. The relationship between the REDD+ activities and IPCC land uses is as follows:
  • Deforestation is an activity that converts forest land to other land uses
  • Degradation, conservation of forest carbon stocks, and sustainable management of forests are activities that occur within forest land that is not converted to other land uses, but remains forest land
  • Enhancement of forest carbon stocks can occur either by converting other land uses to forest land, or within forest land that remains as forest land.
GPG2003 regards deforestation as the sum of conversions from forest land to other land uses. As a default assumption when land is converted to another land use it remains in the land conversion category for 20 years. Therefore as a default assumption deforestation estimates should represent the sum of emissions that occur in the year of conversion of forest to another land use, and any lagged emissions or removals (e.g. due to change in soil carbon or regrowth of biomass on the subsequent non-forest land use) for 20 years thereafter. This is consistent with the advice provided in Chapter 3, Section 3.1.1.
When a REDD+ activity converts land to forest (e.g. through planting of trees on cleared land that results in enhancement of forest carbon stocks) so long as the forest created remains as forest the simplest procedure is to continue to consider over time that land as part of the original REDD+ category, rather than to later transfer to another category (e.g. sustainable management of forests). This is because i) if the land remains forest there is no subsequent land use change to worry about, and ii) countries may not in fact have selected another REDD+ activity to which the created forest could be transferred.
Countries may wish to depart from the above approach for three reasons. Firstly (and consistent with a stepwise approach) countries may not yet have the capacity to track non-forest land use. In this case if the estimates are based just on the year of conversion they will omit subsequent removals from regrowth or emissions from loss of soil carbon. As tracking capacity improves it should be possible to include lagged emissions and removals. Secondly in the case of conversion of forest that was growing on organic soils that are subsequently drained, countries may wish to continue to count these as deforestation emissions while the drainage continues, even beyond the 20 year period. Thirdly countries may wish at some point to reassign land to various REDD+ activities, probably resulting from changes in methodology or policy. In all cases countries should ensure that the REDD+ emission and removals estimates and the estimation of the FREL and/or FRL use the same methods (Chapter 3). For countries tracking lands and/or making transitions to full land use accounting, reporting challenges will become more obvious as they draw on denser and longer time series of land use change data. Neither the UNFCCC REDD+ decisions, nor GPG2003 describes how to allocate lands and emissions/removals for REDD+ activities in circumstances where there are (multiple) land use (or REDD+ activity) changes through time, but general to avoid double counting and omission of emissions and removals for countries tracking land uses the MGD advice is to:
  • where necessary, develop sub-categories under the relevant IPCC land use classes to allow transparent and consistent reporting where lands under REDD+ activities differ from IPCC land use categories
  • establish and document reporting rules that describe under which land uses emissions and removals will be reported.
Countries should ensure that tracking of lands between IPCC land uses and/or REDD+ activities does not lead bias estimates of emissions or removals, e.g. by selective inclusion. Further advice on full tracking of lands and events which lead to multiple changes in land use or REDD+ activity through time is provided in Chapter 3, Section 3.2.

Box 6: The IPCC tier concept

The IPCC classifies the methodological approaches in three different Tiers, according to the quantity of information required, and the degree of analytical complexity (IPCC, 2003, 2006).
Tier 1 employs the method described in the IPCC Guidelines using country specific activity data and the default emission/removal factors and other parameters provided by the IPCC. There are simplifying assumptions about some carbon pools (e.g. dead wood and litter pools may be combined as ‘dead organic matter’ and dead organic matter stocks are assumed to be steady for non-forest land use categories; though, for Forest Land converted to another land use, default values for estimating dead organic matter carbon stocks are provided). Tier 1 methodologies may be combined with spatially explicit activity data estimated from remote sensing. The stock change method is not applicable at Tier 1 because of data requirements (GPG2003).
Tier 2 generally uses the same methodological approach as Tier 1 but applies emission/removal factors and other parameters which are specific to the country. Country-specific emission/removal factors and parameters are those more appropriate to the forests, climatic regions and land use systems in that country and all five pools are covered explicitly. More highly stratified activity data may be needed in Tier 2 to correspond with country-specific emission/removal factors and parameters for specific regions and specialised land-use categories.
At Tier 3, higher-order methods include models and can utilize data from national ground monitoring programmes to address national circumstances. Tier 3 systems are generally more flexible than Tier 1 or 2 systems as they can more easily accommodate a wide range of different disturbance events. Properly implemented, these methods can provide estimates of greater certainty than lower tiers, and can have a closer link between biomass and soil carbon dynamics. Such systems may be GIS-based combinations of forest type and age class/production systems with connections to soil modules, integration several types and sources of data. Combined with Approach 3 they can provide accurate estimates of carbon stock changes and associated emissions and removals for changes in land use or management over time. These systems may include a climate dependency, and provide estimates with inter-annual variability.
Progressing from Tier 1 to Tier 3 generally represents a reduction in the uncertainty of GHG estimates, though at a cost of an increase in the complexity of measurement processes and analyses. Lower Tier methods may be combined with higher Tiers for pools which are less significant. There is no need to progress through each Tier to reach Tier 3. It may be simpler and more cost-effective to transition from Tier 1 to 3 directly than produce a Tier 2 system that then needs to be replaced. For example, where detailed forest inventory data is available it may be possible to develop empirical growth curves from these data almost as easily as developing emissions/removals factors (see Box 11 and example of the CBM-CFS3).

 (1)
The GPG2003 also provides three alternative methods for dealing with harvested wood products.
 (2)
Table 1.1, vol 4, section 1.3 Opens in new window contains the corresponding carbon pool definitions used in the 2006 Guidelines
 (3)
The category definitions are from section 2.2 in the GPG2003 Opens in new window
 (4)
The forest ecosystem types referred to are, for tropical ecosystems: wet; moist with short dry season; moist with long dry season; dry; montane moist; montane dry.
 (5)
GPG2003 section 3.2 Opens in new window states ‘Lands that have been converted to another land use should be tracked under the appropriate sections for as long as carbon dynamics are influenced by the conversion and follow up dynamics. 20 years is consistent with IPCC Guidelines, but Tier 3 methods may use longer periods where appropriate to national circumstances.
 (6)
In IPCC terms, forest land remaining forest land.
 (7)
 (8)
Spatially explicit means having a location that can be identified on the ground using geographical coordinates and applies to both individual sampling sites and exhaustive tessellations obtained from wall-to-wall remotely sensed data.
 (9)
See fig 5.4.2 in GPG2003 Opens in new window, Decision tree to choose a good practice method
 (10)
 (11)
The stock-change method is called the stock difference method in the 2006GL.
 (12)
See Section 2.3.2 on stratification.
 (13)
 (14)
Or may not possess an NFI with suitable statistical design
 (15)
REDD+ activities are identified in paragraph 70 of decision 1/CP.16 Opens in new window
 (16)
Other auxiliary data such as log input to processing plant together with an estimate of intermediate losses may also be relevant see Chapter 4, Section 4.2.3 or more detail.
 (17)
Because of the data requirements the stock change method is not appropriate at Tier 1.