Grassi, G., House, J., Dentener, F., Federici, S., den Elzen, M., & Penman, J. (2017). The key role of forests in meeting climate targets requires science for credible mitigation. Nature Climate Change, 7(3), 220-228. https://doi.org/10.1038/nclimate3227 Peer reviewed version Link to published version (if available): 10.1038/nclimate3227 Link to publication record in Explore Bristol Research PDF-document This is the author accepted manuscript (AAM). The final published version (version of record) is available online via Springer Nature at https://www.nature.com/nclimate/journal/v7/n3/full/nclimate3227.html. Please refer to any applicable terms of use of the publisher. University of Bristol - Explore Bristol Research General rights This document is made available in accordance with publisher policies. Please cite only the published version using the reference above. Full terms of use are available: http://www.bristol.ac.uk/red/research-policy/pure/user-guides/ebr-terms/ 1 Key role of forests in meeting climate targets but science needed 2 for credible mitigation 3 4 5 Authors 6 Giacomo Grassi*, European Commission, Joint Research Centre, Ispra (VA), Italy 7 Jo House, Cabot Institute, Bristol University, UK 8 Frank Dentener, European Commission, Joint Research Centre, Ispra (VA), Italy 9 Sandro Federici, FAO consultant, Italy 10 Michel den Elzen, PBL Netherlands Environmental Assessment Agency, The Hague, The 11 Netherlands 12 Jim Penman, University College London, UK 13 14 15 *Corresponding author: giacomo.grassi@ec.europa.eu 1 16 ABSTRACT 17 18 Forest-based climate mitigation may occur through conserving and enhancing the carbon sink 19 and through reducing greenhouse gas emissions from deforestation. Yet the inclusion of forests 20 in international climate agreements has been complex, often treated separately or considered a 21 secondary mitigation option. In the lead up to the Paris Climate Agreement, countries 22 submitted their (Intended) Nationally Determined Contributions ((I)NDCs), including climate 23 mitigation targets. Assuming full implementation of (I)NDCs, we show that land use, and forests 24 in particular, emerge as a key component of the Paris Agreement: turning globally from a net 25 anthropogenic source during 1990-2010 (1.3 ± 1.1 GtCO2e/y) to a net sink of carbon by 2030 (up 26 to -1.1 ± 0.5 GtCO2e/y), and providing a quarter of emission reductions planned by countries. 27 Realizing and tracking this mitigation potential requires more confidence in numbers, including 28 reconciling estimates between country reports and scientific studies. This represents a challenge 29 and an opportunity for the scientific community. 2 30 MAIN TEXT 31 32 In December 2015, 195 countries adopted the Paris Climate Agreement1 at the 21st Conference of 33 Parties (COP-21) of the United Nations Framework Convention on Climate Change (UNFCCC). As 34 part of the process, 187 countries, representing more than 96% of global net emissions in 20122, 35 submitted their Intended National Determined Contributions3 (INDCs, which become NDCs with the 36 ratification of the Paris Agreement4). The NDCs are the basis for implementing actions under the 37 Agreement, and the vast majority include commitments in the land-use sector. 38 Land use, including agriculture and forests, accounts for about 10% of global greenhouse gas (GHG) 39 emissions as CO2, and nearly quarter including CH4 and N2O5-9. Also, about one third of the current 40 anthropogenic CO2 emissions are removed by terrestrial ecosystems, mainly forests. While 41 deforestation is estimated to be the main GHG source in many tropical countries, forest sinks are 42 important globally with net sinks dominating in temperate and boreal countries10. 43 Including land use in the UNFCCC process has been long and complex. For forests, uncertainties of 44 GHG estimates and methodological issues such as additionality (i.e. showing that proposed mitigation 45 efforts go beyond Business-as-Usual (BAU) and separation of non-anthropogenic effects) and leakage 46 (displacement of land-use activities to other areas) have often led to controversies and compromises11. 47 The UNFCCC requires that all countries report GHG inventories of anthropogenic emissions and 48 removals using methodologies developed by the Intergovernmental Panel on Climate Change (IPCC) 49 and adopted by UNFCCC12. Developed countries report annual GHG inventories13, using mandatory 50 and voluntary land-use activities towards meeting their emission reduction targets where applicable 51 under the Kyoto Protocol14. Developing countries’ GHG inventories have historically been reported 52 less frequently15, though biennial updates are now required16, and may undertake voluntary mitigation 53 activities, notably through the REDD+ process (Reducing Emissions from Deforestation, forest 54 Degradation, and other forest activities). 3 55 The Paris agreement is a potential game changer for land use mitigation. It calls explicitly for all 56 countries to make use of a full-range of land-based mitigation options, and to take action on REDD+. 57 Based on country information, this analysis quantifies the expected GHG mitigation role of the land- 58 use sector in the (I)NDCs to the year 2030, including activities conditional on finance, technology and 59 capacity-building support. It does not assess specific country policies. It focuses on CO2 emissions 60 and removals and non-CO2 emissions from Land Use, Land-Use Change and Forestry (LULUCF, 61 primarily deforestation and forest management), encompassing most of the land-use sector identified 62 in (I)NDCs. Harvested wood products are included for most developed countries. Non-CO2 emissions 63 from agriculture are not included. 64 65 Country mitigation targets are expressed in different ways 66 Countries express their (I)NDC targets with different combinations of the following elements17-19 67 (Supplementary Tables 1-2), reflecting different national circumstances, i.e.: 68 • Quantifier - targets are expressed as either an absolute quantity e.g. amount of GHG reduction in 69 tonnes of CO2 equivalent (tCO2e), or as a change in the emission intensity, e.g. China and India 70 express a reduction of emission intensity per unit of GDP. 71 • Reference point – Emissions in the target year (e.g. 2025 or 2030) are compared to either a historic 72 base year (e.g. 1990, 2005) or to the target year in a BAU scenario. The BAU scenario assumes 73 either no mitigation activity, or some existing mitigation activity. 74 • Conditionality - While developed country (I)NDC targets are unconditional, most developing 75 countries expressed at least part of their targets as conditional on finance, technology or capacity- 76 building support. 77 The (I)NDCs vary in the way they include LULUCF. It may be fully included as part of the overall 78 target like other sectors, or partially included (e.g., only deforestation), or considered separately with 4 79 special mitigation actions or accounting rules. Consequently, evaluating the expected effect of 80 LULUCF on the (I)NDC mitigation targets is complex. 81 Our analysis is based on information provided on LULUCF in the (I)NDCs3, and also (in order of 82 priority) other country reports to UNFCCC13,15,16,20,21, other official country documents, and FAO- 83 based datasets for forest8,22 and for other land uses23 (Supplementary Tables 4-5). Given the Paris 84 Agreement context of our analysis, we prioritized (I)NDCs and those country reports which are 85 formally reviewed or technically assessed by UNFCCC, with FAO-based datasets used for gap filling, 86 allowing global estimates covering 195 countries (see Methods). We found sufficient information to 87 analyse the LULUCF mitigation contribution for 68 countries (or 41 (I)NDCs, with the EU’s NDC 88 representing 28 countries), representing around 78% of global net emissions in 20122 and 83% of the 89 global forest area22. For the remaining countries, where LULUCF is not expected to offer a large 90 mitigation potential (Supplementary Section 1), the future LULUCF mitigation contribution was 91 assumed to be zero. 92 93 Historical and projected forest emissions and removals 94 Fig. 1 shows, for all 195 UNFCCC countries, historical and future anthropogenic LULUCF emissions 95 and removals from this analysis, based on official country data. The Supplementary Sections 2 and 3 96 provide, respectively, additional country-specific assessments and an analysis of uncertainties for the 97 absolute level of net emissions and their trend24,25, based on information from countries’ reports. 98 While country information on uncertainty up to 2030 is not available, we conservatively assumed that 99 the uncertainty estimated for historical net emissions would also hold for the future. 100 Historically, global LULUCF net emissions decreased from 1.54 ± 1.06 GtCO2e/y (95% CI) in 1990 101 to 0.01 ± 0.86 GtCO2e/y in 2010 (slope of linear trend: -0.08 GtCO2e/y). The trend and the inter- 102 annual variability over this period are influenced by: (i) deforestation in Brazil, with peak years in 103 1995 and 2002-2004 followed by a steep reduction of emissions by about -1.3 GtCO2e/y till 2010; (ii) 104 high deforestation rates (1997-1999) and peak years in peat fire emissions (e.g., 1997) in Indonesia; 5 105 (iii) an increasing sink in managed temperate and boreal forests, of about -0.8 GtCO2e/y from 1990 to 106 2010. By splitting the 1990-2010 period (average emissions: 1.28 ± 1.15 GtCO2e/y) into four sub- 107 periods, we conclude that the historical trend is statistically significant after 2000 (Supplementary 108 Section 3). 109 The wide range of future LULUCF net emissions depends on policy scenarios (Fig. 1). The ‘country- 110 BAU’ scenario foresees a marked increase in global net emissions (Supplementary Table 6), reaching 111 1.94 ± 1.53 GtCO2e/y in 2030. This is because several developing countries assumed BAU to be a no- 112 measures scenario, e.g. ignoring the existing policies to reduce deforestation. Under the ‘pre-(I)NDC 113 scenario’, i.e. considering policies in place prior to COP-21 (including the earlier Copenhagen 114 pledges21), global net emissions increase moderately, up to 0.36 ± 0.94 GtCO2e/y in 2030. For the 115 ‘unconditional (I)NDC scenario’ the global net emissions slightly decrease, reaching a sink of -0.41 ± 116 0.68 GtCO2e/y in 2030. An additional reduction of net emissions is estimated for the ‘conditional 117 (I)NDC’ scenario, leading to a sink of -1.14 ± 0.48 GtCO2e/y in 2030. 118 The analysis of the emission trend over the entire period shows that the difference between the 1990- 119 2010 average and the net emissions in 2030 is not significant for the pre-(I)NDC scenario, but is 120 significant (95% CI) for both the unconditional and the conditional (I)NDC scenarios (Supplementary 121 Figure 3b). This indicates that the reduction of net emissions assumed by the (I)NDCs relative to the 122 historical period, if achieved, is statistically robust. 123 124 Comparison with global datasets 125 Fig. 2 compares the historical LULUCF trend from our analysis to other three well-known global 126 LULUCF datasets: (i) latest country reports to UNFCCC (ref13,15,16,20); (ii) FAOSTAT for all land 127 uses23; and (iii) IPCC Fifth Assessment Report (AR5) Working Groups (WG) I5 and III6 data used for 128 the global carbon budget. 6 129 The difference between this analysis and the UNFCCC country reports is because several (I)NDCs 130 updated past datasets, and because we used FAO-based data for gap-filling, instead of pre-2010 131 National Communications. 132 Differences between this analysis and FAOSTAT include the definition of forest (UNFCCC vs. 133 FAO); coverage of areas and of carbon pools; and differing estimation methods by reporting 134 agencies8 (see Methods). 135 There is a large difference of about 3 GtCO2e/y between this analysis, based on country reports 136 following the IPCC Guidelines for national GHG inventories25,26 (IPCC GL), and the scientific studies 137 summarized by the IPCC AR55,6, For the period 2000-2009, the level of net emissions is on average 138 0.90 ± 1.11 GtCO2e/y (95 % CI) in our analysis and 4.03 ± 2.93 GtCO2e/y (90 % CI, reflecting both 139 methodological and terminological choices27-29) in IPCC AR5 (Fig. 2). The above differences are 140 linked to different scopes of the two IPCC work streams30: the GL focus on internationally agreed 141 methodologies for national anthropogenic GHG estimation, recognizing different countries’ 142 definitions and technical capabilities, whilst the AR5 focuses on assessing the state of the science on 143 the global carbon budget using globally applied data, definitions and modeling methods. 144 Specifically, LULUCF in the IPCC GL includes estimates of GHG emissions and removals from all 145 land uses, reported under either a stable or changed land-use status (typically in the last 20 years), e.g. 146 “forest remaining forest” or “forest converted to cropland” (or vice versa). There is a large scientific 147 challenge of providing a practicable methodology to factor out direct human-induced mitigation 148 action from indirect human-induced and natural effects31,32, such as the natural aging of forests, 149 natural disturbances and environmental change (e.g. climate change, extended growing seasons, 150 fertilizing effects of increased [CO2] and nitrogen deposition). Therefore, the IPCC GL25,26 use the 151 category of “managed land” as a default first order approximation of “anthropogenic” emissions and 152 removals, based on the rationale that the preponderance of anthropogenic effects occurs on managed 153 land32. The GHG inventories should report all emissions and removals for managed land, while GHG 7 154 fluxes from unmanaged land are excluded. What is included in “managed land” varies from country 155 to country, although the countries’ definition must be applied consistently over time. 156 In contrast, global models such as those used in IPCC AR5 and the Global Carbon Project take a 157 different approach to separate anthropogenic from natural effects. Anthropogenic fluxes (referred as 158 “net land-use change”5,9, or “Forestry and Other Land Uses”6), are estimated by a bookkeeping 159 model27 or by dynamic global vegetation models9 based on changes in land cover (i.e. between forest 160 and agriculture), forest regrowth and, depending on the modeling capability, some forms of 161 management (wood harvest and shifting cultivation). The difference between this modeled 162 “anthropogenic” flux and the estimated total net flux of CO2 between the land and atmosphere9 is the 163 “residual terrestrial sink”5,6,9, which is generally assumed to be a natural response of primary or 164 mature regrowth forests to environmental change9,27. 165 The above methodological differences are reflected in the net emissions from developed countries, 166 where most of the ≈ 3 GtCO2e/y difference between our analysis and IPCC AR5 occurs for the period 167 2000-2009: while these countries report a substantial “anthropogenic” sink (-1.9 GtCO2e/y in 168 “UNFCCC Annex 1” countries), the bookkeeping model (IPCC AR5) finds a small net source (0.1 169 GtCO2e/y, “OECD” in Fig. 11.7 of ref.6). This difference lies essentially in whether the large sinks in 170 areas designated by countries as “managed forest” (following IPCC GL), well documented in forest 171 inventories10, are attributed to “anthropogenic” (in the GHG inventories) or to “natural” fluxes (in 172 IPCC AR5). 173 To explore, at least in part, the impact of these different attribution methods, Fig. 3a compares what is 174 considered undisputedly “anthropogenic” by both IPCC AR5 (land-use change) and the country 175 reports (land converted to other land uses). These estimates, both predominated by tropical 176 deforestation, are of similar magnitude, especially after 2000. The other fluxes, where the attribution 177 differs more between IPCC AR5 and the countries, are shown in Fig 3b. Thus much of the sink that 178 countries report under ‘forest remaining forest’, the global models consider part of the natural flux. 179 This disaggregation suggests that the residual sink is at least partly influenced by management 8 180 practices not captured by global carbon models33, but also that countries consider anthropogenic what 181 is partly influenced by environmental change and by recovery from past disturbances. 182 There are many reasons for the lower sink reported by countries in Fig 3b compared to the residual 183 sink from IPCC AR530, including the fact that countries do not report sinks for unmanaged lands (e.g., 184 a large sink in tropical and boreal unmanaged forests10) and their reporting for managed land may be 185 incomplete, i.e. ignoring fluxes (e.g. sink in grasslands, wetlands or forest regrowth) or carbon pools. 186 There would be other factors to consider, including treatment of legacy fluxes from past land-use and 187 other definitional and methodological differences. These would require a more detailed analysis, 188 which is outside the scope of this paper. 189 Finally, the projections from this analysis can be compared to RCP scenarios used in IPCC AR5 up to 190 2030 (Fig. 3, dashed lines). For the undoubtedly “anthropogenic” fluxes (Fig. 3a), our country data 191 analysis falls broadly within the IPCC AR5 scenarios, supporting previous qualitative findings34. 192 Overall, our analysis shows 1) that various global LULUCF datasets may be more consistent than 193 apparent at first glance, 2) unless the scientific and GHG inventory community appreciate these 194 definitional and methodological issues, conflicting numbers and messages are likely to appear in the 195 coming years, and 3) that several reasons for the differences among datasets can be further reconciled 196 in collaboration between the two communities, which would be a very useful contribution to science 197 and policy. 198 199 Different perspectives on mitigation contribution by forests 200 To reflect the complexity of approaches to (I)NDCs, this analysis assesses three different perspectives 201 on LULUCF mitigation: 202 (A) 2030 (I)NDC vs. 2005, i.e., the expected impact of full (I)NDC implementation. The year 2005 is 203 chosen as historically reliable in terms of data. Fig. 1 shows that the global LULUCF net emissions to 204 the atmosphere would transition from an estimated net anthropogenic source of +0.8 GtCO2e/y in 9 205 2005 to a net sink of -0.4 GtCO2e/y (unconditional (I)NDCs) or -1.1 GtCO2e/y (conditional (I)NDCs) 206 in 2030. 207 (B) 2030 (I)NDC vs. 2030 alternative scenarios: country-BAU or pre-(I)NDC, i.e., the additional 208 LULUCF contribution relative to alternative scenarios (Fig. 1). The magnitude of the difference 209 between country-BAU and pre-(I)NDC (1.6 GtCO2/y) may raise concerns about the expected results- 210 based payments under REDD+, which should be based on credible baselines and not on a no- 211 measures scenario. Clarification of the role of REDD+ in (I)NDCs should therefore be seen as a 212 priority by countries. Compared to the estimated pre-(I)NDC scenario, net emissions in 2030 are 213 lower by 0.8 GtCO2e/y or 1.5 GtCO2e/y for unconditional and conditional (I)NDCs, respectively. For 214 the ‘conditional (I)NDC vs. 2030 pre-(I)NDC’ scenario (Fig. 4a), this LULUCF contribution of 1.5 215 GtCO2e/y (Fig. 4a, last column) represents 26% of the total mitigation expected from all GHG sectors 216 (5.9 GtCO2e/y35, Fig. 4a, third column). The countries contributing most to LULUCF mitigation 217 under this perspective are Brazil and Indonesia, followed by other countries focusing either on 218 avoiding carbon emissions (e.g. Ethiopia, Gabon, Mexico, DRC, Guyana and Madagascar) or on 219 promoting the sink through large afforestation programs (e.g. China, India). 220 (C) Country perspective on emissions reduction in the (I)NDC, i.e. what each country might consider 221 its “LULUCF contribution to the overall (I)NDC”, as part of its mitigation package, e.g. if a country 222 commits to reduce its all-sectors emissions by x% relative to y (reference point: base year or BAU- 223 scenario), what fraction of x is attributable to LULUCF? This approach looks at the way countries 224 define their (I)NDCs (e.g. reference point) and the way LULUCF is included within the (I)NDC (as 225 any other sector or with special accounting rules). Globally, under this perspective the LULUCF 226 contribution is 3.1 GtCO2e/y (unconditional) or 3.8 GtCO2e/y (conditional). The latter case (Fig. 4b, 227 last column) corresponds to 24% of total all-sectors emission reduction relative to the reference point 228 (i.e. 15.8 GtCO2e/y, Fig. 4b, third column). 229 The emission reductions from a country perspective (Fig. 4b) are greater than the deviation from the 230 pre-(I)NDC scenario (Fig. 4a), because countries’ choices of reference point in their (I)NDCs tend to 10 231 maximize the accounted mitigation, i.e. countries that already reduced emissions used a historical 232 base year, whereas countries expecting a future increase of emissions used a future BAU-scenario. 233 This is evident under perspective C, where nearly one third of the contribution comes from Brazil, 234 followed by Indonesia and Russia (Fig 4b, last column). In Brazil, where total emissions have 235 declined after 2004 due to successful implementation of policies to reduce deforestation36, the NDC 236 target (-43%) is relative to 2005. Our analysis suggests that in Brazil the LULUCF contribution to 237 NDC is greater than the all-sectors NDC target for 2030, i.e. the NDC allows emissions from other 238 sectors to increase. In Indonesia the conditional NDC target (-41%) is relative to the BAU-scenario in 239 2030. LULUCF represents about 65% of current (2010) total emissions and is expected to contribute 240 nearly two-thirds of the NDC emission reduction (relative to BAU) foreseen in for 2030. Brazil and 241 Indonesia are representative examples of GHG emission trends in developing countries: with an 242 expanding and industrializing economy, the currently high LULUCF emissions are expected to 243 decrease, and be superseded by growing emissions from the energy sector. The (I)NDC target of 244 Russia (-30%) is relative to 1990, with LULUCF contributing by about two-fifths to this emission 245 reduction. Russia is more important in perspective C than in B because its specific accounting method 246 for LULUCF gives prominence to the contribution of the current forest sink to climate mitigation. 247 The (I)NDCs of the three countries above may be assessed also in terms of clarity and trust of 248 information provided (see Supplementary Section 2). Overall, Brazil’s NDC is transparent on the 249 land-use sector and the underling GHG estimates are based on a well-developed monitoring system. 250 The recent relevant upward revision of historical deforestation emissions in Brazil opens new 251 questions on the implementation of the NDC target and on how and when data consistency between 252 NDC, REDD+ and National Communications will be ensured. The relative ambiguity of Indonesia’s 253 NDC on how it would address land use emissions is improved by the information in more recent 254 documents. Furthermore, recent monitoring efforts have improved the GHG emission estimates, 255 especially from peatland drainage and from forest degradation, whereas emissions from peat fires 256 remain very uncertain. These improvements are mainly due to the REDD+ process, which in many 257 developing countries is triggering unprecedented monitoring efforts. The challenge is increasingly to 11 258 transfer these improvements into the NDC process, and to clarify the often uncertain relationship 259 between the financially-supported REDD+ activities and the NDCs. For Russia, transparency of 260 mitigation efforts will crucially depend on clarifying the accounting method chosen for LULUCF. In 261 addition, credible GHG estimates will require reconciling or explaining the currently large difference 262 in the forest sink between the reports submitted by Russia to UNFCCC and to FAO. 263 In summary, the full implementation of (I)NDCs would turn LULUCF globally from a net source 264 during 1990-2010 (1.3 ± 1.1 GtCO2e/y) to a net sink by 2030 (up to -1.1 ± 0.5 GtCO2e/y). The 265 absolute LULUCF mitigation contribution in 2030 is very different depending on the way that 266 mitigation is calculated, ranging from 0.8 to 3.1 GtCO2e/y for unconditional (I)NDCs and from 1.5 to 267 3.8 GtCO2e/y for conditional (I)NDCs (Supplementary Table 3). However, in relative terms, 268 LULUCF would provide about a quarter of total emission reductions planned in countries’ (I)NDCs 269 irrespective of the approach to calculating mitigation. 270 Whereas a similar trend of decreasing LULUCF net emissions with full (I)NDCs implementation has 271 been suggested also by other analyses (ref34,37), the absolute level of net emissions differs 272 significantly: e.g., ref37 reports net emissions about 3 GtCO2e/y higher than ours, due to the 273 ‘harmonization’ of different datasets (country projections and (I)NDCs were aligned to historical 274 FAOSTAT data). By contrast, our study is the first so far showing a global picture of country-based 275 LULUCF net emissions that is consistent between historical and projected periods, including 276 discussing the differences with other global datasets and different mitigation perspectives. 277 278 Science can help countries to keep the forest mitigation promise 279 Several studies suggest a theoretical mitigation potential from land use6,35,38 higher than in this 280 analysis, others suggest limits posed by ecological and socio-economic constraints (including land 281 availability)39,40. Irrespective of the potential, in the past UNFCCC negotiations the LULUCF sector 282 has often been treated separately and considered as a secondary mitigation option, largely due to its 283 complexity and limited trust in data. 12 284 Our analysis shows a wide range of future LULUCF net emissions, depending on policy scenarios. 285 Through the implementation of (I)NDCs countries (especially developing ones) expect a key 286 contribution from LULUCF in meeting their (I)NDC targets, with a clear focus on forests. Achieving 287 this will require increasing the credibility of LULUCF mitigation, through more transparency in 288 commitments and more confidence in estimates. To this regard, the Paris Agreement includes a 289 “Framework for transparency of actions”, key for its credibility41, aimed at providing clarity on GHG 290 estimates and tracking of progress toward achieving countries’ individual targets. 291 More transparent commitments means that future updates of the NDCs should provide more details 292 on how LULUCF mitigation is calculated towards meeting the target and how the financially- 293 supported REDD+ activities contribute to the pledges. More confidence in LULUCF estimates will 294 require improving the country GHG inventories in terms of transparency, accuracy (including 295 information on uncertainties), consistency, completeness and comparability42, especially in 296 developing countries. 297 This is a challenge and an opportunity for the scientific community. Supporting country GHG 298 estimation includes regular reviews of the latest science (e.g. ref43), expanding the scope of the 299 operational methods in the IPCC guidance, as has been done for REDD+44, and incorporating 300 opportunities offered by emerging satellite data45,46 available through highly accessible products47. 301 More confidence also requires independent checks of the transparency and reliability of data, e.g. by 302 reproducing and verifying countries’ GHG estimates. According to IPCC guidance25, verification of 303 GHG inventories is key to improve scientific understanding and to build confidence on GHG 304 estimates and their trends. This can be achieved by comparing GHG inventories with scientific 305 studies using partially or totally independent datasets and/or different methods (e.g. ref48), including 306 greater integration of modeling and measurement systems of land use-related net emissions9. 307 Meaningful verification requires improving mutual understanding and cooperation between the 308 scientific community and the developers of national GHG inventories. 13 309 Finally, increasing trust in proposed LULUCF mitigation options will require reconciling the current 310 differences in global LULUCF net emissions between country reports and scientific studies (as 311 reflected in IPCC reports). Among the many possible reasons for these differences30,49, we suggest 312 that what is considered “anthropogenic sink” is key and deserves further analyses. While recognizing 313 differences in scopes among these communities, reconciling differences in estimates is a necessity, as 314 the “Global stocktake”, i.e. the foreseen five-yearly assessment of the collective progress toward 315 achieving the long-term goals of the Paris Agreement, will be based on both country reports and 316 IPCC reports. Without speaking the same language, the “balance between anthropogenic GHG 317 emissions by sources and removals by sinks in the second half of this century"1, needed to reach the 318 ambitious “well-below 2oC” target, cannot be properly assessed. 319 320 321 Correspondence and requests for materials: giacomo.grassi@ec.europa.eu 322 Disclaimer: The views expressed are purely those of the writers and may not in any circumstances be 323 regarded as stating an official position of the European Commission. 324 Author Contributions G.G. conceived the analysis on (I)NDCs, executed the calculations and 325 drafted the paper. J.H., F.D., M.d.E. and J.P. contributed to the analysis and to the writing of the paper. 326 S.F. provided data from FAO FRA-2015 and contributed to the analysis. J.H. was supported by 327 Leverhulme Foundation and EU FP7 through project LUC4C (GA603542). 328 Competing financial interests. The authors declare no competing financial interests. 329 14 330 References 331 332 1 333 334 UNFCCC. Adoption of the Paris Agreement. Report No. FCCC/CP/2015/L.9/Rev.1, http://unfccc.int/resource/docs/2015/cop21/eng/l09r01.pdf. (UNFCCC, 2015). 2 European Commission, Joint Research Centre (JRC)/Netherlands Environmental Assessment 335 Agency (PBL). Emission Database for Global Atmospheric Research (EDGAR), release 336 version 4.3.1 http://edgar.jrc.ec.europa.eu/overview.php?v=431, 2016. 337 3 UNFCCC. INDCs as communicated by Parties, 338 http://www4.unfccc.int/submissions/indc/Submission%20Pages/submissions.aspx. 339 (UNFCCC, 2015). 340 4 341 342 UNFCCC. Paris Agreement - Status of Ratification. http://unfccc.int/paris_agreement/items/9444.php (UNFCCC 2016) 5 Ciais, P. et al. in Climate Change 2013: The Physical Science Basis. Contribution of Working 343 Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change 344 (eds T.F. Stocker et al.) Ch. 6, 465-522 (Cambridge University Press, 2013). 345 6 Smith, P. et al. in Climate change 2014: mitigation of climate change. Contribution of 346 Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate 347 Change (eds O. Edenhofer et al.) Ch. 11, 811-886 (Cambridge University Press, 2014). 348 7 349 350 Tubiello, F. N. et al. The Contribution of Agriculture, Forestry and other Land Use activities to Global Warming, 1990-2012. Global Change Biology 21, 2655-2660 (2015). 8 Federici, S., Tubiello, F. N., Salvatore, M., Jacobs, H. & Schmidhuber, J. New estimates of 351 CO2 forest emissions and removals: 1990-2015. Forest Ecology and Management 352, 89-98 352 (2015). 353 9 354 355 356 Le Quéré, C. et al. Global Carbon Budget 2015. Earth System Science Data 7, 349-396 (2015). 10 Pan, Y. et al. A large and persistent carbon sink in the world's forests. Science 333, 988-993 (2011). 15 357 11 Schlamadinger, B. et al. Options for including land use in a climate agreement post-2012: 358 improving the Kyoto Protocol approach. Environmental Science and Policy 10, 295-305 359 (2007). 360 12 UNFCCC. Decision 2/CP.17. Outcome of the work of the Ad Hoc Working Group on Long- 361 term Cooperative Action under the Convention, FCCC/CP/2011/9/Add.1, 362 http://unfccc.int/resource/docs/2011/cop17/eng/09a01.pdf (UNFCCC, 2011). 363 13 UNFCCC. Greenhouse Gas Inventories, 364 http://unfccc.int/national_reports/annex_i_ghg_inventories/national_inventories_submissions/ 365 items/8812.php. (UNFCCC, 2015). 366 14 Grassi, G., den Elzen, M. G. J., Hof, A. F., Pilli, R. & Federici, S. The role of the land use, 367 land use change and forestry sector in achieving Annex I reduction pledges. Climatic Change, 368 1-9 (2012). 369 15 370 371 annexinatcom/submittednatcom/items/653.php (UNFCCC, 2015). 16 372 373 UNFCCC. Biennial Update Reports, http://unfccc.int/national_reports/nonannex_i_natcom/reporting_on_climate_change/items/8722.php (UNFCCC, 2015). 17 374 375 UNFCCC. National Communications Non-Annex 1, http://unfccc.int/nationalreports/non- WRI. CAIT Climate Data Explorer. Paris Contributions Map, http://cait.wri.org/indc/. (World Resources Institute, 2015). 18 Admiraal, A. et al. Assessing Intended Nationally Determined Contributions to the Paris 376 climate agreement – what are the projected global and national emission levels for 2025– 377 2030? Report No. PBL 1879, http://www.pbl.nl/en/publications/assessing-intended- 378 nationally-determined-contributions-to-the-paris-climate-agreement. (PBL, 2015). 379 19 2 °C. Nature 534, 631-639, doi:10.1038/nature18307 (2016). 380 381 Rogelj, J. et al. Paris Agreement climate proposals need a boost to keep warming well below 20 UNFCCC. National Communications Annex 1, 382 http://unfccc.int/nationalreports/annexinatcom/submittednatcom/items/7742.php; (UNFCCC, 383 2015). 16 384 21 UNFCCC. Compilation of information on nationally appropriate mitigation actions to be 385 implemented by Parties not included in Annex I to the Convention, 386 FCCC/AWGLCA/2011/INF.1, http://unfccc.int/resource/docs/2011/awglca14/eng/inf01.pdf. 387 (UNFCCC, 2011). 388 22 FAO-FRA. Country reports (2015), http://www.fao.org/forest-resources-assessment/current- 389 assessment/country-reports/en/. (Food and Agricultural Organization of the United Nations, 390 2015). 391 23 392 393 FAOSTAT. Land use emissions. (Food and Agricultural Organization of the United Nations (FAO), Rome, Italy, http://faostat3.fao.org/download/G2/*/E, 2015). 24 Grassi, G., Monni, S., Federici, S., Achard, F., and Mollicone, D. Applying the 394 conservativeness principle to REDD to deal with the uncertainties of the estimates. 395 Environmental Research Letters, 3(3) (2008) 396 25 IPCC. in 2006 IPCC Guidelines for National Greenhouse Gas Inventories (eds H.S. 397 Eggleston et al.) (National Greenhouse Gas Inventories Programme, Institute for Global 398 Environmental Strategies, Hayama, Japan, 2006). 399 26 400 401 Penman et al.) (Institute for Global Environmental Strategies, 2003). 27 402 403 Houghton, R. A. et al. Carbon emissions from land use and land-cover change. Biogeosciences 9, 5125-5142 (2012). 28 404 405 IPCC. in Good Practice Guidance for Land Use, Land-Use Change and Forestry (eds J. Hansis, E., Davis, S. J. & Pongratz, J. Relevance of methodological choices for accounting of land use change carbon fluxes. Global Biogeochemical Cycles 29, 1230-1246 (2015). 29 Pongratz, J., Reick, C. H., Houghton, R. A. & House, J. I. Terminology as a key uncertainty 406 in net land use and land cover change carbon flux estimates. Earth System Dynamics 5, 177- 407 195 (2014). 408 409 30 Federici, S. et al. GHG fluxes from forests: An assessment of national reporting and independent science in the context of the Paris Agreement, 17 410 http://www.climateandlandusealliance.org/reports/ghg-fluxes-from-forests/ (Climate and 411 Land Use Alliance, 2016). 412 31 IPCC. in IPCC Meeting on Current Scientific Understanding of the Processes Affecting 413 Terrestrial Carbon Stocks and Human Influences on them (eds D. Schimel & M. Manning) 414 (National Oceanic & Atmospheric Administration, 2003). 415 32 IPCC. in Revisiting the use of managed land as a proxy for estimating national 416 anthropogenic emissions and removals (eds S Eggleston, N Srivastava, K Tanabe, & J 417 Baasansuren) (IGES, 2010). 418 33 419 420 Erb, K. H. et al. Bias in the attribution of forest carbon sinks. Nature Climate Change 3, 854856 (2013). 34 UNFCCC. Synthesis report on the aggregate effect of the intended nationally determined 421 contributions, FCCC/CP/2015/7, http://unfccc.int/resource/docs/2015/cop21/eng/07.pdf 422 (UNFCCC, 2015). 423 35 UNEP. The Emissions Gap Report 2015, 424 http://uneplive.unep.org/media/docs/theme/13/EGR_2015_301115_lores.pdf. 1-98 (United 425 Nations Environment Programme, 2015). 426 36 427 428 Change 4, 27-35 (2014). 37 429 430 38 Houghton, R. A., Byers, B. & Nassikas, A. A. A role for tropical forests in stabilizing atmospheric CO2. Nature Climate Change 5, 1022-1023 (2015). 39 433 434 Forsell, N. et al. Assessing the INDCs’ land use, land use change, and forest emission projections. Carbon Balance Manage 11:26, doi: 10.1186/s13021-016-0068-3 (2016). 431 432 Lapola, D. M. et al. Pervasive transition of the Brazilian land-use system. Nature Clim. Mackey, B. et al. Untangling the confusion around land carbon science and climate change mitigation policy. Nature Climate Change 3, 552-557 (2013). 40 Clarke, L. et al. in Climate Change 2014: Mitigation of Climate Change. Contribution of 435 Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate 436 Change (eds O. Edenhofer et al.) Ch. 6, 413-510 (Cambridge University Press, 2014). 18 437 41 Tollefson, J. Paris climate deal hinges on better carbon accountancy. Nature 529, 450 (2016). 438 42 UNFCCC. Guidelines for the preparation of national communications from Parties not 439 included in Annex I to the Convention, Bonn, Decision 17/CP.8, 440 http://unfccc.int/resource/docs/cop8/07a02.pdf (UNFCCC, 2003). 441 43 GOFC-GOLD. A sourcebook of methods and procedures for monitoring and reporting 442 anthropogenic greenhouse gas emissions and removals associated with deforestation, gains 443 and losses of carbon stocks in forests remaining forests, and forestation, GOFC-GOLD 444 Report version COP21-1,University of Wageningen, the Netherlands, 445 http://www.gofcgold.wur.nl/redd/. (GOFC-GOLD, 2015). 446 44 GFOI. Integrating remote-sensing and ground-based observations for estimation of emissions 447 and removals of greenhouse gases in forests: methods and guidance from the Global Forest 448 Observations Initiative. (Global Forest Observations Initiative, 2014). 449 45 450 451 Hansen, M. C. et al. High-resolution global maps of 21st-century forest cover change. Science 342, 850-853 (2013). 46 Achard, F. & House, J. I. Reporting carbon losses from tropical deforestation with Pan- 452 tropical biomass maps. Environmental Research Letters 10, doi:10.1088/1748- 453 9326/10/10/101002 (2015). 454 47 Global Forest Watch http://www.globalforestwatch.org/. (2016). 455 48 Pilli, R., Grassi, G., Kurz, W.A., Abad Viñas, R., Guerrero, N. Modelling forest carbon stock 456 changes as affected by harvest and natural disturbances. I. Comparison with countries’ 457 estimates for forest management. Carbon Balance Manage, 11:5, doi: 10.1186/s13021-016- 458 0047-8 (2016) 459 49 Roman-Cuesta, R.M. et al. Multi-gas and multi-source comparisons of six land use emission 460 datasets and AFOLU estimates in the Fifth Assessment Report, for the tropics for 2000–2005. 461 Biogeosciences, 13, 5799–5819 (2016). 462 50 Friedlingstein, P. et al. Update on CO2 emissions. Nature Geoscience 3, 811-812 (2010). 19 463 51 IPCC. in Climate Change 2013: The Physical Science Basis. Contribution of Working Group 464 I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (eds 465 T.F. Stocker et al.) Ch. SPM, 1-29 (Cambridge University Press, 2013). 466 20 467 METHODS 468 469 This analysis quantifies the mitigation role of Land Use, Land Use Change and Forestry (LULUCF, 470 mainly forests), based on the (I)NDCs3,4 submitted by Parties in the context of the Paris Climate 471 Agreement1, complemented with information from other countries’ official documents. This analysis 472 does not aim to assess specific country policies or the quality of country data in comparison with 473 independent sources. 474 Our analysis of LULUCF net emissions over time covered all 195 UNFCCC countries, with 475 assumptions necessary in some cases (i.e. using the latest historical data where no (I)NDC projection 476 was available, see below). However, due to constraints, our estimation of the LULUCF mitigation 477 contribution was possible only for 68 countries (41 (I)NDCs), covering 83% of global forest area 478 (based of FAO-FRA 201522). Other countries were not included either because LULUCF was not 479 clearly included in the target or because the LULUCF contribution was not entirely clear or directly 480 quantifiable (see Supplementary Section 1, Supplementary Information). 481 Our analysis is based on countries’ documents submitted up to February 2016. However, the most 482 relevant recalculations made by countries after that date and before December 2016 (e.g. Brazil, 483 Indonesia and USA) are briefly discussed in the Supplementary Section 2. 484 485 486 Information used in this analysis 487 The methodological approach applied in this analysis required collecting information on: 488 (i) Countries’ historical data and projections up to 2030 (for all 195 UNFCCC countries), 489 using countries’ documents submitted up to end of February 2016, with the following 490 priority: (I)NDCs3; other country data submitted to UNFCCC (2015 GHG Inventories13 21 491 (GHGI) 492 Communications15,20 (NC) and in Biennial Update Reports16 (BUR) for developing 493 countries); other official countries’ documents (e.g. ref.21); FAO-based datasets (for 494 forests8,22 and non-forest emissions23). Despite gaps in country reports (especially for 495 non-forest land uses in developing countries), this priority is justified by the fact that 496 country reports to UNFCCC are formally reviewed or technically assessed by UNFCCC 497 (GHGIs of developed countries are formally reviewed annually, with biennial technical 498 assessment for developing country inventories), and are the means by which countries 499 assess their progress towards targets. Furthermore, FAO-FRA reports22 are not primarily 500 for reporting CO2 emissions and removals, while UNFCCC country reports specifically 501 address emissions and removals. The range of historical country datasets (dotted line in 502 Fig. 1) reflects alternative selections of country sources, i.e. GHGIs for developed 503 countries (selected for both the lower and the upper range), plus FAO-based datasets 504 (upper range) or NCs (lower range) for developing countries. This alternative selection 505 assumes a high reliability of GHGIs for developed countries, while providing an idea of 506 the impact of choosing only NCs (including old NCs) vs. FAO-based datasets for 507 developing countries. See Supplementary Table 4 for an overview of historical datasets 508 used. 509 For historical data, GHGIs with a time series from 1990 to 2013 were available for all 510 developed countries, in most cases including Harvested Wood Products. For developing 511 countries, data are from BURs when available or from latest NCs, typically not including 512 Harvested Wood Products. When only few years were available (typically at least two 513 between 1990 and 2010), 5 or 10 years averages were used. Sometimes, especially for 514 older NCs, data from NCs contain ambiguities, or are not fully comparable across 515 countries (e.g. while most countries implicitly report only emissions and removals from 516 “managed forests”, in accordance with the recent IPCC guidance, a few countries include 517 the sink from apparently unmanaged forests). To reduce the risk of using old or for developed counties, and GHGIs included in recent National 22 518 inappropriate data, the more recent FAO datasets were used instead of NCs prior to 2010. 519 Net emissions from forests (e.g., sink from forest management and emissions from 520 deforestation) usually dominate the LULUCF fluxes in country reports, although in some 521 case emissions from croplands and grasslands (rarely reported by developing countries) 522 are also relevant, especially from organic soils. 523 Based on available information from countries’ reports to UNFCCC complemented by 524 expert judgment, we performed an analysis of the uncertainties for LULUCF absolute 525 GHG net emissions (level) and for the associated trends (see Supplementary Section 3, 526 Supplementary Table 7 and Supplementary Figures 2 and 3). 527 The FAO-based datasets include country data on forest carbon stock change from the 528 Forest Resource Assessment (FAO-FRA 201522, as elaborated by ref8) and FAOSTAT23 529 data on country-level non-forest land use emissions (CO2, CH4 and N2O from biomass 530 fires, including peatlands fires, and from drainage of organic soils). The overall small 531 difference between the FAO-FRA 2015 forest carbon stock data used in our analysis 532 (based on ref.8) and the FRA-2015 forest carbon stock data included in FAOSTAT23 is 533 that the gap-filling methods differ (although for the biomass pools such difference does 534 not impact the total net CO2 emissions/removals across the time series), and that we 535 include both living biomass (above and below-ground) and dead organic matter if 536 reported by countries, while FAOSTAT only considers living biomass. Overall, for the 537 historical period we only used FAO-based datasets to fill the gaps for a relative large 538 number (60), but typically rather small developing countries (covering 11% of global 539 forest area). The significant difference between this analysis and FAOSTAT (Fig. 2 of the 540 paper) is due to several factors, including higher non-forest land use emissions in 541 FAOSTAT for developing countries (especially in Indonesia, Sudan, Zambia) and higher 542 forest land use emissions in FAOSTAT for both developing countries (e.g. Colombia, 543 Liberia, Madagascar, Myanmar, Nigeria, Philippines, Zimbabwe) and developed ones 544 (USA and Russia). 23 545 For projections, data from (I)NDCs (with some expert-judgment interpretation when 546 needed), or NCs20 were available for almost all developed countries. For developing 547 countries, if no projection was available in the (I)NDCs, BURs or NCs, FAO-FRA 2015 548 country projections8,22 were used in few cases. Where no projection was available, the 549 latest historical country data available were used (i.e. continuing the recent estimates). 550 While almost no country provided formal information on uncertainties in their 551 projections, in the analysis of uncertainties (see Supplementary Section 3) we 552 conservatively assumed that the uncertainties estimated for the past will hold for the 553 future. In addition, the different scenarios that our analysis identified up to 2030 (Fig. 1) 554 may also give an order of magnitude of the uncertainties. The range “LULUCF 555 projections min-max” shown in Fig. 1 is slightly broader than the various scenarios (by 556 about 500 MtCO2e/y, or 0.5 GtCO2e/y, in 2030) because in few cases countries provide a 557 range of projections and not all the various projections can be associated with the four 558 scenarios analyzed. The overall difference of about 500 MtCO2e/y is essentially due to 559 the range of projections from the US (the difference between the “high” and a “low” 560 sequestration scenario in their latest National Communication amounts to 370 MtCO2e/y 561 in 2030), and due to Russia (the difference between the various sequestration scenario in 562 their latest National Communication amounts to about 150 MtCO2e/y in 2030). 563 With regards to the GHGs considered, this paper focuses on CO2 emissions and removals 564 and on available data on non-CO2 emissions (CH4 and N2O), based on the information 565 included in countries’ documents. National GHGIs are required to report on all GHGs, 566 but in some developing countries the information on non-CO2 gases is incomplete. Based 567 on available information, and excluding agricultural emissions, the importance of non- 568 CO2 gases is typically small relative to the total GHG fluxes (see ref30 for details), 569 representing about 2-3% of total CO2-equivalent forest flux, with slightly higher values 570 found where forest fires are important in the overall GHG budget. This suggests that, 571 when comparing different datasets (Fig. 2), the possible different coverage in the 24 572 (I)NDCs and other documents of non-CO2 gases does not represent a major reason for 573 discrepancy. 574 (ii) Type of mitigation target elaborated in each countries’ (I)NDC (Supplementary Table 1), 575 i.e. change in absolute emissions or intensity, either relative to a base year or to a BAU 576 scenario (i.e. 2025 or 2030 scenario year); target ‘unconditional’ or ‘conditional’ (i.e. 577 related to the provision of finance, technology or capacity-building support). (I)NDCs 578 expressing only ‘policies and measures’ (without quantitative targets) were not taken into 579 account. 580 (iii) Modality of inclusion of LULUCF within each countries’ (I)NDC (Supplementary Table 581 1), i.e. it may be treated in the same way as other sectors (fully included as part of the 582 overall target), or partially included (only forest activities), or considered separately with 583 special mitigation actions and/or accounting rules. 584 Some additional expert evaluation was included where necessary. 585 586 (I)NDC cases 587 The (I)NDCs were classified into four ‘(I)NDC cases’ (Supplementary Table 2). Based on the 588 availability of country LULUCF information, enough information was found to assign 68 countries to 589 these different “(I)NDC cases”, and to quantify directly the expected LULUCF mitigation. These 68 590 countries include all countries with a major forest coverage and correspond to 78% of global 591 emissions in 2012 (including LULUCF emissions and international aviation and marine emissions)2. 592 593 Different mitigation perspectives 594 The quantification of the mitigation role of LULUCF has been undertaken using different approaches, 595 reflecting different perspectives, according to the questions addressed (Supplementary Table 3). 25 596 597 Estimation of LULUCF mitigation 598 Whereas estimates for perspective ‘A’ (LULUCF net emissions over time) could be made for all 195 599 UNFCCC countries, the information needed for the LULUCF mitigation contribution under 600 perspectives ‘B’ ((I)NDC compared to alternative future scenarios) and ‘C’ (country perspective on 601 calculating emissions reduction (I)NDC) was available only for the 68 countries (41 (I)NDCs) 602 included in Supplementary Table 1. For the remaining countries, the additional mitigation in 603 perspectives ‘B’ and ‘C’ were assumed to be zero relative to other sectors. This assumption is 604 probably conservative (see Supplementary Section 1). 605 Based on the four (I)NDC cases (Supplementary Table 2), and using the available country 606 information (generally with limited expert judgment), this analysis quantified the LULUCF mitigation 607 perspectives (Supplementary Table 3) following the method illustrated in Supplementary Fig. 1. In 608 the very few cases where the target is expressed for 2025, we assumed that the same target applies to 609 2030, allowing us to sum up all the countries’ contribution to 2030. 610 611 Contribution of the land sector to mitigation activity across all sectors 612 The LULUCF mitigation perspectives ‘B’ and ‘C’ were compared to the expected (I)NDC mitigation 613 efforts across all sectors, for each country and at a global level. The global-level all-sectors ‘pre- 614 (I)NDC’ and ‘(I)NDC unconditional + conditional’ are taken from UNEP35. All-sector emissions at 615 the ‘reference point’ (i.e. base year or BAU scenario for target year 2025 or 2030) are from: (i) 616 countries or (ii) from ref18 (for the BAU estimates for China and India). These two sources of 617 information were sufficient for countries representing 87% of global GHG emission in 2012. 618 Emissions for the remaining countries were approximated by assuming the same ratio of emissions at 619 reference point (i.e. estimates from available sources were multiplied by 100/87). 620 26 621 Comparison of this analysis with IPCC AR5 622 In order to make a meaningful comparison of country data (this analysis) with IPCC AR55,6, we 623 disaggregated country data between “land converted to another land use” and “land remaining under 624 the same land use”. While this disaggregation was directly available in all developed country reports, 625 and was largely available for the most important developing countries (e.g. Brazil, Indonesia, India, 626 China, Mexico), for the remaining developing countries information was generally available only for 627 deforestation. In these cases, unless specified otherwise, the other emissions and removals were 628 assigned to “land remaining under the same land use”. 629 630 Data availability 631 This study is primarily based on countries’ (I)NDCs3,4 and other GHG reports submitted to 632 UNFCCC13,15,16,20,21, complemented by FAO-based datasets8,22,23. A large part of elaborated data used 633 to support our findings are available in the Supplementary Information, including: 634 (i) Country-specific information for 68 countries (41 (I)NDCs), in terms of general features 635 of the (I)NDCs (Supplementary Tables 1 and 2) and of data and sources of information of 636 LULUCF net emissions for the historical period 1990-2010 and for 2030, as expected for 637 unconditional and conditional (I)NDC targets (Supplementary Table 5). 638 (ii) Aggregated information on uncertainties (Supplementary Figures 2 and 3), on LULUCF 639 mitigation perspectives (Supplementary Table 3) and on LULUCF net emissions 640 (Supplementary Table 6). 641 Any other raw or elaborated data used in this study are available from the corresponding author upon 642 request. 643 27