3.1.1 Estimation of emissions from deforestation
Deforestation is the conversion of Forest Land to another land category. In IPCC terms the possibilities are Cropland, Grassland, Wetlands, Settlements or Other Land. The total emissions from deforestation will depend on how much carbon was in the forest at the time of clearing, how the land was cleared and the subsequent land use. For example loss of soil carbon is likely to be greater under cropping than under permanent pasture, and will continue for some time as the disturbed pools come to new dynamic equilibrium. If deforestation is accompanied by drainage of organic soils, emissions will persist as long as the soil remains drained or organic matter remains(1).
Chapter 3 of the GPG2003 includes guidance for estimating emissions and removals associated with conversion from one land category to another covering all pools and gases with some simplifications at Tier 1. It does not include deforestation as a single conversion category because the guidance is organised around making estimates of the effect of conversion to the new category, rather than away from the previous one. This means that Chapter 3 of the GPG2003 has no specific methodological guidance for deforestation labelled as such. Since deforestation is an activity recognised under the KP, Chapter 4 of GPG2003 , which contains supplementary guidance for estimating and reporting on KP activities, does cover deforestation in the KP context, as does section 2.6 of the IPCC 2013 KP Supplement .
The MGD advice is to estimate deforestation as the sum of conversions from Forest Land to other land uses (usually Cropland, Grassland, or Settlements). Section 4.2.6 in Chapter 4 of GPG2003 cross references the sections in Chapter 3 of GPG2003 needed to do this. The relevant sections are shown in Table 9 below and can be used in conjunction with the advice below to estimate emissions from deforestation. The steps are:
- consider successively the five potential forest conversions identified by the index i
- if the conversion corresponding to the current value of i does not occur then its additional contribution to deforestation emissions for the year in question is zero
- if the conversion does occur then emissions from the newly converted area should be estimated using the methodology provided in the corresponding section of GPG2003 or where applicable the 2006GL
Even if the ith conversion did not occur in the current year, there may be emissions arising from the delayed effects, e.g. in the soil carbon pool(2) of conversions of this type that occurred in previous years. In these cases it is necessary to use historical data in estimating deforestation emissions and an assessment made of the eventual land use following deforestation. IPCC Tier 1 methods generally assume that the changes occur over 20 years and that land ceases to be in a conversion category 20 years after the conversion occurred. Therefore it would be reasonable to base deforestation emissions on conversion data covering the past 20 years unless (as discussed in Box 5 above) a country does not yet have the tracking capacity required, or wishes to use a longer period, e.g. to capture on-going emissions from drained organic soils, or wishes to reassign land to various REDD+ activities, probably of methodological or policy rationalization. In all cases countries should ensure that the REDD+ emission and removals estimates and the estimation of the FREL and/or FRL are on a consistent basis
If data are not available for such a period then deforestation emissions can still be estimated, but they will show a transient effect as the estimated lagged emissions accumulate. In all cases it is important that the actual emissions estimates and the FREL or FRL are estimated on a consistent basis. Not accounting for these lagged emissions can lead to bias in the FREL/RL and emissions reporting. Where the forests are stratified, for example according to the Forest Resources Assessment (FAO & JRC, 2012) into primary forest(3), modified natural forest(4) and planted forest(5) (which may also have various sub-strata such as wet, moist, montane etc.) the guiding steps above are repeated for each of the strata or sub-strata used.
Emissions from deforestation in the year in question are then the sum of conversions from each forest type that occurred in the current year, plus lagged effects from conversions that occurred in any category over the previous 20 years, or for the historical time period being used.
The IPCC methods identified in Table 9 cover all pools and gases for which Tier 1 methodologies are available and which may be considered the source of significant emissions from deforestation. Advice on the interpretation of the term significant in the REDD+ context in provided in Chapter 2, Section 2.2.3.
Advice on estimating the areas converted (which are the activity data required) and on estimating biomass on the Forest Land prior to conversion (this appears in the IPCC calculations for each potential conversion type as the quantity CBEFORE) is provided in Chapter 5, Section 5.1. In applying the IPCC methods listed in Table 9, the MGD process is described in Figure 7 and advice is as follows:
- Stratify the national forest area. The suggested basic stratification is into primary forest, modified natural forest and planted forest. Other stratifications may be used, but should enable reporting of these three forest categories to maintain consistency with the FAO Forest Resource Assessment. Modified natural forest may be distinguished by coupe and concession records as well as signs of canopy disturbance, detected using remote sensing data showing a shift in spectral reflectance (Margono et.al., 2012; Zhuravleva et.al., 2013), or changes in radar backscatter, or signs of disturbance such as fire scars or logging roads; or by using an NFI. Primary forests do not show these signs, although they may have been affected by natural disturbances such as fire or storms. Signs of disturbance should be treated as evidence of modified natural forest unless there is evidence that the disturbance is natural. Planted forests are identified using information on planted areas or concessions, which should be available via the NFMS from plantation companies or local or national authorities, or by using remote sensing data. There should be sub-stratification to capture ecosystems that vary in biomass density within the three main strata, which may also take account of different disturbance levels including the effect of different management types. Stratification should aim to reduce significantly variation in biomass density within a stratum.
- Obtain average biomass carbon densities for each sub-stratum identified at Step 1:
- For primary forest and modified natural forest the biomass carbon densities are referred to as CBPF, CBMNF respectively. They can be estimated by sampling or from the most recent NFI if there is one with sufficient sampling intensity, plus supplementary sampling if necessary (Appendix B). These possibilities will be referred to collectively as the sampling. The sampling should take account of previous impacts such as selective logging (in the case of modified natural forests), and natural disturbances, which will have reduced biomass carbon densities. This will require the construction of a map of logging history and prior natural disturbances, using remote sensing and ground observations (e.g. spatial records of prior harvesting, areas impacted by wildfire or cyclone). This should be used for sub-stratification to obtain relatively uniform biomass density. If the sampling comes from an NFI, it may provide merchantable volume data, in which case expansion factors (to convert forest inventory data to total above-ground biomass) and root-to-shoot ratios (to estimate root biomass from estimates of above-ground biomass) are needed to estimate total biomass(6). Many countries are developing individual tree biomass models based on basic measurements of tree diameter and height. The NFMS should be consulted to ensure that expansion factors, root-to-shoot ratios, carbon per unit of biomass, and other quantities and models are being used consistently across data sources, so that consistent estimates of biomass carbon density are obtained.
- For planted forest identified at Step 1 the carbon density can be referred to as the CBPlantF, and should be sub-stratified as necessary. CBPlantF, will depend on the age-class structure of existing planted forests and rate of growth of the species concerned, and the time of harvest and the average delay between harvest and replanting in specific planting cycles. This information should be sought via stakeholder engagement in the NFMS, and can also be supplemented using historical time series of remotely-sensed data.
- In applying the IPCC methods referenced in Table 9 use successively as CBEFORE referred to by IPCC the average values CBPF, CBMNF and CBPlantF, for each relevant sub-stratum of primary forest, modified natural forest and planted forest respectively that is deforested.
- Use remotely sensed data, plus (if available) NFI data with additional sampling if needed (Appendix B), and information available from the NFMS, to estimate the area converted from sub-stratified forest type j to another land use i. If the area A(j,i) is zero then there is no additional contribution to deforested land in the year in question, but there may be contributions to current emissions from non-zero A(j,i) values from past years. Use A(j,i) values for the current year and past years in the historical period being considered as activity data in the emission estimation method referenced in Table 9. As described in the IPCC guidance there is a need to take account of the fate of felled biomass (used either for wood processing or fuel wood, burnt or left to decay in situ).
- Emissions from each land use change stratum are estimated by multiplying the area deforested by the average change in forest carbon stocks per unit area (ΔCLC) estimated as the difference between the forest carbon stocks per unit area before conversion and the forest carbon stocks per unit area for the new land use after conversion. These are called CBefore and CAfter by IPCC. Default CAfter values are available in the 2003GL(7). Uncertainty in biomass carbon densities will lead to correspondingly uncertain emission estimates.
Table 9: Potential conversions contributing to deforestation and corresponding IPCC Guidance on emissions estimation
Section of GPG2003 where estimation method is found
Corresponding section in 2006GL
Forest to Cropland
Forest to Grassland
Forest to Wetland
Forest to Settlements
Forest to Other Land
Considerations at the decision points in the tree are as follows:
Decision Point 1: Has there been a land use change?
A land use change is determined by applying national forest definition thresholds and descriptions in combination with remote sensing and other auxiliary data. A loss of forest cover in one year does not necessarily lead to a deforestation event. Sufficiently frequent time series information combined with ground based information relating to national land use practices within all land uses can greatly assist in distinguishing land use change events from land management activities (i.e. distinguishing deforestation and forest degradation events from temporary unstocked forest lands following timber harvest and lands subject to shifting agriculture).
Lagged effects are considered in the soil carbon pool at Tier 1. Higher Tiers may consider the dynamics of other pools explicitly.
Essentially intact natural forest
Forests with native tree species that have grown naturally where there is evidence of human activities. FRA, 2015 refers to Primary Forest, Other Naturally Regenerated Forest and Planted Forest – see Forest Resources Assessment Working Paper 180
Forests composed of trees established through planting or seeding by human intervention. They include semi-natural plantation forests with indigenous species and plantation forests comprised of exotic species.
For Tier 1, factors are given in 3A.1.10 and 3A.1.8 of the GPG2003 and the corresponding tables in volume 4 of the 2006GL are Table 4.4 (for root-to-shoot ratios) and Table 4.5 (for biomass expansion factors) . At higher Tiers country specific data should be used.
Refer to the respective sections of the GPG2003 listed in Table 9 for default carbon stocks in biomass immediately after conversion (CAFTER; tC ha-1) for the post deforestation land use.
Revised carbon stock and carbon stock change factors for gleysols are provided in chapter 5 of the 2013 IPCC Wetlands Supplement
Revised CO2 and N2O emissions factor as well as additional method for off-site CO2 emissions and for CH4 emissions are provided in chapter 2 of the 2013 IPCC Wetlands Supplement
Methods for estimating CO2 and CH4 emissions from rewetted soils are provided in chapter 3 of the 2013 IPCC Wetlands Supplement