Reference Emissions Level [MGD Sections]

REDD+ Reporting [MGD Sections]

3.1.2   Estimation of emissions from degradation Previous topic Parent topic Child topic Next topic

There is wide agreement that forest degradation represents long-term loss of forest values, and that temporary loss due to harvest or natural disturbance in sustainably managed forest is not degradation.
For reporting on REDD+, carbon stock is the primary value under consideration, so degradation is interpreted here as the processes leading to long-term loss(1) of carbon without land-use change, otherwise there would be deforestation. Since sustainable management may take other forest values(2) into account, degradation based on long term loss of carbon is not necessarily the same as unsustainable forest management, more broadly defined. In this case any decreases in forest carbon stocks would be estimated through sustainable management of forests, using the method described below in Section 3.1.3. Degradation may occur in any of the forest types considered. In terms of the stratification suggested by the FAO FRA it may start from primary forest but does not have to do so. Modified natural forests, and planted forests are not degrading if the long-run average carbon stock is maintained, or is increasing. Degradation, as interpreted here, occurs in areas where long-run average carbon stock is decreasing, even if temporary increases of carbon stock occur. Regional estimates of degradation have been made in the range 5% to 132% of deforestation emissions (Houghton et al., 2009) and other estimates have been made at 25% and 47% of deforestation emissions (Asner et al., 2005, Asner et al., 2010, FRA, 2015). Forest degradation is likely to be a significant source of GHG emissions globally. Degradation is typified by a change in forest structure and species composition which may result in:
  • sustained loss of carbon from biomass and dead organic matter (DOM) pools(3);
  • sustained loss of soil C, especially from peat forests following drainage, fire or exposure after a reduction of canopy density;
  • sustained increase in emissions of non-carbon dioxide GHGs, especially from fire.
Neither the GPG2003 nor the 2006GL identifies forest degradation by name, but since it occurs on forest land and does not entail deforestation, net GHG emissions associated with it should be estimated using the methodologies described for Forest Land remaining Forest Land set out in section 3.2.1 of the GPG2003 Opens in new window(4). Detecting forest degradation and then estimating the resulting net GHG emissions, requires reliable forest observation techniques, data and resources. Countries should build upon existing systems and capacities where these are available, and integrate degradation measurement systems into their NFMS so that forest degradation is detected and measured in a manner consistent with detection and measurement of other REDD+ activities.
Multiple human-induced and natural processes can cause or contribute to forest degradation, e.g. unsustainable biomass removal from selective logging or fuelwood gathering, over-frequent prescribed burning, or drainage of peat soils. Factors such as climatic stress, wildfire and pest infestation or diseases, though they also occur in forest areas that are not degrading, may also contribute. Degradation will be more apparent where the capacity to regrow is impaired (e.g., following soil erosion, through loss of seed banks, or fragmentation caused by deforestation in adjacent areas).
Degradation may be localised (e.g. where it involves the loss of individual trees or groups of trees) or widespread (e.g., through wildfires covering many thousands of hectares or shortening of harvesting cycles for entire forest types or regions). Patterns vary from selective removal of individual trees or groups of trees, with the latter often leading to the creation of fragments which (unless part of silvicultural strategy leading to regeneration and enhanced growth) are likely to be more susceptible to further degradation. Degradation can take place after a single disturbance event or through gradual processes. Notwithstanding that temporary openings in forest cover can be part of sustainable forest management practices, use of remote sensing may significantly underestimate the extent of degradation (indicated by partial canopy cover reduction) for several reasons, including limited spectral range, the pixel size of the imagery used and the time between image acquisitions over the area of interest. For example, in cases where there is canopy closure after disturbance there may only be a short time period in which degradation can be detected by remote sensing. In other cases, the nature of partial canopy reduction may be below the minimum extent detectable by the satellite. The extent of underestimation can be reduced by using high spatial and temporal resolution data (which is more likely to detect disturbances) and by constraining data analysis so that the transition from MNF to primary forest is not allowed – that is to say once forest has been disturbed, it is assumed to remain so.
In applying the IPCC methods countries may wish to follow the steps set out below. If both forest degradation and deforestation are considered, estimates need to be consistent. In particular, the stratification called for is the same as for deforestation, and steps 1) and 2) below are common with steps 1) and 2) identified above for estimating emissions from deforestation. Step 4) below is not exactly the same as step 3) under deforestation, because the former refers to a long-run average carbon density and the latter to a current value, but the calculation methods are similar and should be consistent. Degradation as estimated by the steps below takes account of long-term reductions of carbon densities due to transitions between forest strata and sub-strata, and within the strata and substrata affected by human activity (i.e. MNF and planted forests). For estimating degradation the steps are:
  1. See Step 1 under Deforestation (Section 3.1.1)
  2. See Step 2 under Deforestation (Section 3.1.1)
  3. Estimate the annual change in CBMNF. Call this quantity BMNF. It may be estimated from repeated NFIs if these exist, by sampling as set out below, by using the gain-loss method as set out in section of GPG2003 Opens in new window. It should take account of sub-stratification and factors including forest growth, logging, fuelwood harvest and fire. ΔCBMNF will be positive if ΔCBMNF is increasing, and zero or negative otherwise. In order to ensure the terms in the following equation have the correct sign, set factor fMNF = 0 if ΔCBMNF is positive or zero and fMNF = +1 if ΔCBMNF is negative.
  4. Estimate the annual change in the long-run (LR) average carbon density in planted forests. The long-run average carbon density is the carbon density averaged across the forest rotation taking account of both growth and harvesting events, and over successive forest rotations. This implies assessment of anticipated forest growth and removals due to harvest especially when there is a significant proportion of newly established planted forest in the planted forest estate. Call this quantity LRCBPlantF and the annual change ΔLRCBPlantF. First estimate LRCBPlantF for the current year, which will depend on the rate of growth of the species concerned, the frequency of harvest and the average delay between harvest and replanting all as anticipated in the current year. This information should be available via the NFMS, from national forest authorities or from commercial operators. Box 9 gives an example of the type of the calculations required. Subtract from the current value the value of LRCBPlantF in the previous year to obtain ΔLRCBPlantF. This will be positive if LRCBPlantF is increasing, and zero or negative otherwise. Set fPlantF = 0 if ΔLRCBPlantF is positive or zero and fPlantF = +1 if ΔLRCBPlantF is negative.
  5. Using the methods described in Chapter 5 to estimate the annual transfer of areas from primary forest to modified natural forest. Call this quantity ΔAPF>MNF.
  6. Using the methods described in Chapter 5 to estimate the annual transfer of areas from primary forest to planted forest. Denote this quantity ΔAPF>PlantF.
  7. Using the methods described in Chapter 5 to estimate the annual transfer from modified natural forest to planted forest. Denote this quantity ΔAMNF>PlantF.
  8. Estimate annual carbon dioxide emissions from degradation (CO2degrad) using the following equation. The significance of the individual terms is described in the steps above and summarized in the Table 10:
Equation 1
Inclusion of a quantity in square brackets means that, if negative, the quantity should be treated as zero, so that the corresponding term will not then affect the total emissions from degradation. The fPlantF and fMNF multipliers perform a similar function so that only long-run decreases in carbon density contribute to degradation. Vertical lines mean that the absolute value of the quantity which they enclose should be used. The table below shows the degradation processes to which the five terms on the right hand side of the equation respectively correspond. Since the terms are separately identified, degradation may be disaggregated by process or treated as a sum over processes. For example, if countries wish to distinguish between degradation which may occur in primary and modified natural forest (on the one hand) and that which may occur in planted forest (on the other) then the 5th term in Equation 1 should be removed, and treated separately. The terms in the equation should be sub-divided to take account of sub-stratification.
At Tier 1, GPG2003 assumes that for Forest Land remaining Forest Land, mineral soil, dead wood and litter pools are in equilibrium. If higher Tier methods are being used, national data should enable Equation 1 to be expanded to include them. If organic soils are drained to establish planted forest, emissions should be estimated for the corresponding planted forest areas as set out in section of GPG2003 Opens in new window. Tier 1 carbon dioxide emission/removal factors reported in the IPCC guidance and guidelines for organic soils under different circumstances are summarised in Table 11.

Table 10: Terms used in Equation 1

Number of terms in RHS of Equation 1.
Degradation process
Term on the right hand side of Equation 1
Multiplies the whole of the right-hand side of the equation and converts from mass of carbon to mass of carbon dioxide
Conversion of primary forest to modified natural forest
Conversion of modified natural forest to planted forest
Conversion of primary forest to planted forest
Decrease in long-term carbon density of modified natural forest
Decrease in long-term carbon density of planted forest

Table 11: Sources of emission/removal factors of organic soils

Chapter and Section Number
Table Number
Description of default emissions factors
Annual CO2-C emission factor for drained organic soils in managed forests
Annual CO2-C emission factor for cultivated organic soils
Annual CO2-C emission factor for managed grassland organic soils
Annual CO2-C and N2O-N emission/removal factors for drained organic soils in managed forests
Annual CO2-C emissions factor for cultivated organic soils
Annual CO2-C emission/removal factors for drained grassland organic soils
Annual CO2-C on-site emissions/removals factor and CO2-C off-site emission factor for drained organic soils in all land-use categories
Annual N2O-N emissions factor for drained organic soils in forest land
CO2-C and CH4 emissions/removals factors for peat fires in all land-use categories

Box 9: Estimating long-term average biomass density in planted forests

Biomass density (above- and below ground) in a planted forest subject to multiple harvest and subsequent growth will show the saw-tooth pattern illustrated in the figure below. The long-term average carbon density is the carbon density averaged over the initial subsequent rotations. If replanting is immediate this will be a fraction f1 of the above-ground biomass density at the time of each harvest. The fraction f1 is commonly about 0.5. If there is significant delay (say δt) between harvest at the time of replanting and the time from replanting to harvest is t1 then the long-run average biomass density is P.(f1.(t1/(t1+δt))+r) where P is the above-ground biomass density at the time of harvest and r is the root-to-shoot ratio. P and r will depend on species, site conditions and management inputs. If there are 0.5 tonnes of carbon per tonne of biomass then LRCBPlantF = (0.5) P.(f1.(t1/(t1+δt))+r).The basic information required from stakeholders is growth rates and the timing and nature (biomass removed) of harvest, and whether there are significant delays in replanting. Better values can be obtained using growth models which can take account of the effect of disturbance on r. Other carbon pools are taken into account at higher Tiers.

That is to say, increase in the extent of forest strata with lower carbon density, averaged over harvest cycles if appropriate, or declining carbon density within strata as revealed by sampling over time. The application of these ideas is discussed below.
For example biodiversity, fire control, water management or productive capacity
See Table 2 for pool definitions
The IPCC Task Force on National Greenhouse Gas Inventories (TFI) has developed additional national-level inventory methodological guidance on wetlands, including default emission factor values, with the aim to fill gaps in the coverage of wetlands and organic soils in the 2006 IPCC Guidelines. This document is called 2013 Supplement to the 2006 IPCC guidelines for National greenhouse gas inventories: Wetlands (the 2013 IPCC Wetlands Supplement) Opens in new window.