Table of contents

1.Introduction

2.Context for Forest Reference Levels

3.Procedural Steps for laying down Forest Reference Levels (FRLs)

3.1    Assessment of draft NFAPs    

3.2    Assessment of revised NFAPs    

4.Assessment of the revised NFAPs and FRLs

4.1    Key issues in the assessment of revised NFAPs    

4.2    Corrigenda received by Member States    

4.3    Recalculation by Commission    

ANNEX A: Technical assessment

Forest definition    

Managed Forest Land area    

Consistency of Carbon Pools and greenhouse gases with GHG Inventory    

Dynamic age-related forest characteristics    

Continuation of sustainable forest management practices    

Methodological consistency with Greenhouse Gas Inventory    

ANNEX B: Recalculations

Bulgaria    

Czech Republic    

Germany    

Cyprus    

Poland    

ANNEX C: FRL and reference level without HWP

1.Introduction

The land use, land use change and forestry (‘LULUCF’) sector has the potential to contribute in three key ways to climate mitigation and the long-term climate goal of the Paris Agreement to limit global warming to less than 2 degrees:

·As a net sink, at least partly offsetting remaining emissions towards a net zero target of emissions by 2050

·As a means to substitute fossil- or carbon-intensive materials with bio-materials, in the transition to a low greenhouse gas-emitting economy

·By avoiding, to the extent feasible, emissions in agriculture, thus linked to carbon-friendly food production systems

Sustainable land management practices are key to effective carbon sequestration and precursors to carbon storage in long-living products. In addition, such practices can maintain the resilience, productivity, regeneration capacity and vitality of the land covered by the LULUCF sector, thereby underpinning economic and social development while reducing the carbon and ecological footprint of related sectors.

In line with the Paris Agreement, Member States should ensure that sinks and reservoirs, including forests, are conserved or enhanced with a view to meeting the ambitious and to be increased greenhouse gas emissions reduction targets of the Union by 2030 1 , the proposed climate neutrality target by 2050 2 and pathways to get there 3 . To this end, Regulation (EU) 2018/841 4 (hereafter ‘LULUCF Regulation’) sets out a robust accounting system to ensure that for each Member State accounted emissions from land use are at least compensated by an equivalent accounted removal of CO₂ from the atmosphere.

With regard to the accounting of managed forest land 5 , the LULUCF regulation requires the Commission to set a forward-looking benchmark that should reflect the forest age-structure, without unduly constraining future forest management intensity, such that long-term carbon sinks can be maintained or strengthened. This is achieved by applying sustainable forest management practices, as documented in the reference period from 2000 to 2009, and projecting into the future to set a forward-looking benchmark. It is against this benchmark value (or Forest Reference Level, FRL) that reported emissions and removals from managed forest land from the period 2021 to 2025 will be accounted.

Figure 1. EU LULUCF net emissions and removals by IPCC land use categories extracted from 2019 EU GHG inventory for the period 1990-2017. The reference period 2000-2009 is marked.

The LULUCF sector of the European Union (EU) 6 is a net sink, removing more CO2 from the atmosphere than emitting GHG 7 ( Figure 1 ). From 2000 to 2009 the LULUCF sector shows net removals of, on average, ‑308 million tonnes of CO2-equivalents (Mt CO2-eq) 8 . In the EU, only forests (‑414 Mt CO2-eq) and products from forests, so-called harvested wood products (HWP, ‑50 Mt CO2-eq) are net removals; all other land uses are net emissions. Most of the net removals by forests originate from stable forest land (‑369 Mt CO2-eq), i.e. managed forest land which in the EU mostly undergoes cycles of tree growth, harvest and regrowth but also includes areas under conservation status and old-growth forests, with the rest coming from land converted to forests (afforestation).

Figure 2. A: Proportion of Managed Forest Land area by Member States’ EU territory. B: Emission factor for Managed Forest Land (net removals by Managed Forest Land against Managed Forest Land area). All data are averages over the period 2000-2009 from the 2019 GHG inventory.

Member States contribute to this managed forest sink differently, depending on their geographical and economic circumstances ( Figure 2 ). National territories of Estonia, Finland and Sweden have more than 50% area coverage by Managed Forest Land; Croatia, Latvia, Austria, Portugal, Slovenia and Slovakia have shares greater than the EU average of 34%. The picture is different for removals by unit area, so-called emission factors. In the period 2000-2009, Germany, Luxembourg, Slovenia and the United Kingdom have emission factors of, on average, more than -5 t CO2 ha-1; removals per area are above the EU average of -2.7 t CO2 ha-1 in Bulgaria, Croatia, Italy, Lithuania, Latvia, the Netherlands, Poland, Romania and Slovakia.

Member States with large proportions of their territory being forested remove sizable amounts of CO2 from the atmosphere, which, depending on their total area, can make substantial contributions to the EU LULUCF sink. However, the removals per area unit, that is the effectiveness of CO2 removals by forests, is higher in central and eastern European Member States. Forest age, climatic conditions, tree species, past and present management practices, and disturbance frequency are the main reasons for these differences.

2.Context for Forest Reference Levels

Article 8 of the LULUCF Regulation sets out the accounting rules for Managed Forest Land and the FRL accounting benchmark. Article 2 defines Managed Forest Land as ‘land use reported as forest land remaining forest land’ as reported to the United Nations Framework Convention on Climate Change (UNFCCC). Greenhouse gases covered include CO2, CH4 and N2O, and carbon pools of above-ground and below biomass, litter, deadwood, soil organic carbon and harvested wood products must be reported 9 .

Member States must account their emissions and removals from Managed Forest Land for the periods 2021-2025 and 2026-2030 against the FRLs (Article 8(1)). Credits from managed forest land accounting are limited to 3.5% of the total emissions in the base year (Annex III of the LULUCF Regulation) although removals from dead wood and harvested wood products (except for the paper category) are excluded from this limit (Article 8(2)).

Article 8(5) sets out principles for setting the FRLs including:

·the continuation of sustainable forest management practice for the future compliance periods

·the reference period from 2000 to 2009

·the use of best available data

·the taking into account of the future impact of dynamic age-related forest characteristics in order not to unduly constrain forest management intensity

·the aim of maintaining or strengthening long-term carbon sinks

·the demonstration of consistency between the methods and data used to determine the proposed forest reference level and those used in the reporting to the UNFCCC

3.Procedural Steps for laying down Forest Reference Levels (FRLs)

Figure 3 illustrates the governance process for laying down Forest Reference Levels:

Figure 3. Governance process for laying down Forest Reference Levels.

3.1Assessment of draft NFAPs

Article 8(3) of the LULUCF Regulation requires Member States to submit to the Commission their NFAP, including a proposed FRL for the period from 2021 to 2025. All Member States submitted such draft NFAPs and published this document on a public website 10 . All Member States except Slovenia submitted their NFAP in English 11 .

The LULUCF Expert Group 12 , established on 30 of October 2018, consists of 10 Type A members with individuals appointed in their personal capacity, five Type C members from research institutions and non-governmental organizations, 27 13 Type D members from competent authorities of EU Member States with up to two representatives per Member State, three Type E members from other public entities and third countries’ authorities and seven observers. This Expert Group assisted the Commission with the technical assessment of the national forestry accounting plans. Specifically, it assisted the Commission in:

·Carrying out the technical assessment with the view of assessing the compliance of the FRLs with the LULUCF Regulation

·Preparing the delegated act laying down the FRLs for the first compliance period (from 2021 to 2025)

·Providing expertise when preparing the technical recommendations to the Member States

The Expert Group adopted Rules of Procedure 14 and Terms of References 15 in its first meeting on 6 February 2019, which together with the horizontal rules 16 and the call 17 set out the governance for Expert Group operation.

From 1 to 12 April 2019, the LULUCF Expert Group assessed the technical aspects of the draft NFAPs by working in five sub-groups 18 . Experts categorized issues as transparency (information missing or incomplete but likely no impact on the FRL) and accuracy if an impact on the FRL is expected. At the end of the Expert Group meeting sub-groups reported back to the whole expert group to discuss and adopt the findings 19 .

Taking into account the Expert Group report, harmonizing the views across the Member States, and own analysis, the Commission issued technical recommendations to 27 Member States 20 on 18 June 2019. This document also contains the Commission’s interpretation – without prejudice to the LULUCF Regulation – of the principles of Article 8(5) and requirements of Annex IV with criteria in Section A and elements in Section B.

During summer and fall 2019 the Commission followed up with voluntary technical bilaterals, explaining in detail the reasons for the technical recommendations and discussing possible ways forward 21 . On 2 and 3 October 2019 the Commission organized the third LULUCF Expert Group meeting to discuss possible solutions to frequently occurring issues.

3.2Assessment of revised NFAPs

Article 8(7) requires Member States to communicate their revised proposed FRL to the Commission, taking into account the technical assessment and technical recommendations. To this end, 27 Member States resubmitted revised NFAPs and 23 Member States also revised their FRL 22 . 25 Member States submitted, on a voluntary basis, explanatory notes responding to the Commission recommendations and findings from the Expert Group 23 . In accordance with Art 8(7) the Commission published the FRLs proposed by Member States on 25 February 2020 24 . Nine Member States submitted, on their own initiative or following discussions with the Commission or the Expert Group, corrections to their revised NFAP, including their FRL, or additional material for better understanding (in this document termed Corrigendum 25 ).

The Commission undertook a technical assessment of all revised NFAPs, also taking into account, if applicable, corrigenda, in two steps:

I.Have the technical recommendations been addressed?

II.Have the requirements of Art 8(5) and Annex IV of the LULUCF Regulation been met?

While the assessment of the technical recommendations (step i.) was crucial to understand the progress made by the Member State in its revision of the NFAP, it was also necessary to fully assess if all the specific regulatory requirements (step ii.) were compliant.

The Commission proposed a recalculation, therefore, where the FRL proposed by Member States was assessed as significantly not in compliance with the principles and requirements as set out in Article 8(5) and Annex IV of the LULUCF Regulation. All recalculations were communicated to the respective Member State and discussed in at least one bilateral meeting, between February and June 2020. All Member States concerned engaged in the discussions and communicated additional information to the Commission. In some cases, recalculations were withdrawn by the Commission, after supplementary information or discussion, or otherwise put forward as Corrigendum by the Member State.

Issues with no or minor impact on the FRL proposed by Member States were noted (see ANNEX A: Technical assessment ), to be addressed in 2027 either by a technical correction 26 or in the LULUCF compliance report to be submitted the Member State in accordance with Article 14(1) of the LULUCF Regulation.

The Commission’s assessment, including recalculations, were discussed in the fourth LULUCF Expert Group meeting on 26 and 27 May 2020. This expert group meeting was preceded by a preparatory meeting with Type-A members (individuals appointed in their personal capacity). Extensive discussion, flagging issues or questions and replies, took place before, during and after this expert group meeting 27 . Overall, the Expert Group found the Commission’s assessment and recalculation appropriate. Following the meeting, specific observations by the Expert Group were taken into consideration for the draft delegated act. On 22 June 2020 the Commission submitted this draft via a written procedure to the LULUCF Expert Group and also informed the European Parliament and Council of the European Union 28 . The draft act was revised, where applicable, after the Expert Group written procedure and consultation of Commission services. The general public was consulted between 17 August and 14 September 2020 29 .

In October 2020, the Commission adopted the Forest Reference Levels in a delegated act amending Annex IV of the LULUCF Regulation. The power to adopt is given by Article 8(8) if Member States submitted an NFAP and Article 8(9) if no NFAP was submitted 30 .

4.Assessment of the revised NFAPs and FRLs

Revised NFAPs and, where applicable, revised FRLs were assessed against the recommendations issued on draft NFAPs and the legal requirements of the LULUCF Regulation. Table 1 provides an overview of the proposed and adopted FRL and Corrigendum or recalculation.

According to Art 3(1) definition (7) of the LULUCF Regulation, FRLs shall be expressed in tonnes of CO2 equivalent per year. The Commission assumed “zeros” for all missing significant figures of FRLs proposed by Member States in kilo tonnes or mega tonnes. The Commission also applied mathematical rounding to the nearest tonne for Malta’s proposed FRL.

Table 1. Overview of the FRL proposed by Member States in their draft and revised NFAPs, the FRL in the delegated act, and transmitted Corrigendum by the Member State or Recalculation of the FRL by the Commission 31 . FRLs are provided in t CO2-eq yr-1 and include emissions and removals from HWP using the first-order decay function, the methodologies and the default half-life values as defined in the LULUCF Regulation.

In order to adopt sound and technically correct FRLs, the Commission undertook an in-depth assessment whether the revised NFAPs addressed the recommendations and fulfilled the legal requirements of the LULUCF Regulation. The following section “ Key issues in the assessment of revised NFAPs ” (and ANNEX A: Technical assessment ) explain the six key issues identified in the technical assessment, where:

·Issues with a substantial numerical impact were either addressed by the Member State in a Corrigendum or by the Commission in a recalculation. The following sections “ Corrigenda received by Member States ” and “ Recalculation by Commission ” (and ANNEX B: Recalculations ) explain the issues that were addressed through Corrigenda or recalculations.

·Other possible issues with no or minor impact proposed by Member States are noted in this document (see also ANNEX A: Technical assessment ), and shall be addressed in 2027 either through a technical correction or in the LULUCF compliance report. 

The Commission’s assessment concluded that, taking into account Corrigenda submitted by Member States, 23 submitted Forest Reference Levels to be in accordance with the LULUCF Regulation. The Commission undertook recalculations on issues that significantly underestimated the sink of managed forest land for Bulgaria, the Czech Republic, Germany, Cyprus and Poland.

4.1Key issues in the assessment of revised NFAPs

This section explains the key issues that the Commission identified in the assessment of the revised NFAPs and how they were addressed – through the submission of a Corrigendum by the Member State, a recalculation by the Commission, a future technical correction or clarification in the compliance report: 

· Forest definition : Article 8(5) and Annex IV, Section A (h) requires the FRL to be consistent with the GHG inventories, hence the forest definition used for the emission reporting under the Convention to the UNFCCC. In Annex II of the LULUCF Regulation Member States laid down specific minimum values for area size, tree crown cover and tree height to define forest.

Slovenia and Finland currently use forest definitions in their GHG Inventory that are different from the definitions in Annex II. This inconsistency must be addressed in the LULUCF compliance report to be submitted by Member States in accordance with Article 14(1) of the LULUCF Regulation.

· Managed Forest Land area : Article 2(1) defines Managed Forest Land as “land use reported as forest land remaining forest land” and Annex IV, Section B (e)-i requires Member States to specify the area under forest management. To ensure consistency with the GHG inventory as required by Article 8(5) and Annex IV, Section A (h), the area for Managed Forest Land must match the area of “forest land remaining forest land” as reported in the GHG inventory.

The Commission undertook a detailed assessment for all Member States to identify for which the Managed Forest Land area did not match with the area reported in the GHG inventory, for the year preceding the start of the FRL projection. Due to the significance of this parameter and absence of adequate explanation, the Commission recalculated the Managed Forest Land area for Germany and Cyprus . All Member States are in any case requested to address the Managed Forest Land area through technical corrections in 2027.

· Consistency of Carbon Pools and greenhouse gases with GHG Inventory : According to Article 8(5), “Member States shall demonstrate consistency between the methods and data used to determine the proposed forest reference level in the national forestry accounting plan and those used in the reporting for managed forest land”. Annex IV, Section B (b) states that the NFAP shall identify “the carbon pools and greenhouse gases which have been included in the forest reference level, reasons for omitting a carbon pool from the forest reference level determination, and demonstration of the consistency between the carbon pools included in the forest reference level.”

The Commission thoroughly assessed the GHG inventory consistency and as a consequence recalculated FRLs for Bulgaria , Germany , Cyprus and Poland . Ireland addressed a clerical error through a Corrigendum. Inconsistencies for other Member States mostly related to neglecting emissions from biomass burning, hence resulting in conservative FRL proposals by Member States. As in most cases technical corrections will be necessary, e.g. by making use of the provision for natural disturbances (Article 10), the Commission requests Member States to address all remaining inconsistencies through technical corrections in 2027.

· Dynamic age-related forest characteristics : According to article 8(5), the FRL “shall be based on the continuation of sustainable forest management practice, as documented in the period from 2000 to 2009 with regard to dynamic age-related forest characteristics in national forests, using the best available data”. Annex IV, section B, element (e-iii) sets out that the NFAP shall contain a description of “forest characteristics, including dynamic age-related forest characteristics, increments, rotation length and other information on forest management activities under ‘business as usual’”.

The Commission analysed and categorized how Member States addressed dynamic age-related forest characteristics. All Member States were found to be in line with this legal requirement of the LULUCF regulation.

· Continuation of sustainable forest management practices : According to Art 8(5), “the forest reference level shall be based on the continuation of sustainable forest management practice, as documented in the period from 2000 to 2009 with regard to dynamic age-related forest characteristics in national forests, using the best available data.” Further, the FRL “shall take account of the future impact of dynamic age-related forest characteristics in order not to unduly constrain forest management intensity as a core element of sustainable forest management practice, with the aim of maintaining or strengthening long-term carbon sinks.”

The Commission thoroughly reviewed how Member States described forest management activities and in particular from which period information was used. Next, the Commissioned assessed how Member States defined harvest intensity, to project past management activities from the reference period to the compliance period. The assessment focussed on the parameters and periods used to define this. Inconsistencies of significance were recalculated for the Czech Republic and Germany . Other issues were noted but found to be broadly in line with the legal requirements of the LULUCF Regulation.

· Methodological consistency with Greenhouse Gas Inventory : According to Article 8(5), “Member States shall demonstrate consistency between the methods and data used to determine the proposed forest reference level in the national forestry accounting plan and those used in the reporting for managed forest land”, and to Annex IV, Section A (h), “[…] the model used to construct the reference level shall be able to reproduce historical data from the National Greenhouse Gas Inventory”.

Member States ensure methodological consistency by ex-ante or ex-post model calibration, along with qualitative and quantitative demonstration of modelling outputs against the GHG inventory. The Commission undertook an analysis of the periods used and if methodological consistency with the GHG inventory was assured. Greece and Finland provided Corrigenda to their NFAP addressing methodological inconsistencies through ex-post calibration. The Commission recalculated for model inconsistencies for Bulgaria , Germany and Poland .

ANNEX A: Technical assessment contains the detailed analysis for all six key issues and all Member States. This also includes specific recommendations and requests for technical corrections and adjustments in the LULUCF compliance report in 2027. More detailed technical analysis of the individual NFAPs will be available in a JRC report 33 .

4.2Corrigenda received by Member States

The following list (and also Table 1 ) provides an overview of the nine corrigenda received from the Member States. These corrigenda either address corrections of specific information in the NFAP, including the FRL, or provide additional information for better understanding. The following Member States submitted corrigenda on specific matters:

·The Czech Republic noted the potential use of the natural disturbance provision. The Corrigendum is unrelated to the recalculation for the Czech Republic 34 .

·Germany provided a detailed description of the model used for estimating emissions from living biomass in the FRL and additional tables of the area development by age-volume classes, including several corrected tables and data of higher precision of Annex I of the German NFAP. This corrigendum indirectly underpins the recalculation.

·Ireland corrected a clerical error of the FRL including HWP and provided additional explanations regarding the Managed Forest Land area.

·Greece corrected the FRL for methodological consistency with GHG Inventory using ex-post calibration.

·France provided extended information and a correction of the Managed Forest Land area with no impact on the FRL.

·Latvia provided extensive additional explanation for the modelling of living biomass and other carbon pools. This document adds confidence to the Latvian NFAP but has no impact on the FRL.

·Malta provided a clerical correction of the sign of the FRL.

·Poland provided clarifications, clerical corrections and additional explanations of harvest intensity with no change of the FRL.

·Finland corrected the FRL for methodological consistency with GHG Inventory using ex-post calibration.

4.3Recalculation by Commission 

The following list (and Table 1 ) provides an overview of recalculations:

·The FRL of Bulgaria was corrected to ensure Consistency of Carbon Pools and greenhouse gases with GHG Inventory and Methodological consistency with Greenhouse Gas Inventory .

·The recalculation of the FRL for the Czech Republic concerned the Continuation of sustainable forest management practices .

· Germany ’s FRL was recalculated to ensure the Managed Forest Land area , the Consistency of Carbon Pools and greenhouse gases with GHG Inventory , the Continuation of sustainable forest management practices and Methodological consistency with Greenhouse Gas Inventory .

·For Cyprus , the Managed Forest Land area was corrected and Consistency of Carbon Pools and greenhouse gases with GHG Inventory was ensured.

·The FRL of  Poland was recalculated to ensure Consistency of Carbon Pools and greenhouse gases with GHG Inventory and Methodological consistency with Greenhouse Gas Inventory .

ANNEX B: Recalculations contains the technical documentation for recalculations by the Commission.

ANNEX C: FRL and reference level without HWP provides a comparison between estimates by Member State.

ANNEX A: Technical assessment

This Annex provides a detailed overview for all six key issues and for all Member States. This includes, where necessary, Member State-specific analysis, recommendations and requests for technical corrections or clarification in the compliance report to be submitted in accordance with Article 14(1) of the LULUCF regulation in 2027.

Forest definition

Credible accounting for Managed Forest Land requires the use of the same forest definition for laying down the FRL and for reporting data in Member State GHG inventories. Article 8(5) and Annex IV, Section A (h) requires the FRL to be consistent with the GHG inventories, hence the forest definition used for the emission and removal reporting under the Convention to the UNFCCC. Annex II of the LULUCF Regulation laid down specific minimum values for area size, tree crown cover and tree height to define forest, for each Member State.

The parameters for the forest definition in Member States’ GHG inventories 35 was found to be in line with the minimum values of Annex II of the LULUCF Regulation for all Member States except Slovenia 36 and Finland 37 (see Table 2 ). The Commission considers that in such cases of discrepancy, the FRL should be made consistent with the current GHG inventory 38 . The Commission requests Slovenia and Finland to align the forest definition between Annex II and the GHG inventory and to apply a technical correction to all relevant parameters in the compliance report to be submitted in accordance with Article 14(1) of the LULUCF regulation in 2027.

Table 2. Minimum values for defining forests in Annex II of the LULUCF Regulation and Member States GHG inventories’ National Inventory Reports (NIR) for Slovenia and Finland.

Managed Forest Land area

The LULUCF Regulation applies to the EU territories of the Member States, for which the area of Managed Forest Land sets the geographic scope of the FRL for each Member State. Article 2(1) defines Managed Forest Land as “land use reported as forest land remaining forest land”. Annex IV, Section B (e)-i requires the specification of how the area under forest management was considered as an element in the NFAP. To ensure consistency with the GHG inventory as required by Art 8(5) and Annex IV, Section A (h), the area for Managed Forest Land must match the area of “forest land remaining forest land” as reported in the GHG inventory.

The Managed Forest Land area is key to correctly setting the geographic scope for accounting of Managed Forest Land against the emissions or removals reported in the GHG inventory, and to avoid both double counting and accounting gaps. Nevertheless, neither the static nor the dynamic approach 39 are likely to predict the correct area for the compliance period at the point of laying down the FRL. Consistency of the area for Managed Forest Land with the GHG inventory at the end of the compliance period is of paramount importance. Consequently, all Member States will apply technical corrections following Article 8(11) in the LULUCF compliance report (Article 14(1)) for Managed Forest Land area once the average area in the compliance period is known from GHG inventories. Nevertheless, Member States noted below, also including Member States with inconsistencies smaller than 1000 ha, must take into account additional technical corrections, because their FRLs likely include a technical bias.

An overview of the geographic coverage, including the EU coverage of each Member State’ GHG inventory, is described in the GHG inventory of the European Union 40 . Specific inventories and inventory submission years used as reference for the FRL proposed by Member States are indicated in Table 3   41 . From that GHG inventory, the reporting year corresponds with the year preceding the start of the projection of the FRL. As a matter of principle, at the start of the projection the Managed Forest Land area must correspond to the area reported in CRF Table 4.A for row “forest land remaining forest land”. Exceptions may apply in case of:

·Exclusion of area reported as unmanaged forests: France 42 , Greece

·Exclusion of non-EU territories by sub-categories in CRF reporting tables: United Kingdom

·“Forest land remaining forest land” with no emission estimation 43 : Bulgaria, Italy, Hungary, Austria

The default reporting year is 2009, because the projection should begin after the reference period (2000-2009), but depending on specific modelling circumstances exceptions may apply. Member States with notable exceptions of very late model starts justified by best available data are: Germany, Estonia, Ireland, Croatia and Portugal.

Table 3. Overview of Reference GHG inventory and submission year, the reporting year used to assess the Managed Forest Land area, the area used in the NFAP and the approach chosen to represent Managed Forest Land in the compliance period.

Except for Malta, area differences smaller than the thousand-hectare level have been be disregarded 46 . The following list addresses MS with differences in areas, explanations provided by Member States and the specific Commission views:

·Belgium has a difference of 11% between areas of Managed Forest Land for the FRL and “forest land remaining forest land” in CRF Table 4.A. Belgium provides substantial explanations for each region in the NFAP 47 , where Managed Forest Land is limited to productive forest areas. However, the Commission notes that Belgium does not make a distinction between productive and unproductive forests in the GHG inventory, which leads to inconsistencies and potentially non-compliance in LULUCF accounting. Therefore the Commission requests Belgium to align the area for Managed Forest Land with the area reported in Table 4.A for “forest land remaining forest land” 48 by a technical correction following Article 8(11) in the LULUCF compliance report (Article 14(1)).

·Denmark has a difference of 2.5% between areas of Managed Forest Land for the FRL and “forest land remaining forest land” in CRF Table 4.A and does not provide further explanations. The Commission considers this difference substantial, but given the rather complicated FRL modelling by Denmark a correction could have resulted in unforeseen side effects. Instead, the Commission requests Denmark to align, in a transparent manner, the area for Managed Forest Land with the area reported in CRF Table 4.A for “forest land remaining forest land” by a technical correction following Article 8(11) in the LULUCF compliance report (Article 14(1)).

·Germany did not state an area for Managed Forest Land in its revised NFAP 49 but noted that the area is the same as for “forest land remaining forest land” in the most recent National Inventory Report. However, throughout the NFAP Germany referred to different GHG inventories, and tabular information in the annex on the forest area used for modelling were different from all inventories. This issue was addressed, among others, in the recalculation of the German FRL (see Germany ).

·Ireland has a difference of 6% between areas of Managed Forest Land for the FRL and “forest land remaining forest land” in CRF Table 4.A. This difference is due to the current GHG reporting of Ireland to the UNFCCC with no specific transition period for “land converted to forest land” (corresponding to the Afforested land accounting category) entering “forest land remaining forest land”, i.e. the scope of the FRL for Managed Forest Land (see also Corrigendum to the Irish NFAP). For laying down the FRL the Commission considers the difference in area justified. The Commission requests Ireland to align the area for Managed Forest Land with the area reported in CRF Table 4.A for “forest land remaining forest land” by a technical correction following Article 8(11) in the LULUCF compliance report (Article 14(1)).

·Cyprus has a difference of 7% between areas of Managed Forest Land for the FRL and “forest land remaining forest land” in table 4.A and does not provide further explanations. The Commission found that this difference was not justified and recalculated the FRL of Cyprus for this and other issues (see Cyprus ).

·Hungary has a difference of 1.3% between areas of Managed Forest Land for the FRL and “forest land remaining forest land” in CRF Table 4.A. The Hungarian NFAP notes that the area for Managed Forest Land refers to the area covered by trees and is consistent with the National Inventory Report; though this area does not correspond to the areas for which emissions and removals are estimated in the GHG inventory. The Commission notes this ambiguity and requests Hungary to align the area for Managed Forest Land by a technical correction following Article 8(11) in the LULUCF compliance report (Article 14(1)). This alignment may either be with the area reported in CRF Table 4.A for “forest land remaining forest land” or, alternatively and with substantial justification, the area for which emissions and removals are estimated in the GHG inventory.

·Malta’s area difference is an exception because the Member State currently does not estimate emissions and removals from forests. This area refers to forest for which data were available for estimating emissions and removals. The Commission recommends that Malta improves it GHG inventory and estimates emissions and removals for land use categories as required for compliance with the LULUCF Regulation.

·Portugal does not provide the area for Managed Forest land for 2016, i.e. the year preceding the start of the FRL projection. Instead, Portugal presents in its NFAP 50 the area of “forest land remaining forest land” for 2000-2009 and 2021-2025, including the Managed forest land area used in the compliance period employing the dynamic area approach. Assuming a constant annual increase of 28,000 ha in Managed Forest land area between 2009 and 2021 51 the area in 2016 would be approximately 3,983,000 ha, hence a difference of 12,734 ha or 0.3 % to the area in the GHG inventory for 2016. Although the area difference is likely marginal, the Commission requests Portugal to specify the area of Managed Forest Land for the year preceding the start of the projection, and, if applicable, align the area for Managed Forest Land by a technical correction following Article 8(11) in the LULUCF compliance report (Article 14(1)).

·Sweden has a difference of 1.4% (400,000 ha) between areas of Managed Forest Land for the FRL and “forest land remaining forest land” in CRF Table 4.A. Sweden explains the difference in area by not including low productivity forests, described as being “unmanaged”, in mountain areas in the FRL 52 . The Commission requests Sweden to align the area for Managed Forest Land by a technical correction following Article 8(11) in the LULUCF compliance report (Article 14(1)).

Consistency of Carbon Pools and greenhouse gases with GHG Inventory

According to Article 8(5), “Member States shall demonstrate consistency between the methods and data used to determine the proposed forest reference level in the national forestry accounting plan and those used in the reporting for managed forest land”, and to Annex IV, Section B (b), the NFAP shall identify “the carbon pools and greenhouse gases which have been included in the forest reference level, reasons for omitting a carbon pool from the forest reference level determination, and demonstration of the consistency between the carbon pools included in the forest reference level.”

As a matter of principle, carbon pools and greenhouse gases used for laying down the FRL must be the same as in the reference GHG inventory. The Commission thoroughly checked this precondition against quantities and notation keys provided in the reference GHG inventory submission used for setting the FRL. The Commission also assessed the inclusion of mandatory carbon pools in accordance with Art 5(4) of the LULUCF Regulation 53 and the consistency with carbon pools and greenhouse gases listed in Annex I 54 of the LULUCF Regulation. If omissions or inconsistencies were detected, the Commission notes this issue and indicates a need for a future technical correction in accordance with Art 8(11) of the LULUCF Regulation. If the impact was significant, the Commission sought clarification with the Member State with a view of correction in a Corrigendum or recalculation.

Carbon pools and greenhouse gas reporting in CRF tables

Information on carbon stock changes and emissions and removals from CO2, CH4 and N2O was extracted from each Member States reference GHG inventory submission year (see Table 3 ). Net carbon stock changes were extracted from CRF Table 4.A (Forest Land), row “forest land remaining forest land” for carbon pools of living biomass, dead wood, litter, mineral soils and organic soils. Net CO2 emissions and removals of harvested wood products (HWP) were taken from CRF Table 4.Gs1, Approach B (Production approach). The total HWP was calculated for Member States reporting HWP domestic consumption and export separately. HWP in solid waste disposal sites were not considered.

Emissions from N2O fertilization were extracted from CRF Table 4(I) (Direct nitrous oxide emissions from nitrogen inputs to managed soils), row “forest land remaining forest land” for N2O. Emissions and removals from drainage and rewetting were taken from CRF Table 4(II) (Emissions and removals from drainage and rewetting and other management of organic and mineral soils), row “forest land” 55 for CO2, CH4 and N2O. Emissions from N2O mineralization were taken from CRF Table 4(III) (Direct nitrous oxide emissions from nitrogen mineralization/immobilization associated with loss/gain of soil organic matter resulting from change of land use or management of mineral soils), row “Forest land remaining forest land” for N2O. Emissions from biomass burning were extracted from CRF Table 4(V) (Biomass burning), row “Forest land remaining forest land” and sub-categories “controlled burning” and “wildfires” 56 for CO2, CH4 and N2O.

The full time series of emissions and removals reported in reference Member States’ GHG inventories, including notation keys, was extracted and converted to CO2(-eq) 57 . For assessment against carbon pools and greenhouse gases of the FRL, the average emission value over the reference period (2000-2009) is represented in categorical form as emission or removal 58 . In case of using notation keys for all years in the reference period, the aggregate of notation keys will be retained.

Linking carbon pools and greenhouse gases between FRL and GHG inventory

Member States’ NFAPs were scrutinized for quantitative and qualitative information regarding carbon pools included in the FRL 59 . All Member States except Luxembourg 60 provided numerical information. In several cases there is a one-to-many relationship between information provided in the NFAP and tabular reporting in the GHG inventory. Most frequently, often due to the modelling framework, emissions and removals from drainage and rewetting and emissions from biomass burning have been calculated as CO2-eq; yet Member States indicated to which gases this aggregate corresponds 61 . Almost all Member States provided individual estimates for each carbon pool 62 . Allocation of estimates to carbon pools and greenhouse gases was in unclear for Denmark 63 and Latvia 64 . Portugal and Sweden included indirect N2O emission, corresponding to reporting CRF Table 4(IV), in their FRL 65 .

Croatia and Italy and estimated zero carbon stock change for dead wood and Italy also for litter. The Commission notes that such estimate is unlikely and discourages such an approach to demonstrate accordance of the FRL with Article 5(4) of the LULUCF Regulation. Instead, Member States should undertake proper efforts and introduce those estimates to their annual GHG inventories and apply a technical correction following Article 8(11) in the LULUCF compliance report (Article 14(1)). Germany modelled carbon pools and greenhouse gases with inconsistent starting dates or periods. The recalculation addressed this issue (see Germany ).

The sum of all carbon pools and greenhouse gases reported in the NFAP should correspond to the FRL. In several cases the numerical precision for individual carbon pools and greenhouse gases differs from the FRL as aggregates of all pools and gases, though, differences are usually small 66 and allow for disentangling issues related to precision from genuine numerical errors. A genuine error was found for Poland regarding an incorrect disaggregated estimate for living biomass 67 , which, among other consistency issues, was corrected in a recalculation (see Poland ). The Irish NFAP also contained an aggregation error for HWP 68 which was addressed and corrected in a Corrigendum by Ireland.

Analysis of consistency of carbon pools and greenhouse gases with GHG inventory

Table 4 presents the comparison between FRL and GHG inventory for carbon pools and emission and removal estimates of CO2, CH4 and N2O. “E” and “R” correspond to emissions and removals, respective, for the average of the GHG inventory for the reference period (2000-2009) and notation keys indicated the reason for no estimation. Cells marked in green indicate that emissions and removals were also estimated in the FRL 69 . Cells in blue indicate the inclusion of an estimate in the FRL that is not reported in the reference GHG inventory, and cells in brown show the omission of estimates from the FRL. Cells in yellow correspond to an estimate in the FRL and notation key “IE” (included elsewhere) in the GHG inventory, which could either mark an error or an estimate to ensure consistency between FRL and GHG inventory 70 ; all those cases will not be analysed further and are assumed to reflect consistency between FRL and GHG inventory.

Table 4. Consistency of carbon pools and greenhouse gases between the FRL and GHG inventory. Green…Estimated for FRL and in GHG inventory, Brown…Estimated in GHG inventory but not for FRL. Blue…Estimated for FRL but not in GHG inventory, Yellow…Estimated for FRL but included elsewhere in GHG inventory. Notation keys for estimates in GHG inventory: R…Removal, E…Emission, IE…included elsewhere, NO…not occurring, NA…not applicable, NE…not estimated. Other abbreviations: FLr…”forest land remaining forest land”, LB…living biomass, DW…dead wood, HWP…Harvested Wood Product, LT…litter, min/org SOC…soil organic carbon from mineral or organic soils.

For most Member States the carbon pools (noted as “Carbon stock change and HWP”) are consistent between FRL and GHG inventory. Member States with inconsistencies are:

·Bulgaria: omitted removals from litter and omitted emissions from mineral soils. Along with corrections of other issues the avoided carbon pools were added to the Bulgarian FRL (see Bulgaria ).

·Croatia: estimated dead wood with zero while this pool is not estimated in the GHG inventory. In principle, there is an inconsistency, but the quantitative impact is zero.

·Cyprus: omitted removals from mineral soils. Along with corrections of other issues the avoided carbon pool was added to the FRL of Cyprus 71 (see Cyprus ).

·Latvia: Inclusion of estimates for litter. The FRL includes litter estimates for organic soils, while no emissions or removals are reported in the GHG inventory.

·Malta: There are currently no estimated emissions and removals in the GHG inventory of Malta. Therefore, in principle, the estimate of living biomass is marked as inconsistency but will not be considered as such with a view to forthcoming emissions estimates by the Member State.

·Poland: Inclusion of estimates for deadwood and litter. Along with corrections of other issues related to carbon pools those pools were removed from the Polish FRL (see Poland ).

·Romania: Inclusion of deadwood and omitted emissions for organic soils.

There are no inconsistencies for emissions from N2O fertilization (estimated only by Finland and Sweden). Regarding drainage and rewetting, most Member States with specific estimates in the GHG inventory also included corresponding estimates in the FRL. Inconsistencies are noted for Latvia with omitted emission of CO2 72 and Romania for N2O. Portugal includes an estimate of N2O mineralization consistent with the GHG inventory, but Bulgaria omits this estimate.

Twelve Member States show inconsistencies for biomass burning with omitted emissions for one or several gases 73 . The consistency analysis for biomass burning is complicated by the possibility to make use of the natural disturbance provision in Article 10 of the LULUCF Regulation, which, in the case of forest fires, requires the separation between wildfires and controlled burning the FRL and the GHG inventory. To date, fourteen Member States have indicated the intention to use the natural disturbance provision, four Member States have noted that natural disturbances are implicitly included in the FRL and six Member States calculated background levels which are partly included in the FRL 74 .

Twelve Member States had no inconsistency between carbon pools and greenhouse gases included in the FRL and reference GHG inventory. Severe inconsistencies of carbon pools and greenhouse gases between the FRL and GHG inventory as well as erroneous aggregations for the FRL were corrected in Corrigenda by Member States or in Commission recalculations 75 . All other issues are considered as of minor importance and are not addressed when laying down FRLs. The topic of consistency needs to be revisited at the point of LULUCF compliance against the GHG inventory used for accounting. At that stage, issues of lower importance identified in this analysis shall also be addressed by a technical correction following Article 8(11) in the LULUCF compliance report (Article 14(1)).

Dynamic age-related forest characteristics

According to Article 8(5), subparagraph 1, the FRL “shall be based on the continuation of sustainable forest management practice, as documented in the period from 2000 to 2009 with regard to dynamic age-related forest characteristics in national forests, using the best available data”. Annex IV, section B, element (e-iii) sets out that the NFAP shall contain a description of “forest characteristics, including dynamic age-related forest characteristics, increments, rotation length and other information on forest management activities under ‘business as usual’”.

The Member States must define forest management practices and their main parameters (e.g. rotation length, age or size thresholds, target species or cohort) as documented in the period 2000 to 2009 with regard to dynamic age-related forest characteristics. The NFAP must contain adequate information on how age-related forest characteristics and their development over time were considered in the determination of the FRL.

This section addresses two separate but linked questions:

·How did Member States describe age-related forest characteristics?

·If and how did Member States document the evolution of age-related forest dynamics?

The Commission assessed how the Member States consider age in the determination of the FRL, and which information about the dynamics of age-related forest characteristics (i.e. development in the simulation) is reported in revised NFAPs. The following non mutually-exclusive categories are found:

·Age: Specific years or time periods (e.g. 5-year average) that directly enter the modelling tool as input parameters (e.g. yield tables correlating volume with age)

·Size: Dimensional features of individual trees, stands or forest landscapes that enter the modelling tool directly as input parameters e.g., diameter at breast height, basal area, or height

·Other: Criteria or parameters indirectly representing the age-related dynamics e.g., average increment or growth, or volume of growing stock.

The Commission notes that Member States, overall, adequately considered age or age-related proxies in their modelling frameworks to determine the FRL, either explicitly by age or size or implicitly by other parameters. Most Member States specifically consider age in the modelling framework ( Table 5 ), through the use of yield tables (e.g. Bulgaria, Hungary) or as input in modelling frameworks (e.g. Czech Republic, Ireland, Poland). Five Member States use only size as a proxy for age-related characteristics by diameter at breast height or girth and/or basal area (Belgium, France, the Netherlands, Austria, and Slovenia). Five Member States use other parameters to represent the age in the modelling framework, such as increment or growth (Cyprus, Malta, and Portugal), biomass density (Italy), or a combination of those 76 (Greece). Due to frequent natural disturbances, mostly by fires, in those Mediterranean Member States, forest management and forest characteristics are rather uncorrelated to age or size. Finland and Sweden use a combination of age and size (height and diameter at breast height) and Germany uses age and other parameters (volume).

In addition, the Commission assessed whether the revised NFAP contains a transparent description of the development of the relevant age-related forest characteristics e.g., by increment, area, volume of growing stock or biomass density. This information is used to understand if the Member States demonstrate that there is an age-dependent dynamic in the development of forest characteristics from the starting year of projection until 2021-2025, based on the continuation of forest management practices documented in the reference period.

The outcome of this assessment is shown in Table 5 . Nine Member States report the evolution of the area (Bulgaria, Germany, Ireland, Croatia, Latvia, Luxembourg, Austria, Poland, the United Kingdom), five Member States report the evolution of the volume (Belgium, the Netherlands, Romania, Slovenia, Sweden), Hungary and Finland report the evolution of area and volume, Greece reports the evolution of increment, and Italy the evolution of increment and biomass density. The remaining Member States provide limited or no information about the dynamics of age-related characteristics. In these cases, the Commission analysis concluded that the “limited / not reported” information on the dynamics of age-related forest characteristics does not have an impact on the proposed FRLs, but is considered an issue of transparency in the documentation.

Table 5. Information on the consideration of age or age-related proxies in modelling and on the development of age-related forest characteristics in NFAPs for the simulation period.

Continuation of sustainable forest management practices

According to Article 8(5), “the forest reference level shall be based on the continuation of sustainable forest management practice, as documented in the period from 2000 to 2009 with regard to dynamic age-related forest characteristics in national forests, using the best available data.” Further, the FRL “shall take account of the future impact of dynamic age-related forest characteristics in order not to unduly constrain forest management intensity as a core element of sustainable forest management practice, with the aim of maintaining or strengthening long-term carbon sinks.” As detailed in the Commission’s recommendations on draft NFAPs, allowing the age-related characteristics of the managed forest land to develop over time in the FRL modelling means that the harvest volumes may differ in the compliance period relative to the reference period. Any projected variation in harvested volumes in the compliance period must therefore be justified by the age-related dynamics of forest characteristics.

The assessment of the NFAPs found that the Member States put commendable effort into documenting the forest management practice that took place in the reference period 2000-2009. The majority of Member States defines forest management practices of the reference period in two distinct parts: determining modalities for forest management activities (e.g. rotation lengths, age or size thresholds, target species or cohort), and determining the relative harvest intensity. For example, rotation lengths and age or size thresholds for forest management activities are used to determine the biomass available for wood supply, while the final harvest volume projected for the FRL is modelled to continue the relative harvest intensity as documented in the reference period. Six Member States (Belgium, Denmark, Germany, Lithuania, the Netherlands and Austria) use methodologies based on age- or size class transitions, which are understood to consider the forest management activities and relative harvest intensity simultaneously.

Forest management activities

Art 8.5 requires the Member States to define the forest management practices in national forests using the best available data for the period from 2000 to 2009. These practices are then used in the estimation of the FRL. Data and information sources are expected to be national forest inventories, greenhouse gas inventories, and national statistics.

The Commission thoroughly assessed to what extent the period chosen for the definition of forest management activities, such as rotation lengths, age or size thresholds, target species or cohort, is aligned with the reference period 2000–2009 in the revised NFAPs. The period chosen to define the forest management activities can differ from the period chosen to set the harvest intensity (see Assumptions regarding harvest intensity ), and the period covered by the comparison between modelled and reported estimates as in the greenhouse gas inventories (see Methodological consistency with Greenhouse Gas Inventory ), in the framework of model calibration.

Fifteen Member States define the forest management activities including the period from 2000 to 2009. Few Member States define the management activities by using data from a period shorter but within the period from 2000 to 2009 (Germany, Latvia, Lithuania) or for specific years within the period from 2000 to 2009 (Cyprus, Luxembourg, Romania). Malta assumes no forest management activities on managed forest land. Denmark, Ireland, France, the Netherlands, Poland and Slovenia use data to define forest management activities from periods that partially overlap with the period from 2000 to 2009. The outcomes of this assessment are reported in Figure 4 .

The Commission notes that the use of data from a period which is shorter or partially outside the period from 2000 to 2009 may have an impact on the FRL, depending on the modelling assumptions and input parameters, but acknowledges that in the vast majority of cases Member States used appropriate data sources that can be considered the best available data to transparently define and document the forest management practices in the period from 2000 to 2009.

Figure 4. Reference periods used for the documentation of forest management activities.

Assumptions regarding harvest intensity

The Commission assessed two elements regarding harvest intensity:

·The parameters used to describe harvest intensity in the reference period and projected to the compliance period

·The reference period used to define harvest intensity

The assessment of the revised NFAPs found that the Member States have used a number of different parameters to determine the harvest intensity for the FRL, all relative to the forestry parameters documented for the reference period ( Table 6 ). Thirteen Member States determined intensity as the ratio between harvest and growing stock, either by total growing stock or by growing stock available for wood supply. Six Member States (Belgium, Denmark, Germany, Lithuania, the Netherlands and Austria) set the modelling framework to use the same harvest- and age class transition probability as in the reference period. Estonia, Finland 77 and the United Kingdom 78 computed the ratio between harvested area and area available for harvest, for instance in terms of maturity and management category, to remain the same as in the reference period. Greece and Sweden projected the ratio between harvest and increment to continue as in the reference period. A similar proxy for harvest intensity was used by Ireland, who calibrated the harvests projected for the FRL using the ratio between harvests and biomass increment observed in the reference period. Cyprus and Portugal projected the harvest volume per hectare to remain as in the reference period, and Malta did not project harvests for the FRL.

There is some variation in the reference period used to define the values for the parameters that determine the harvest intensity ( Figure 5 ). Nineteen Member States used the reference period 2000-2009 to define the values for these parameters, seven Member States employed data from outside 2000-2009 and Germany limited the period to 2002-2007. Member States using data outside the reference period justified their choice, mostly by issues concerning data availability and best available data, and demonstrated that the deviations do not have an impact on the forest management practice as compared to data from the reference period 2000-2009. In the case of Croatia, the impact of the war period was taken into account and documented accordingly. For Germany, the assessment concluded that the shortened period does not represent appropriately the whole reference period 2000-2009, which was therefore corrected (see Germany ).

Most Member States considered the reference period data consistently for all parameters used to determine the harvest level, usually calculating the ratio from the averages of the numerator and denominator or using average transition probabilities over the reference period. The assessment noted that Slovenia used harvest data from 2007-2012 (mid-year 2010), and defined harvest level based on the growing stock in 2000. Similar approaches were used by Poland with the average harvest from 2000 to 2009 in relation to growing stock in 2000, and by Latvia with two 5-year intervals for harvest in 2000-2004 and 2005-2009 to growing stock in 2000 and 2005, respectively. The Commission notes reservations against such choices because of potential bias introduced to the estimation of the ratio. However, these approaches described above are considered within the legal interpretation of the regulation.

The Czech Republic used salvage felling only from the period 2005-2009 when this felling type was highest, while all other harvests were based on the whole reference period 2000-2009. This model assumption was found to lead to an inconsistent representation of the reference period 2000-2009 practice, and is addressed in a recalculation (see Czech Republic ).

Table 6. Parameters chosen by Member States to define harvest intensity.

Figure 5. Reference period to determine harvest intensity.

Methodological consistency with Greenhouse Gas Inventory

According to Article 8(5), “Member States shall demonstrate consistency between the methods and data used to determine the proposed forest reference level in the national forestry accounting plan and those used in the reporting for managed forest land”, and to Annex IV, Section A (h), “[…] the model used to construct the reference level shall be able to reproduce historical data from the National Greenhouse Gas Inventory”.

The Member States must ensure that the methods and approaches used to determine the FRL are consistent with those used in the GHG inventories. To this end, the Member States must demonstrate that the model used to determine the FRL is able to reproduce historical estimates, in particular for the period 2000 to 2009.

The Commission thoroughly assessed whether or not the model output is consistent with estimates reported in the greenhouse gas inventory ( Table 7 ), and for which period this comparison is done in the NFAP ( Figure 6 ). If an inconsistency was found, the Member States should ex-post calibrate the model to ensure the alignment between the model and reported estimates.

Sixteen Member States used the period 2000 to 2009 to compare the modelled estimates with reported estimates. Estonia, Lithuania, the Netherlands, Slovakia and Finland used a period shorter or partially outside the period from 2000 to 2009. Denmark, Ireland, Spain, Cyprus and Slovenia used a comparison period completely outside the 2000 to 2009. Germany and Malta did not provide information on the comparison with historical estimates. Figure 6 provides a summary of assessment outcomes.

Figure 6. Periods used to compare modelled and reported estimates (blue) and starting year of projection (yellow).

The Commission considers that most Member States achieved consistency between the model output and historical estimates in the greenhouse gas inventories ( Table 7 ). Nine Member States 80 report information on ex-ante efforts to align input data, including ex-ante calibration. Among these, Spain provides very little information on ex-ante model calibration. Slovenia shows inconsistent figures on model consistency with the greenhouse gas inventory. Greece, France, Hungary and Finland performed an ex-post calibration. Some Member States did not explicitly show quantitative consistency between model outputs and historical estimates reported in the greenhouse gas inventory (Bulgaria, Germany, Cyprus, Lithuania, Malta, Austria, and Slovakia). In some cases ( Bulgaria for living biomass and Germany and Poland for mineral soils) the Commission decided to correct the model outputs. In the other cases:

·For Lithuania, the Commission notes that the trend in modelled estimates at first diverges from and then converges with the estimates reported in the greenhouse gas inventory. However, the Commission notes and accepts the explanations provided by Lithuania on ex-ante model calibration, in particular about the alignment with historical wood removals;

·Malta does not report emissions and removals on forest land remaining forest land, hence consistency with the GHGI could not be ensured. The Commission notes that this inconsistency will be assessed at the time of compliance, and where necessary, may be subject to a technical correction according to art 8(11).

·Austria provided only qualitative description on consistency between modelled and reported estimates, and showed a comparison about standing stock and increment between modelled and historical estimates;

·For Slovakia, despite some discrepancy between the model output and the greenhouse gas inventory, the Commission notes and accepts the explanations provided by Slovakia in the revised NFAPs, including on the ex-ante model calibration.

Table 7. Model consistency and model calibration.

ANNEX B: Recalculations

Commission recalculations only address specific issues with a notable or significant numerical impact on the FRL, while maintaining the overall structure of a Member State NFAP.

The Commission explored various options for proposed corrections and recalculations, including supplementary proposals and data provided by Member States. The following recalculations, however, only demonstrate the approach selected for estimating the FRL as laid down in the delegated act.

Bulgaria

Table 8. FRL for Bulgaria (t CO2-eq yr-1).

There is a difference of, on average, -2.6 Mt CO2-eq yr-1 in the period 2001-2009 between the model output and the GHG inventory estimates (submission 2018) for emissions and removals in forest biomass (Figure 42, page 90 of Bulgarian NFAP; Table 9 ). It is not possible to calculate the carbon stock change for years 1999-2000 because of missing information. Therefore, year 2000 is excluded from this calculation.

Table 9. Detailed calculation of the carbon stock changes in living biomass (above- and belowground) for total forest land in the period 2001-2009, and comparison with GHG inventory estimates (submission 2018). The calculation of the offset value is also provided.

Notes:

(1)     Total Forested Lands’ values, Table 7, p. 46 of NFAP of Bulgaria.

(2)     Calculated using the stock-difference method (see Equation 2.5, Chapter 2, IPCC 2006 Guidelines).

(3)     Annual values of the area of Total Forested Lands, Table 4, p. 41-42, NFAP of Bulgaria.

(4)     Calculated as follow:

where: is the carbon stock change for the year (t CO2-eq yr-1); is the carbon stock change per hectare for the year as calculated using the stock-difference method (t C ha‑1); is the cumulative area of Total Forested Lands in the year as retrieved from Table 4, p. 41-42 of NFAP of Bulgaria (ha yr-1).

(5)     Calculated as the difference between reported (2018 GHG inventory submission) and modelled emissions and removals in the period 2001-2009, as follow:

As recommended in Forsell 81 (section 2.4.4), in cases where the two trends show a consistent behaviour across time (in this case, as shown in Fig. 42, p. 90 of the Bulgarian NFAP), an adjustment factor of the projected estimates can be calculated by comparing the overlap between a set of annual estimates composed from as many years as possible. In this case, the calculated offset value in Table 9 of this document is used to correct the modelled emissions and removals for living biomass in the period 2021-2025 (see Table 10 ). Calibrated consistently with the GHG inventory estimates (2018 GHGI submission) for the period 2001-2009, the model adjustment simulates a biomass sink of -5.2 Mt CO2-eq yr‑1 (see Table 10 ).

Table 10. Detailed implementation of the offset value in the period 2021-2025 in t CO2-eq yr-1.

Notes:

(1)     Table 34, p. 90, NFAP of Bulgaria.

(2)     Calculated (see Table 9 ).

(3)     Calculated by adding the proposed offset value to the annual modelled emissions and removals in the living biomass pool.

Based on previous calculation (see Table 10 ) the FRL value for Bulgaria is adjusted (see Table 11 ). Please note that the adjusted FRL also includes previously missing emissions and removals in the litter and mineral soil carbon pools as averages over the period 2000-2009. Detailed explanation are included in Table 11 .

Table 11. Forest Reference Levels (proposed and adjusted), including and excluding the emissions and removals associated with Harvested Products in t CO2 year-1.

Notes:

(1)     Table 34, page 90, NFAP of Bulgaria.

(2)     Emissions and removals in living biomass pool, corrected by using the offset value (see Table 10 ).

(3)     Average value of emissions and removals in the period 2000-2009 (Source: Tables 4.A, 2018 GHG inventory submission, and inventory years from 2000 to 2009).

Czech Republic

Table 12. FRL for the Czech Republic (t CO2-eq yr-1).

Wood removals and harvested wood products

The Czech Republic considers wood removals separately for planned fellings and unplanned fellings (salvage fellings). In the Czech FRL, the wood removal in planned fellings is based on the ratio of harvest to available biomass observed in the reference period 2000-2009, while unplanned fellings (salvage felling) are based on the period 2005-2009 ( Figure 7 ).

Figure 7. Extract from the Czech NFAP (Fig. 8 on p. 27), showing the reference period used for final fellings and thinnings (‘Planned’) and salvage logging (‘Unplanned’), mostly due to insect infestations. As described in the Czech NFAP (p. 26), the harvest volume in ‘Other’ (residue extraction) is allocated between the planned and unplanned harvests.

The wood removals as reported in the Czech NFAP are shown in Figure 8 of this document, and the underlying data is detailed in Table 13 of this document.

In 2000-2009, the total average harvest volume was equal to 3.952 MtC yr-1 (incl. roundwood and residue extraction from planned and unplanned fellings). In the FRL, the Czech Republic considered the harvest volume from the reference period as a sum of the average salvage logging in 2005-2009 (2.276 MtC yr-1) and average planned harvests in 2000-2009 (sum of thinning and final felling, equal to 2.228 MtC yr-1), which equals to 4.504 MtC yr-1. In other words, the approach used by the Czech Republic overestimates the total harvests during 2000-2009 by 0.552 MtC yr-1, or 12.263% ( Table 13 ).

Figure 8. Extract from the Czech NFAP (Fig. 10 on p. 31). The figure on the right-hand side shows the wood removals in 2000-2025. In this document, we consider the right-hand side figure (calibrated results).

For the compliance period 2021-2025, the total wood removals reported by the Czech Republic are 4.627 Mt C yr-1. In this estimate, the Czech Republic has used the average of 2000-2009 for determining final felling and thinning volumes, but a shorter period of 2005-2009 to determine the salvage felling volume. As determined above, the approach used by the Czech Republic overestimates the harvests by 12.263%. To remove this overestimate from the FRL, the total harvest in the compliance period 2021-2025 is reduced by 12.263%, corresponding to 0.567 Mt C yr-1 (4.627 Mt C yr-1* 12.263%). As a result, the EC calculates that using the period 2000-2009 as reference for all fellings (incl. salvage fellings), the total fellings in the period 2021-2025 are 4.060 Mt C yr-1 ( Table 14 ).

The same reduction of total harvest volume, 12.263%, is reflected directly in the harvested wood product (HWP) estimates. To estimate the impact on HWP pool, this reduction in the inflow of wood is applied as a relative reduction of 12.263% in the HWP pool, reducing thereby the HWP carbon stock change from 434.49 kt C yr-1 to 381.21 kt C yr-1 (12.263%, equal to the reduction in wood removals, Table 14 ).

Table 13. Wood removals as reported in Fig. 10 of the Czech NFAP, in MtC yr-1.

Notes:

(1)    The values used by the Czech Republic for determining the FRL.

(2)    12.263% of the avg. 2005-2009 for salvage and 2000-2009 for planned felling

Table 14. The wood removals and HWP in the Czech NFAP and the JRC recalculation for the compliance period 2021-2025.

Notes

(1)      Figure 7 and Table 13 .

(2)     Value reported in the NFAP minus the overestimated amount.

(3)     Relative overestimate in the RP (12.263%) multiplied by the total removals reported by the Czech Republic for the compliance period 2021-2025.

(4)     The overestimated share relative to total wood removals.

(5)     Table 13 of the NFAP.

(6)     The overestimated share in the HWP pool is equal to overestimated share of the total fellings.

Living biomass

As detailed in Table 16 of this document, the reduced wood removals in the CP, 0.567 Mt C yr-1, are added to the stock change reported for living biomass in table 13 of the NFAP (p. 36). This addition reflects the impact of the reduced wood removal estimate on above-ground biomass.

In addition, the reduced wood removals affect the below-ground biomass pool. The stock-change in below-ground biomass is adjusted using the root-to-shoot ratios reported in the National Inventory Report 2019 of the Czech Republic, weighted by the area share reported for different tree species in the Czech NFAP ( Table 15 ). For the below-ground biomass, this calculation results in an adjustment of the stock change by +0.123 Mt C yr‑1, as shown in Table 16 of this document.

Table 15. Root/shoot ratio used in the calculations for estimation of below-ground biomass.

Notes:

(1) NIR 2019 (p. 283).

(2) Relative share of species groups reported in the NFAP for 2021-2025 (Figure 4 on p. 16).

Table 16. The estimates of the contribution of different carbon pools, as reported in the Czech NFAP and after the adjustments by the Commission.

Notes

(1)     Table 13 on p. 36 of the NFAP of the Czech Republic.

(2)     Δ C x [-44/12]

(3)     The overestimated share in % multiplied by the total removals reported by the Czech Republic for the CP (12.263% x 4.627 Mt C yr-1), as determined in Table 13 . 

(4)     Adjustment of the above-ground biomass x Rweighted average in Table 15 of this document.

Germany

Table 17. FRL for Germany (t CO2-eq yr-1).

Taking into account the work of the Expert Group, as reflected in the technical recommendations 82 , the Commission noted inconsistencies regarding Art 8(5) and Annex IV, Part A (h) of Regulation (EU) 2018/841, specifically with respect to data and methods used in the German NFAP and the German GHG inventory and the reference period 2000-2009. This recalculation addresses those issues for all carbon pools and greenhouse gases included in the German GHG reporting. The Commission notes that Germany has provided corrected data sets and additional information in an Addendum and Corrigendum to its revised NFAP as well as directly to the Commission which is taken into account in this recalculation. Germany also contributed by delivery of methodological proposals and supporting calculations, regarding at least for issues addressed in sections Emission factors for living biomass for the reference period 2000-2009 , Compliance period and Adjustments for HWP .

General approach

In recalculating the German FRL, and after consultations with Germany, the Commission used the following approach. For meeting consistency as stipulated in Art 8(5) and Annex IV, Part A (h) of Regulation (EU) 2018/841:

·The Reference period 2000-2009 is applied to all pools and gases included in the GHGI 2019.

·The pools and gases included in Germany’s GHGI 2019 are: Living biomass, Deadwood, Mineral soils and Litter, Organic soils and Drainage (CRF Table 4(II)), Harvested wood products, and Forest fires.

Emission factors for living biomass for the reference period 2000-2009

The matrix model used in Germany’s FRL proposal describes the stock-change of living biomass in 2002-2007 and is based on the German National Forest Inventory 2002 and Inventory Study 2008 (see German NFAP). The weighted emission factor of this model (NFAP Addendum and Corrigendum, Table 6, amended to NFAP, Annex 1, Table I‑2: ‑1.72 t CO2-eq ha-1 yr-1), however, is not representative for the full reference period (2000-2009).

In consultations with Germany it was decided to adjust net emissions from living biomass to reflect the full reference period, using a correction factor based on emission factors estimated by the logging factor method as described by Roehling et al. (2016)  83 . This method provides annual emission factors by using the periodical averages of stock changes in GHG reporting as a basis, but modulating these with annual harvest data. This can be interpreted as a surrogate method described in IPCC guidelines. The method was applied using updated data on harvest, which are consistent with the data as referenced in the GHG inventory (referenced there as http://faostat3.fao.org/home/E ). The same harvest data also serve for the proportional adjustment of the HWP inflow (see Adjustments for HWP ). 

Table 18 (provided by Germany) shows the emission factors used in this recalculation of the German FRL, and the underlying harvest data. The ratio for emission factors was used to correct for the reference period for living biomass (see Overview of the approaches for pools and gases ); the ratio for harvests for the adjustment in the HWP pool (see Adjustments for HWP ).

Table 18. Emission factors for the reference period as used in the recalculated German FRL, and the underlying harvest data.

Area of Managed Forest Land (MFL)

The revised German NFAP, section “Assumptions concerning area of Managed Forest Land”, states that the area of managed forest land corresponds to the area of “forest land remaining forest land” in the most recent German National Inventory Report. The same section in the NFAP states that afforested areas are added to the managed forest land after 20 years while deforestation is set to zero.

For the estimation of living biomass contribution to the FRL, the Commission estimated the managed forest land area in the compliance period 2021-2025 following the German approach described in the NFAP and illustrated in Table 19 . The area of “forest land remaining forest land” in 2017 from the GHGI 2019 was taken as starting point. For 2018 to 2025 the area entering the scope of managed forest land area 84 was added. The yearly values of the period 2021-2025 were used to estimate emissions and removals from living biomass. For simplicity with regards to other corrections, emissions and removals of all other pools and gases were estimated with the average area of reference period (10,607,070 ha); emissions and removals for HWP are independent of area.

Table 19. Area of managed forest land for living biomass.

Compliance period

The matrix model employed by Germany is only able to estimate net emissions in living biomass in full five-year cycles. In the NFAP Germany estimated net emissions from living biomass with the 5-year cycle 2018-2022 of the matrix model, to represent the emissions and removals in the compliance period.

After consultation with Germany, it was decided to match the compliance period as stipulated in the LULUCF Regulation (2021-2025). Therefore, net emissions for living biomass and HWP are calculated as weighted average of two 5-year cycles: 40% from 2018-2022 and 60% from 2023-2027. Germany also provided, for information, 5-year cycles of areas as modelled in the matrix model for 2008-2012 and 2012-2017 (see Addendum and Corrigendum to German NFAP). The estimates for all other pools and gases will be refined for the period 2021-2025 by technical corrections when the final area for managed forest land is known.

Adjustments for HWP

Changes in the net emissions for living biomass (reference period 2000-2009, compliance period 2021-2025) affect the inflow to the HWP pool.

The ratio for harvest (0.984, see Table 18 for adjustments of the reference period to 2000-2009) was applied to both 5-year cycles of the matrix model and the weighted average for the compliance period 2021-2025 ( Table 20 ).

Table 20. Adjustments in HWP inflow in reference period and compliance period.

Notes:

RSCM…reserve solid cubic meters

(1)    Corresponds to total in table I-6 of German NFAP.

(2)    Estimate provided by Germany.

The value for HWP emissions with application of the adjustment proposed for the compliance period (2021-2025) is -8,157,029 t CO2 (calculation by Germany).

Overview of the approaches for pools and gases

In order to ensure consistency with GHGI 2019 and meeting requirements of Art 8(5) and Annex IV, Part A (h) of Regulation (EU) 2018/841 the following considerations were made.

·Mineral soil and litter: As an approximation for the development of the pool after the reference period (2000-2009) in the absence of any changes in forest management, the average net emissions and removals for the reference period from GHG inventory 2019 were used, assuming that the area of managed forest land remains constant between reference period and compliance period.

·Organic soils and drainage (CRF Table 4(II)): Average net emissions and removals for the reference period (2000-2009) from GHG inventory 2019, assuming that the area of managed forest land remains constant between reference period and compliance period. These emissions include emissions from CRF table 4(II) for “forest land”, which are assumed to occur entirely on “Managed Forest land”.

·Dead wood: Average net emissions and removals for the reference period (2000-2009) from GHG inventory 2019, assuming that the area of managed forest land remains constant between reference period and compliance period.

·Forest fires: Average net emissions and removals for the reference period (2000-2009) from GHG inventory 2019, assuming that the area of managed forest land remains constant between reference period and compliance period.

·HWP: Value calculated by Germany (see Adjustments for HWP ), taking into account the changes implemented for living biomass as described in section ( Emission factors for living biomass for the reference period 2000-2009 ).