December 2018 Pilot ecosystem account for Indonesian peatlands Sumatra and Kalimantan islands Acknowledgements This report is a first, preliminary pilot ecosystem account for Indonesian peatland. It is prepared in order to test and pilot the System of Environmental Economic Accounting – Experimental Ecosystem Accounting Approach (SEEA-EEA), for a specific, policy relevant ecosystem type, i.e. Indonesian peatlands. The report was supported by SarVision, Wageningen and Wageningen University, the Netherlands. The report was provided under auspices of Statistics Indonesia (Badan Pusat Statistik/BPS), with further technical support provided by the Ministry of Environment and Forestry (MoEFRI/KLHK), the Indonesian Ministry of National Development Planning (BAPPENAS), and the Ministry of Agriculture (MoARI/Kementan). The key technical advisors were Resti Salmayenti, MSc, Dr Elham Sumarga and prof. dr Lars Hein. The project has been facilitated and supported by the World Bank project ‘Wealth Accounting and Valuation of Ecosystem Services project in Indonesia (WAVES Indonesia)’. Disclaimer This report is a pilot of an experimental methodology, i.e. the SEEA EEA. The report is not a statistical publication and is meant to examine how the methodology of the SEEA EEA can potentially be applied to a specific policy relevant ecosystem. The account is based on government data. Additional datasets were used for elements for which such data were not available. A major issue was data availability, in particular for the condition and ecosystem services supply and use accounts. Although it was found that all data required to prepare an accurate, detailed account are available, we could not get access to all data that we needed to develop the account. The single dataset that is most crucial for managing and monitoring peatlands is groundwater level. We had several datasets from various sources and used spatial interpolation to estimate groundwater depths in the peatlands of Sumatra and Kalimantan for the year 2013. This influences the accuracy of our results. Moreover, should better data be made available in the future a more accurate peat ecosystem account can be developed. Summary Peatlands cover around 7.8% of Indonesia and support an important agricultural sector. However, activities in peatlands lead to various environmental impacts in particular high carbon emissions as well as, in drained, degraded peatlands, fire and smog formation with associated health impacts. In addition, over time, agricultural activities cannot be maintained because of soil subsidence and subsequent flood risks. Therefore, it is very important that a science-based information system to support peat management is established. The SEEA framework is a comprehensive statistical system for monitoring and analysing environmental information. In order to monitor the changes of peat ecosystems and economic activities concerning their physical and monetary values, this study applies the System of Environmental Economic Accounting – Experimental Ecosystem Accounting (SEEA EEA) framework to develop pilot ecosystem accounts of peatlands in Indonesia. The ecosystem accounting framework comprises a set of connected accounts, dealing with land and ecosystem use (the extent account), the state or health of the ecosystem (the condition account), the supply of ecosystem services including crops and forestry products as well as regulating and cultural services (the physical and monetary ecosystem services accounts) and the monetary ecosystem asset account (depicting the monetary value of the ecosystems). Thematic accounts part of the SEEA EEA include the carbon, biodiversity and water accounts. The SEEA EEA follows a similar approach to monetary analysis as the national accounts, i.e. focusing on measuring the value of ecosystems to economic production and household consumption. It is important that monetary values included in the SEEA EEA do not indicate ‘total economic values’ (in line with the valuation approach of the System of National Accounts on which SEEA is based). Also, externalities are not explicitly covered. Hence, although the peat ecosystem accounts provide useful information to support policy making, the current version of the accounts is not complete and cannot be the only basis for decision making. The report presents an ecosystem account for peatlands in the Indonesian islands of Sumatra and Kalimantan. The study applies the SEEA EEA framework to develop and monitor the changes of peat ecosystems and economic activities concerning their physical and, to some degree, monetary values in Indonesia. The SEEA EEA is a comprehensive system to analyse and report environmental information. The ecosystem accounting framework comprises a set of connected accounts, dealing with land and ecosystem use (the extent account), the state or health of the ecosystem (the condition account), the supply of ecosystem services including crops and forestry products as well as regulating and cultural services (the physical and monetary ecosystem services accounts) and the monetary ecosystem asset account (depicting the monetary value of the ecosystems. It needs to be noted that the peat account is still incomplete. Several important ecosystem services are missing (e.g. flood occurrence, the control of fires in undrained peatlands, hydrological services, and the supply of non-timber forest products). Also there was a shortage of data (in particular for recent years) on drainage levels in the peat. Hence further work on the peat account is required before it can act as a comprehensive basis for decision making – at this point in time information from the account needs to be considered jointly with other datasets (such as the aforementioned ones on hydrology and fire impacts). By integrating and reporting on various aspects of environment and human activities in peat ecosystem, the pilot ecosystem accounts can support a range of policies in Indonesia. First, the accounts can support the rehabilitation of peatland. Targets are set to restore approximately 2.5 Mha degraded peatland by 2020. Pilot ecosystem accounts for Indonesian peatland can, first, facilitate identifying areas that can be considered a priority for rehabilitation. Second, the pilot ecosystem accounts can support monitoring of the physical and monetary 2 impacts of peat rehabilitation. Third, the results of the carbon account can support the National Action Plan to Reduce Greenhouse Gas (GHG) Emissions, by monitoring carbon emissions from peatland. Another relevant application is to support the payment for ecosystem services (PES) in Indonesia by identifying areas of specific relevance for ecosystem services supply, as well as by providing information on co-benefits of payment schemes. The peatland account is based on the Indonesian governments (MoEFRI or KLHK) land cover map. Based on the SEEA EEA, four individual ecosystem accounts are compiled for several years in the period 1990-2015. It should be noted that this is a pilot study, and the account is not yet complete. There is still a lack of data, such as drainage and production data (e.g. on non-timber forest products), and extrapolation was required to fill data gaps. Information is missing on the impacts of fires and smog on people’s health. Nevertheless, the current account presents a first indication of how a peat account for Indonesia can be developed using various valuation techniques and approaches. There is room to improve the quality of data and to increase the number of services (including water) and monetary values (e.g. of health impacts). Indonesian peatland cover approximately 8% of Indonesia’s land surface and support an important agricultural sector. Peatland are important for the cultivation of oil palm, one of the main agricultural commodities currently produced in Indonesia. Other provisions important to Indonesia’s economy include timber and paddy production, and biomass production for pulp. Yet given the increasing scarcity of unused land, the pressure to convert peatland to cropland or plantation forestry areas are still expanding. It is found that 52% of peat forests in Kalimantan and Sumatra have been converted to other land covers from 1990 up to 2015 according to government data for this period. In both Sumatra and Kalimantan plantation areas and agricultural lands expanded drastically during 1990 to 2015. This led to increases in the production of plantation crops such as oil palm fruit, rubber and acacia. However, activities in peatland also lead to various environmental impacts such as high carbon emissions, degraded peatlands, fire and smog formation with associated health impacts. In addition, over time, agricultural activities cannot be maintained because of soil subsidence in drained peatlands and subsequent flood risks. Oil palm plantation areas have expanded significantly and generated the highest monetary value in 2015, meanwhile timber production, CO2 sequestration, and protected land decreased over time. This was established through an ecosystem services account that tracked six main ecosystem services provided by Indonesian peatland, including the production of oil palm, biomass for pulp, paddy, timber, CO2 sequestration, and protected land as biodiversity habitat. However in economic analyses of land use options in peatlands also externalities (such as health effects of peat fires and CO2 emissions) and the long term forecasts of production need to be considered. The current and future increases in flood occurrence in peatlands due to soil subsidence are not yet included in the accounts, and this is a priority for further work so the peat accounts can more meaningfully be used to advice policy makers. Forest conversion and other land use changes in peatland lead to the decrease of carbon stored in vegetation. This was determined through the carbon account, used to monitor the change in carbon stocks and emissions (based on net carbon flux and peat fires) from peatlands. Around 31% of above ground carbon stocks in 1990 was lost by 2015 in Indonesia. Meanwhile, the total emissions from net carbon (CO2) flux increased by 74% during the same period. Additionally, large parts of peatland were burned every year, resulting in more carbon emissions. It needs to be noted that there is a degree of uncertainty in these estimates. For example, the land use mapping conducted for the extent account shows that there is a possibility that the government of Indonesia underestimates the amount of land already converted to plantations and 3 overestimates remaining forest cover. More accurate data on land use, fire and subsidence rates can lead to more accurate estimates of carbon flows. This is a pilot study, and the account is not yet complete. There is still a lack of data, such as drainage and production data (e.g. on non-timber forest products), and extrapolation was required to fill data gaps. Information is missing on the impacts of fires and smog on people’s health. Nevertheless, the current account presents a first indication of how a peat account for Indonesia can be developed using various valuation techniques and approaches. There is room to improve the quality of data and to increase the number of services (including water) and monetary values (e.g. of health impacts). The peat accounts present a useful partial (and incomplete) basis to support monitoring and policy making on Indonesian peatlands. It is noteworthy that this report only shows relatively coarse, aggregated maps, however all maps are available at a fine resolution so that scaling down to individual provinces or potentially districts is possible. The summary table below provides a synthesis of the accounts. Summary table for peat ecosystems (all data are conform government data) Year Peatland Indicator Unit 1990 1996 2000 2006 2009 2013 2014 Ecosystem extent Sumatra Undisturbed forest 1000ha 481 450 378 402 281 225 Disturbed forest 1000ha 4159 3824 2659 2081 1642 1257 Water 1000ha 5 5 5 4 4 4 Degraded peatlanda 1000ha 768 829 1447 1468 1720 1394 Bare ground 1000ha 33 96 213 466 355 380 Urban 1000ha 30 31 35 35 35 34 Forest plantation 1000ha 7 32 48 262 420 864 Perennial crops 1000ha 378 535 941 1007 1211 1398 Dry agricultural land 1000ha 317 365 422 421 479 612 Paddy field 1000ha 192 202 213 214 215 192 Others 1000ha 7 7 16 17 15 15 Kalimantan Undisturbed forest 1000ha 113 80 68 62 58 50 Disturbed forest 1000ha 3790 3234 2978 2799 2565 2308 Water 1000ha 6 5 5 5 5 6 Degraded peatlanda 1000ha 589 1083 1335 1432 1500 1532 Bare ground 1000ha 27 44 44 62 94 203 Urban 1000ha 59 73 83 131 256 336 Forest plantation 1000ha 0 1 0 0 1 300 Perennial crops 1000ha 59 73 83 131 256 336 Dry agricultural land 1000ha 231 243 248 268 270 284 Paddy field 1000ha 68 113 113 115 126 126 Others 1000ha 1 1 1 1 2 3 Ecosystem condition Sumatra Dry biomassb Mt 1475 1409 1170 1079 991 965 Water levelc cm 0-117 Hotspotsd Total pixel 4035 2448 5663 Kalimantan Dry biomassb Mt 1148 1015 959 928 887 835 Water levelc cm 0-96 Hotspotsd Total pixel 3879. 2619. 2635 Ecosystem services (physical values) Sumatra Timber production* 1000m3 1893 1482 1094 777 Oil palm production 1000t 10389 16837 20242 23635 Biomass production 1000t 1011 5503 8833 18161 for pulp* Paddy production* 1000t 620 625 627 561 CO2 sequestration* 1000tCO2 7175 7629 5337 4282 Protected habitat* 1000ha 442 451 423 416 Kalimantan Timber production* 1000m3 794 741 666 576 Oil palm production 1000t 14 2185 4282 8022 4 Biomass production 1000t 0 2 24 624 for pulp* Paddy production * 1000t 192 196 214 214 CO2 sequestration* 1000tCO2 1299 1182 1099 958 Protected habitat* 1000ha - - - - Ecosystem services (monetary values) (IDR billion/year) Sumatra Timber production* 1278 1001 739 525 Oil palm production 1764 2858 3436 4012 Biomass production 1709 for pulp* 95 518 831 Paddy production* 1510 1522 1526 1365 CO2 sequestration* 2498 2656 1858 1491 Protected habitat* 5238 5351 5015 4929 Kalimantan Timber production* 536 500 450 389 Oil palm production 1 88 173 324 Biomass production for pulp* 0 0 2 59 Paddy production* 338 344 375 376 CO2 sequestration* 452 412 383 334 Protected habitat - - - - Carbon Sumatra Carbon stocksb 1000t 2707 2585 2148 1980 1819 1770 CO2 emissions 1000tCO2 131 146 178 195 225 272 (oxidation)e CO2 emissions (fire)f 1000ha 318 183 286 Kalimantan Carbon stocksb 1000m3 2107 1862 1759 1702 1628 1533 CO2 emissions 1000t 91 94 95 99 108 115 (oxidation)e CO2 emissions (fire)f 1000t 386 325 324 Note: *ES based on land cover data, oil palm production was estimated from several sources. a :land covers of wet shrub, dry shrub, savanna and grasses and open swamp, b: vegetation, c: the values are displayed in maps (see details in the chapter of results, d: 1-km fire pixel based on MODIS fire product, calculated for the minimum confidence value at 80%, e:net carbon flux (excluding from peat fires), f: from burned peat (33-cm depth of burned peat for all types of LC) 5 Abbreviations AGB Above-Ground Biomass BGB Below-Ground Biomass CO2 Carbon dioxide ES Ecosystem Services FFB Fresh Fruit Bunches (of oil palm) Ha Hectare LC Land Cover MoARI Ministry of Agriculture Republic of Indonesia MODIS Moderate Resolution Imaging Spectroradiometer MoEFRI Ministry of Environment and Forestry Republic of Indonesia NCA National Capital Accounting SCC Social Cost of Carbon SEEA System of Environmental-Economic Accounting SEEA CF SEEA Central Framework SEEA EEA SEEA Experimental Ecosystem Accounting 6 Table of Contents Acknowledgements.................................................................................................................................................. 1 Disclaimer ................................................................................................................................................................... 1 Abstract ....................................................................................................................Error! Bookmark not defined. Summary table (data are conformed government data) ................................................................................ 4 Abbreviations............................................................................................................................................................. 6 List of Figures .......................................................................................................................................................................... 9 List of Tables ............................................................................................................................................................................ 9 List of Equations................................................................................................................................................................... 10 List of Annexes ...................................................................................................................................................................... 10 Chapter 1. Introduction ........................................................................................................................................11 1.1 Background ....................................................................................................................................................................... 11 1.2 Objective of the report.................................................................................................................................................. 12 1.3 Scope of work ................................................................................................................................................................... 12 Chapter 2. Theoretical framework and indicators.....................................................................................13 2.1 SEEA-EEA framework ................................................................................................................................................... 13 2.2 Accounts and indicators .............................................................................................................................................. 13 2.2.1 Ecosystem extent account ...................................................................................................................... 13 2.2.2 Ecosystem condition account ................................................................................................................ 14 2.2.3 Ecosystem service account ..................................................................................................................... 14 2.2.4 Carbon account............................................................................................................................................ 15 2.3 Study area .......................................................................................................................................................................... 15 Chapter 3. Methodology .......................................................................................................................................16 3.1 Ecosystem extent account........................................................................................................................................... 17 3.2 Ecosystem condition account .................................................................................................................................... 18 3.2.1 Vegetation biomass ................................................................................................................................... 18 3.2.2 Groundwater level ..................................................................................................................................... 18 3.2.3 Hotspots ......................................................................................................................................................... 19 3.3 Ecosystem services account ....................................................................................................................................... 19 3.4 Carbon account ................................................................................................................................................................ 21 3.4.1 Carbon stocks (vegetation) .................................................................................................................... 21 3.4.2 Carbon emissions ....................................................................................................................................... 21 Chapter 4. Results...................................................................................................................................................23 4.1 Ecosystem extent account........................................................................................................................................... 23 4.2 Ecosystem condition account .................................................................................................................................... 27 4.2.1 Vegetation biomass ................................................................................................................................... 27 4.2.2 Groundwater level ..................................................................................................................................... 27 7 4.2.3 Fire hotspots................................................................................................................................................. 28 4.3 Ecosystem service account ......................................................................................................................................... 29 4.4 Carbon account ................................................................................................................................................................ 31 4.4.1 Carbon stocks (vegetation) .................................................................................................................... 31 4.4.2 Carbon emissions ....................................................................................................................................... 32 Chapter 5. Discussions ..........................................................................................................................................36 5.1 Uncertainties and Limitations ................................................................................................................................... 36 5.2 Policy applications ......................................................................................................................................................... 38 Chapter 6. Conclusions .........................................................................................................................................40 References .................................................................................................................................................................42 Annexes ......................................................................................................................................................................46 8 List of Figures Figure 1. Connections between ecosystem and related accounts (Source:20) ........................................ 13 Figure 2. Distribution map of Indonesian peatlands (Source: 41) ............................................................ 16 Figure 3. Spatial distribution of land cover types in Sumatra in 1990 and 2015. .................................. 23 Figure 4. Spatial distribution of land cover types in Kalimantan in 1990 and 2015. ............................. 24 Figure 5. Total biomass (vegetation) in different land cover types of peatland in Sumatra and Kalimantan 1990-2015. .............................................................................................................................. 27 Figure 6. Estimated water level map of Sumatra and Kalimantan peatlands in 2013. .......................... 28 Figure 7. Number of hotspots (with confidence value ≥ 80%) in different land cover types of peatland in Sumatra and Kalimantan for 2006, 2009 and 2014. ............................................................................. 28 Figure 8. Spatial distribution of carbon stocks (vegetation) in Sumatra and Kalimantan peatlands in 1990 and 2014. ............................................................................................................................................ 31 Figure 9. Social cost of carbon of carbon stocks (vegetation) in peatlands of Sumatra and Kalimantan 1990-2014/2015. ........................................................................................................................................ 32 Figure 10. Social cost of carbon of carbon emissions from oxidation in peatlands 1990-2014/2015 . 33 Figure 11. Trend of CO2 emissions and CO2 sequestrations in Sumatra and Kalimantan peatlands 1990-2015 based on land cover types. The bars depict the emission from each land cover type. These emissions are a function of drainage. Note that it is assumed that all plantations and dry agricultural land are drained. Disturbed forests and degraded peatlands only emit CO2 when they are drained. .. 33 Figure 12. Social cost of carbon of carbon emissions from burned peat 1990-2014/2015 .................. 34 Figure 13. Carbon (CO2-eq) emissions from burned peat in several types of peatland covers in 2006, 2009, and 2014. Note that the emissions were from 33-cm burned peat, applied for all types of land cover. ............................................................................................................................................................ 35 Figure 14. Spatial distribution of precipitation in Sumatra and Kalimantan peatlands in 2014. estimated from TRMM (2011). ................................................................................................................... 56 Figure 15. Flow chart of kriging interpolation technique ........................................................................ 58 Figure 16. Selected variogram model for Sumatra with parameters: partial sill 187.35, range 1.9 and nugget 56.39................................................................................................................................................. 59 Figure 17. Selected variogram model for Kalimantan with parameters: partial sill 457.15, range 1.1 and nugget 85.8............................................................................................................................................ 59 List of Tables Table 1. The class description of Indonesian peatland cover ................................................................................. 17 Table 2. The meaning of confidence interval in hotspot information ................................................................. 19 Table 3. Indicators for physical valuation of ES provided by peatland ecosystem in Indonesia ............. 20 Table 4. Indicators for monetary valuation of ES provided by peatland ecosystem in Indonesia .......... 20 Table 5. Extent account of peatlands in Sumatra ......................................................................................................... 25 Table 6. Extent account of peatlands in Kalimantan ................................................................................................... 26 Table 7. Total vegetation biomass in Sumatra and Kalimantan peatlands 1990-2015. .............................. 27 Table 8. Physical values of ecosystem services in Sumatra and Kalimantan peatlands. ............................. 29 Table 9. Monetary values of ecosystem services in Sumatra and Kalimantan peatlands. .......................... 29 Table 10. Total area of oil palm plantation in Sumatra and Kalimantan peatlands in 2000, 2005, 2010, and 2015 ........................................................................................................................................................................................ 30 Table 11. Physical and monetary value of oil palm production in Sumatra and Kalimantan peatlands in 2000, 2005, 2010, and 2015............................................................................................................................................. 30 Table 12. Total carbon (CO2-eq) stocks as vegetation in Sumatra and Kalimantan peatlands 1990- 2015. ................................................................................................................................................................................................ 31 Table 13. Total CO2 emissions in Sumatra and Kalimantan peatlands 1990-2015. ...................................... 32 Table 14. Total burned peatlands and total carbon (CO2-eq) emissions from burned peat in Sumatra and Kalimantan in 2006, 2009, and 2015. ...................................................................................................................... 34 9 Table 15. Total plantation area in peatlands 2014 (MoEFRI and WRI maps) ................................................. 52 Table 16. Total plantation area in peatlands of Central Kalimantan 2012 ....................................................... 53 Table 17. The average of annual precipitation in Sumatra and Kalimantan peatlands. .............................. 55 List of Equations Equation 1 ..................................................................................................................................................................................... 18 Equation 2 ..................................................................................................................................................................................... 21 Equation 3 ..................................................................................................................................................................................... 21 Equation 4 ..................................................................................................................................................................................... 22 Equation 5 ..................................................................................................................................................................................... 55 List of Annexes Annex 1. Estimated carbon stocks as vegetation in several land cover types ................................................. 46 Annex 2. Hotspots ...................................................................................................................................................................... 47 Annex 3. Estimated total area of oil palm plantation in Sumatra and Kalimantan peatlands. ................. 48 Annex 4. Annual production rate of ecosystem services in peatlands of Sumatra and Kalimantan ...... 49 Annex 5. Estimated net carbon flux in several land cover types. .......................................................................... 49 Annex 6. Production cost and market price of commodities .................................................................................. 49 Annex 7. Spatial distribution of ecosystem services (physical values) in Sumatra and Kalimantan peatlands for 2000 and 2015 ................................................................................................................................................ 50 Annex 8. Comparison of total area and spatial distribution of plantation in peatlands .............................. 51 Annex 9. Burned peatlands in Sumatra and Kalimantan for 2006. 2009 and 2015 ...................................... 54 Annex 10. Precipitation ........................................................................................................................................................... 55 Annex 11. Oceanic Niño index 1950-2017 ...................................................................................................................... 57 Annex 12. Estimated emission factor and biomass consumption of burned peat ......................................... 57 Annex 13. Kriging interpolation method ......................................................................................................................... 58 Annex 14. Selected water table depth data ..................................................................................................................... 59 10 Chapter 1. Introduction 1.1 Background Peatlands in Indonesia support several economic activities, in particular agriculture and plantation forestry. Peatlands are important for the cultivation of oil palm, one of the main agricultural commodities produced in Indonesia. Given increasing scarcity of unused mineral land, the pressure to convert peatlands to cropland or plantation forestry areas is still increasing. This has also been facilitated by the decentralisation policies applied as of 1999, that shifted responsibilities for land conversion to lower administrative levels 2-5. However, unsustainable peatland uses are cause of environmental concern6. The tropical peatlands of Indonesia are one of the world’s largest carbon pools, storing around 30–700 tC/ha carbon per every meter of peat soil depth7. Burning peat forest, peat drainage, and fertilisation in plantation and agricultural lands lead to substantial emissions of carbon dioxide (CO2)8-10. Cultivation on peatlands requires drainage, which, in turn, leads to irreversible soil subsidence 11, 12. Over time, this will lead to flooding of these areas, and consequently soil subsidence will severely hamper crop production in peatlands. In particular, subsidence will lead to peatlands becoming relatively low-lying areas in the landscape, where rain and potentially river and or sea water will collect during the rainy season. Many plantation crops including oil palm, and forestry crops such as acacia are highly sensitive to flooding and production cannot be maintained in these areas in the near to medium future. Furthermore, peat drainage and cultivation also have major social/environmental impacts on biodiversity and health issues. The environmental problems in peat ecosystem have drawn the attention of the government. Better management plans to move towards sustainable development including in peatlands are required. In 2011, a Norway-Indonesia partnership focusing on the reduction of GHG emissions was followed by a Presidential Instruction (inpres) No.10/2011 about a two-year suspension of new licences for primary natural forest and peatland clearing. This instruction has succeeded in protection of carbon and biodiversity in 71% or 11.2 Mha of Indonesian highly threatened peatlands 13. In 2014, a Government Regulation (PP) No.71/2014 about peatland protection and ecosystem management was issued to protect 30% of Indonesian hydrological unitary peatlands 14. As a further step, a peat restoration agency (BRG) has been formed by the president of Indonesia with a target to restore approximately 2.5 Mha of degraded peatlands by 2020 (SK.05/BRG/Kpts/2016). This was followed by a Presidential Decree (Perpres) No.1/2016 about restoration priority in seven provinces (12.9 Mha of peatlands in Riau, Jambi, South Sumatra, West Kalimantan, Central Kalimantan, South Kalimantan and Papua provinces). These priority areas also include most severely burned parts in 2015, shallow peat areas with canals (3 Mha), as well as peat domes with canals and without canals (2.8 Mha and 6.2 Mha) 15. A range of studies on Indonesian peatlands have been conducted. Peatland maps for Indonesia, for example, have been published by several research institutions including the governmental agency, the Ministry of Agriculture (MoARI) and BRG. The Ministry of Environment and Forestry (MoEFRI) also continuously publishes Indonesian land cover maps, in which peatlands are included. Furthermore, several research activities have been implemented to analyse the environmental impacts and economic benefits from activities in peatlands, for instance regarding carbon emissions 9, 10, and ecosystem services 6. However, this information needs to be integrated and disseminated in order to effective monitor and report on the changes of environmental and economic conditions in the peat ecosystem and support policy making on peat management. 11 The United Nations System of Environmental-Economic Accounting - Experimental Ecosystem Accounting (SEEA EEA) is a framework for monitoring the interaction between ecosystems and economic activities. The SEEA framework is comprehensive (covering ecosystem extent, condition, services and assets), coherent (aligned with the System of National Accounts) and flexible (can be implemented at different institutional scales or for different ecosystem types). This peat account is based upon the SEEA EEA. It is the world’s first pilot ecosystem account following the SEEA EEA developed for peatlands. The Indonesian government has been applying national capital accounting to support analysing the relation between natural resources and economic development in the last two decades. This is supported by regulation (UU) No.32/2009 about environmental protection and management, and regulation (PP) No.46/2017 about economic instruments for the environment. In line with these regulations Statistics Indonesia (BPS) has developed an integrated environmental and economic balance system, called SISNERLING, to account the timber, mineral and energy at national level. SISNERLING is based upon the SEEA Central Framework (CF). The SEEA CF and the SEEA EEA are complementary information systems, covering different ecosystem- environment dependencies. In addition, the SEEA (EEA) is, contrary to the SNA and the SEEA CF, a spatial account, where maps as well as accounting tables are presented for each account. 1.2 Objective of the report The objective of this study is to develop pilot ecosystem accounts for Indonesian peatlands, following the framework of SEEA EEA. Based on the technical recommendations of SEEA EEA, there are four pilot accounts (ecosystem extent account, ecosystem condition account, ecosystem services account and carbon account) included in this study. Specific indicators for each account are selected in this report, based on technical consistency with the SEEA EEA framework, data availability and policy relevance. The accounts identify the ecosystem in physical (applied for all accounts) and monetary (only applied to ecosystem services) terms. This study establishes pilot ecosystem accounts for Indonesian peatlands based on data mainly from governmental institutions. The accounts integrate different statistics and thereby provide new insights. They can also guide the monitoring process in the future as new data are collected continually. Several types of missing data were collected from literature reviews. The description of data sources and methods used are explained in the methodology. The results are displayed in tables and maps that allow tracking the temporal and spatial changes of the selected indicators. 1.3 Scope of work The scope of this study is the peat area of Indonesia, specifically in Sumatra and Kalimantan, based on peatland map from MoARI (2011). The scope is a soil type of characterization which is fixed in time. There is no additional or reductional size of peatlands. It is noted that there are still uncertainties about peatland area and cover 16-19. Where government data are available, this account exclusively uses government data. For indicators for which there were no government data available other sources have been used. The peat ecosystem accounts are established for following years: 1990, 1996, 2000, 2006, 2009, and 2014. The most recent data for 2016 are not presented in this report due to a lack of access to land cover data. In addition, several indicators are limited to certain years due to data limitations. In particular, there are relatively few data on peat drainage and production rate of agricultural products. Since the account is incomplete (not all services, values and externalities are considered), it needs to be kept in mind that the accounts only provide a partial insight in natural capital provided by peat under different uses. 12 Chapter 2. Theoretical framework and indicators 2.1 SEEA-EEA framework Pilot ecosystem accounts for Indonesian peatlands are developed based on SEEA EEA framework. This framework is based on the System of National Accounts (SNA) with a focus on environmental resources and their interactions with economic (human) activities20. The framework consists of several individual accounts: ecosystem extent, ecosystem condition, ecosystem services, ecosystem asset and thematic (land, water, carbon and biodiversity) accounts. Technical recommendations for SEEA EEA have been published, and these have been applied in developing the peat ecosystem account. Figure 1 illustrates the relationship between the accounts and the terms of valuation. Physical terms are applied to all accounts, and monetary terms are applied only to ecosystem services, ecosystem asset and thematic accounts20. Figure 1. Connections between ecosystem and related accounts (Source:20) Referring to SEEA EEA framework, this study develops four specific pilot peatland ecosystem accounts, covering the ecosystem extent account that explains states and changes of land cover (LCs), the ecosystem condition account that monitors the state of peat ecosystems, the ecosystem service (ES) account which estimates the physical and monetary values of main ecosystem services provided by Indonesian peatlands, and lastly carbon account that tracks the stocks, emissions and sequestration of carbon in peat. For reasons of data availability, only the peatlands in Kalimantan and Sumatra are included in the peat account. 2.2 Accounts and indicators 2.2.1 Ecosystem extent account Ecosystem extent accounting is an initial step that explains the distribution of environmental characteristics and human activities in the ecosystem. In this account, land cover (LC), land use (LU) or ecosystem use are classified. This information facilitates monitoring and reporting the changes in LC over time. This information is also essential to analyse ecosystem services provided by peat, since these services are closely linked to land cover 21, 22. The indicator of this account 13 follows the same LC categories as the land account of Indonesia published recently (see details in the methodology chapter). 2.2.2 Ecosystem condition account The ecosystem condition account focusses on the physical state or condition of the ecosystem. For this account, three indicators (vegetation biomass, water level and hotspots) were selected based on the characteristics of Indonesian peatlands, policy relevance and data availability. Vegetation biomass In addition to carbon in the soil, peatlands also store noteworthy amounts of carbon in the vegetation. The reduction of vegetation density in peat forests by fires, deforestation and land conversion, decreases the carbon content due to biomass loss and peat decomposition 23. This indicator is included in the condition account to track the changes of vegetation biomass in peat ecosystem. Water level Drainage is required to convert peatland into a plantation or agricultural land. For example, a drainage depth of around 50-70 cm is typically needed for oil palm plantation 24, even though drainage can be shallower or deeper in the case of poor water management. The groundwater level varies within a year, and this account includes the annual average groundwater depth. Drainage depth is closely related to CO2 emissions and subsidence and also increases the risk of fire 25-28. Therefore, water level is included in this account. Fire hotspot Forest fires have negative impacts on people’s health, and lead to ecosystem damages and a reduction of agricultural production 29, 30. It was estimated that forest fires in Indonesia, including peatlands, also become a large source of CO2 emissions 10, 23, 31. This study uses the number of fire (hotspots) as an indicator in order to track the temporal and spatial distribution of fire accidents in peatlands using satellite product. 2.2.3 Ecosystem service account The ecosystem service account provides information about the main ecosystem services (ES) generated in peatlands. This account encompasses the values both in physical and monetary terms. The values of ES can be estimated using several valuation techniques, depending on the characteristic of selected indicators and data availability20, 32, 33. Importantly, these indicators should be consistent with the valuation principles of the System of National Accounts, as explained in the SEEA EEA Technical Recommendations. The detailed description of methods used for this account is presented in the methodology chapter. ES (based on MoEFRI land cover data) The estimation of the main ES (physical value) provided by Indonesian peatlands at national level has been conducted using several data sources 6. To maintain the consistency of data collected by the government, firstly this study focuses on ES estimated based on MoEFRI land cover data (published continuously). From diverse ecosystem services provided by Indonesian peatlands, there are five ES that can be identified. These ES are divided into three provisioning services (biomass (acacia) production for pulp, timber production, and paddy production) measured in terms of annual product harvested, carbon sequestration (a regulating service) measured in terms of total carbon (CO2) sequestration in undisturbed forests, and a cultural service measured in terms of total area of protected peatlands (peat swamp forest). The protected area network 14 consists of protected forests and conservation areas (national park, recreation park, nature reserve and wildlife sanctuary) in 2000 which have not been changed to non-forest areas6. Comparing to the previous study 6, this report only selects the CO2 sequestrations instead the CO2 emissions. The carbon emissions are categorised as an ecosystem disservice and explained independently in the carbon account, which is more aligned with the SEEA EEA Technical recommendations. ES: oil palm production Plantation areas in Indonesian peatlands have been expanded to support the economics of this country. However, LC maps used in this study are not able to distinguish the variations of plantation. Diverse types of the plantation that have been widened in Indonesian peatlands such as oil palm, rubber and coconut plantations 6, 34-37. Therefore, this study tries to identify more types of ES using other data sources, which might have inconsistency with data from the LC map. However, this report only presents one additional ecosystem service, oil palm production, due to limitation of data sources. 2.2.4 Carbon account Carbon account is presented to monitor the trend of carbon stocks and emissions in peatland ecosystem. This account is divided into several indicators, including carbon stocks (vegetation), carbon emissions from peat fires as well as oxidation of soils, and a net carbon flux is assessed. 2.3 Study area In this section, the boundaries of the study area are defined. The territory analysed in this research comprises all peatlands of Sumatra and Kalimantan as identified in government publications, which together account for around 75% of Indonesian peatlands. The area is distributed between 60 08’ north latitude and 110 15’south latitude, and between 950 45’ and 1410 05’ of east longitude. Located in the equatorial zone makes Indonesia experiences tropical climate with an average temperature around 24-280C and precipitation ranging from 906-4627 mm/year with three seasonal patterns distributed spatially 39, 40. The population of Indonesia is still increasing, from around 87.8 million in 1960 to 257.6 million in 2015 38. The ecosystem account of this study is the tropical peatland with peatlands in Sumatra and Kalimantan as the specific ecosystem type (ET). According to peatland map from the Centre for Research and Development of Agricultural Land Resources (BBSDLP), the MoARI, Indonesia has 14.9 Mha peatlands (approximately 7.8% of Indonesian land surface) that are mainly spread in seventeen provinces (of thirty-three provinces) in the three biggest islands (Sumatra, Kalimantan and Papua) 16. As seen in Figure 2, Indonesian peatlands are scattered, 43% in Sumatra, 32% in Kalimantan, and 25% in Papua 16. 15 Figure 2. Distribution map of Indonesian peatlands (Source: 41) 2.4 Limitations in the approach The peat account covers a range of important ecosystem services provided by peat, but not all services are as yet included (they could be added in the future should the Government of Indonesia decide to continue the approach). In particular, what is missing are first the health effects of peat fires. The costs of fires are an externality of peat use, related to drainage. They lead to smog that can cover large parts of Indonesia and parts of nearby countries. The account (chapter 4) shows in which land cover types the fires are generated, the area burned, and the amount of CO2 emitted due to the fires, but the health effects are not yet analysed. Second, flooding is not well captured in the accounts yet. Seasonal flooding occurs in peatlands where the surface has subsided. Following several decades of subsidence, peatlands may become the lowest parts in the landscape, and thereby be under water for periods of weeks or months each year. These peatlands can no longer be used for plantations and will be abandoned (this is happening already in parts of South Sumatra). The flood risk and the associated impacts on production are not yet in the accounts. Third, the list of ecosystem services covered in the account is incomplete. For example, the production of non-timber forest products such as rattan, as well as a range of crops such as sago and liberica coffee are missing. Fourth, the accounts use an SNA-consistent monetary valuation approach based on exchange value (aligned with market prices); they do not measure welfare effects of peat use. Hence, the monetary information in the accounts cannot be used for an environmental CBA without further economic analysis. Fifth, the accounts have as one of the key inputs the land cover map of MoEFRI. However, the land and extent accounting report suggests that the plantation area may be underestimated in these maps, and the forest cover could be overestimated. This adds uncertainty to the figures (among others it may lead to an underestimate of the CO2 emissions since drainage levels were modelled based on among others land use). Finally, it is important to report that soil subsidence is an irreversible process. Choices made to drain peatland now will be felt for decades to come, also because restoring peatlands by rewetting them is very expensive. It is not yet clearly shown in this account that over time the value of drained peatlands will convert to zero, since production of plantation crops cannot be maintained when the area is seasonally flooded. An alternative is to use profitable crops in peat that do not require drainage (note that even low drainage levels will lead to irreversible soil subsidence over time), such as jelutung rubber or sago (which can be used for starch-based food products as well as bioplastics). These options are also not yet included in the account (the current production of these so-called paludiculture crops is low and further data are needed to include them). 16 Chapter 3. Methodology This section describes the methods and the data sources that were used to estimate the value of the indicators in the ecosystem extent account (land cover), the ecosystem condition account (vegetation biomass, water level and hotspots), the ecosystem services account (physical and monetary values) and the carbon account, for peatlands in Sumatra and Kalimantan. 3.1 Ecosystem extent account The types of land cover (LC) were classified based on the regulation of director general of forestry planology No. P.1/VII-IPSDH/2015 42 as seen in Table 1. In this study, the land cover of dry shrub, wet shrub, open swamp and savanna are grouped as degraded lands. Following the land account of Indonesia, the ecosystem extent account is developed for the years, 1990, 1996, 2000, 2006, 2009, and 2014. Along with the total area, this account also monitors the additional and the reduction areas of each LC type for every time-period (1990-1996, 1996-2000, 2000-2006, 2006- -2009, and 2010-2014. The results are presented as annual data in tables and maps. It is important to note that the account is based on land cover data from MoEFRI. In this dataset, information on land cover for the year 2010 is mostly based on remote sensing images of 2009; and information for the year 2015 is mostly based on remote sensing images of 2014. Hence, the land cover changes would lead to inaccuracies in this account (since there will in some areas have been further land use conversion from 2009 to 2010 and from 2014 to 2015). This underestimation also occurs in the published government data and may lead to an inaccuracy of several percent in the land cover data (thereby affecting estimates of ecosystem services and carbon stocks and emissions). In order to align the estimates of carbon as close as possible to the actual land cover, the carbon stocks and emissions are calculated for 2009 and 2014. The peatland area is included based on government data, including the peatland map of the MoARI 41 and the land cover maps of the MoEFRI 43. These maps were overlaid to calculate the total area of each LC category using ArcMap 10.5. To measure the total area, projected coordinate system of Asia South Lambert Conformal Conic was selected by modifying the parameters with central meridian at 115, standard parallel 1 at 2, standard parallel 2 at -7, and latitude of origin at 0. Table 1. The class description of Indonesian peatland cover Type of land cover Description Undisturbed forest Primary and well-preserved secondary forest Disturbed forest Secondary (degraded) natural forest Water Open water areas Degraded peatland Bush (dry shrub), shrub swamp (wet shrub), savanna and grasses, and open swamps Bare ground Bare areas Urban Transmigration and settlement areas Forest plantation Forest vegetation in large areas, dominated by homogeneous trees species, and planted for specific purposes, including reforestation area, industrial plantation forest and community plantation forest Perennial crops Estate areas, mostly with perennials crops or other agriculture trees commodities Dry agricultural land Pure and mixed dryland agriculture Paddy field Agricultural land for paddy Other LCs Fish ponds (aquaculture), mining areas, ports and harbours No data Clouds and no-data areas 17 3.2 Ecosystem condition account Following the ecosystem extent account, this study tries to develop ecosystem condition account for the same years. However, the available data are limited for specific years: vegetation biomass for 1990, 1996, 2000, 2006, 2009, and 2014, water level only for 2013, and hotspots data for 2006, 2009, and 2014. 3.2.1 Vegetation biomass The estimated biomass is the sum of dry biomass of tree and dead organic matter as displayed in Equation 1. The measurement of dry tree biomass consists of two parts, aboveground biomass (AGB) and belowground biomass (BGB)44. The biomass of dead dry matter is the sum of litter biomass (L) (such as fallen leaves, fruits and flowers) and biomass of woody debris (WD) (such as dead trees, fallen trees and part of trees like stems, branches, twigs on the ground). Vegetation biomass (tdm) = At × (AGBt + BGBt + Lt + WDt) Equation 1 The calculation was conducted based on the total area of each LC type in the extent account (At). Biomass data from a report of the MoEFRI 45 were applied. However only the estimated AGB values (22 different LC classes) were available. This study presents a more detailed information by adding the values of BGB, L and WB collected from various data sources (see details in Annex 1). The values are expressed in an average tonne of dry biomass (tdm) per hectare for each LC category. Some of data are available as the total carbon (C) stored in the vegetation, therefore this unit was converted to tdm with an assumption that 1 tdm contains 0.5 tC (tonne of carbon) 45. 3.2.2 Groundwater level The distribution of peat drainage depth, indicated by data on water table depth, was mapped by a combination of two mapping techniques: kriging interpolation and lookup tables. The selection of the method was based on land cover types. Kriging interpolation was applied in perennial crop area, plantation forest, bare land, and degraded peat swamp forest in areas less than 500 m from perennial crop or plantation forest areas (given that groundwater only gradually – in space - reverts to the natural level, close to the surface). For the rests of land cover and land use types, a lookup table technique was applied. The availability of spatial information (coordinates) of water table depth data measured from various drained peat areas allows the application of kriging interpolation. This technique uses some parameters derived from analysis on the spatial autocorrelation between measured data to estimate the water table depth in un-measured locations. In total 166 depth data points for Sumatra and 40 depth data points for Kalimantan were used as the input for interpolation. The depth data represent the average level of water table depth in a year. The procedures of kriging interpolation technique include spatial autocorrelation analysis, variogram modelling, kriging interpolation, and accuracy assessment. A brief explanation of the procedures is described in the Annex 13. The period of measurement is for the year of 2013 only. A lookup table technique was applied by assigning a value of water table depth to a related land cover type, using the spatial analysis tool of ArcMap 10.5. The water table depth data were selected based on a literature review. Annex 14 lists the water table depth values used in this mapping. 18 3.2.3 Hotspots The number of hotspots in peat areas was detected from MODIS (Moderate Resolution Imaging Spectroradiometer) collection 6 near real-time fire data from Fire Information for Resource Management System (FIRMS), NASA 46. The fire products contain temporal and spatial distribution information globally with 1 km pixel resolution from two satellites, Aqua (MOD14) and Terra (MYD14). These data are available on The Earth Observing System Data and Information System (EOSDIS) website from NASA. Fire products for Terra satellite can be accessed from November 2000 and from July 2002 for Aqua satellite. The determination of a hotspot in MODIS fire product is explained in Giglio et al. 47. The number of fires in peat areas was extracted using ArcMap 10.5. Peatland distribution map from the MoARI 41 and MODIS fire data were intersected to show the distribution of hotspots in Sumatra and Kalimantan peatlands. It is assumed that hotspots in the same or nearest location (approximately 1 km resolution) within one year are assumed as one pixel hotspot 48, 49. To improve the certainty of the result, this study also compares the number of hotspots in different confidence levels according to the information in Table 2. This indicator is divided into three classes, which are hotspots with confidential level 0-100%, 30-100% and 80-100%. Table 2. The meaning of confidence interval in hotspot information Confidence value Class Action 0% ≤ C < 30% Low Important to note 30% ≤ C < 80% Nominal Alert 80% ≤ C ≤ 100% High Immediate countermeasures Source: 47, 50 It is noted that this indicator indicates the detected number of fire accidents in peat areas, not the total burned area. The measurement of burned area was done separately using different satellite product (see details in section 3.4.2). 3.3 Ecosystem services account An ecosystem services account is developed for 2000, 2006, 2009 and 2014. The methods to estimate the physical and monetary values are explained in Table 3 and Table 4. For physical values, the total production of each commodity was estimated by multiplying the production rate in peat area (Annex 4 and Annex 5) and the total area (ecosystem extent account and Annex 3) 6. Next, monetary values were calculated based on the physical values, using resource rent (RR) approach for provisioning services, the social cost of carbon (SCC) for CO2 sequestrations, and restoration cost for the protected habitat. Several studies and sources (Annex 6) were referred to estimate the production costs and the total revenue. Monetary values of ES were standardised into IDR for 2017 by considering inflation rates (6.96 in 2010; 3.79 in 2011; 4.3 in 2012; 8.38 in 2013; 8.36 in 2014; 3.35 in 2015; 3.02 in 2016)51 and exchange rate of Bank Indonesia (BI)52. 19 Table 3. Indicators for physical valuation of ES provided by peatland ecosystem in Indonesia Type of ES ES specification Indicator Method Provisioning Timber production Annual timber harvested (m /year) a 3 Timber production rate × services total forest area (excluding forest in protected area) Oil palm production Annual fresh fruit branch (FFB) of oil Oil palm (FFB) production palm harvested (ton/year) rate × total area of oil palm plantation Biomass production Annual acacia biomass harvested Biomass (acacia) production for pulpa (m3/year) rate × total area of acacia plantation Paddy productiona Annual paddy harvested (ton/year) Paddy production rate × total area of paddy field Regulating CO2 sequestration a Net carbon (CO2) flux of undisturbed Net carbon (CO2) flux of services forests (ton CO2/year) undisturbed forests × total area of undisturbed peat forest Cultural Protected habitata Total area of peat swamp forests inside Total area of forest in services protected areas that are not converted protected area to other land uses since 2000 (ha) Note: methods were taken from 6 aES estimated from MoEFRI land cover data Table 4. Indicators for monetary valuation of ES provided by peatland ecosystem in Indonesia Type of ES ES specification Indicator Method Provisioning Timber production Resource rent (IDR/year) (Timber price × total production) - a services production cost Oil palm production Resource rent (IDR /year) (FFB price × total production) - production cost Biomass production Resource rent (IDR /year) (Biomass price × total production) - for pulpa production cost Paddy production a Resource rent (IDR /year) (Paddy price × total production) - production cost Regulating CO2 sequestration a Social cost of carbon Total CO2 sequestration × cost of services (IDR /year) carbon (CO2) Cultural Protected habitata Restoration cost Total area of forest in protected area × services (IDR /year) restoration cost aES estimated from MoEFRI land cover data Resource rent (RR) The value of RR is counted as the total revenue from the market price of output minus the total production cost 35, 53, 54 (Table 4). Data needed for this measurement are the market price (P), the total production (the physical value of ES) and the total production cost (see details in Annex 6). Restoration cost (not applied) It needs to be noted that the restoration cost approach is not recommended for the SEEA EEA. This because it is unclear if society would indeed restore the peatlands (or any other ecosystem) for these costs if given the choice. If not, it is not realistic to use these costs in the context of accounting. Therefore, the Restoration cost method has been excluded from the accounts, and the habitat service is expressed only in physical indicators. In order to give an idea of potential values that would accrue for restoration, costs for reforestation in peat swamp forest in tropical peatlands have been estimated at 1054 US$/ha (value for 2014) 55, 56. The cost consists of planning 20 cost, planting cost of 500 trees and the average of annual maintenance cost. Based on the inflation rates of Statistics Indonesia 51 and exchange rate in BI 52 (US$ is equal to 11081 IDR in 2014), the restoration cost would be equivalent to 11,856,417 IDR/ha (value for 2017). In addition, costs may be required for restoring the hydrology of the peatlands. These costs may be substantially higher but will vary strongly as a function of local characteristics (e.g. density and size of drainage canals). Social cost of carbon The monetary value of CO2 sequestration is valued based on marginal social damage cost technique, the social cost of carbon (SCC)35. This study selects the value of SCC for 3% discount rate at 35 US$/t CO2 (value in 2007)57. Based on the inflation rates 51 and exchange rate in BI 52 (1 US$ is equal to 9,136 IDR in 2007), therefore the SCC value for 2017 was equal to 348,17 IDR/t CO2. 3.4 Carbon account A carbon account for peatland is developed for 1990, 1996, 2000, 2006, 2009 and 2014 for carbon stocks and net carbon flux, while the indicator of peat fire emissions is only presented for 2006, 2009, and 2014. 3.4.1 Carbon stocks (vegetation) The total of carbon stocks in vegetation was estimated from biomass data in the ecosystem condition account and the references (in Annex 1) using Equation 2 and Equation 3. Carbon fraction used in this report is equal to 50% of dry biomass 45. The total carbon stocks were converted into CO2-eq by multipying estimated carbon content with 3.67 (44/12, ratio molecular weight of CO2 and carbon). Second, using ArcMap 10.5, mapping the carbon stocks applies look up table methods by asigning the value based on land cover types (Annex 1). Vegetation Carbon Content (tC) = Dry matter (t) × Carbon Fraction (C) Equation 2 CO2-eq = Carbon Content × 3.67 Equation 3 3.4.2 Carbon emissions Oxidation The total net carbon flux was calculated as the sum of carbon sequestration (positive value) and carbon emission (negative value) multiplied by the total area for every type of land cover. The references of net carbon flux for specific land cover are summarised in Annex 5. Peat fires Carbon emissions from peat fires were estimated from the emission of burned scar in peatlands. The size of the burned scar was estimated from burned area products by MODIS (MCD64A1) collection 6 58. This product contains the information of burned area (daily) with a resolution of 500 m. It can be accessed online on http://modis-fire.umd.edu/index.php. Data (.shp) that cover regional win 19 were selected for South-East Asia, including Indonesia. Burned area products then were overlaid with the peat map from MoARI and land cover maps from MoEFRI for the selected year and type of LC (forest, plantation area, agricultural land, bare ground and degraded land). The method was conducted in a software, ArcMap 10.5. In the measurement of total burned 21 peatlands, it is assumed that overlapped burned area in the same location (pixel) within the same year was calculated as one datum. Equation 4 was applied to estimate the total carbon (CO2-eq) emissions from burned peat (Efire). The formula was taken from IPCC supplement 2013 59 as reviewed by MoEFRI 48. The total carbon emissions (Efire) are the sum of the burned area (BPF) multiplied with the mass of fuel available for combustion per area (MB), the combustion factor (CF, default factor = 1.0) and the CO2 emission factor (Gef). Assuming the depth of burned peat is 33-cm for all land cover types48, MB is equivalent to 504.9-ton biomass/ha, and Gef 1,828.2 CO2 kg/ton dry biomass burned 48, 60, 61. Efire = BPF × MB × CF × Gef Equation 4 22 Chapter 4. Results 4.1 Ecosystem extent account Sumatran peatlands Table 5 shows the total areas of seventeen land covers (LCs) in Sumatra peatlands 1990- 2014/2015, with information of additions and reductions in LC classes between the selected years (five periods). Total peat areas in Sumatra is approximately 6.4 Mha. It can be seen from the table that the total LC changes are diverse in each period. A peatland area of 0.46 Mha (7%) has changed during the first period (1990-1995/1996). This number went up dramatically during the second period (1995/1996-2000) to 1.42 Mha (22%). In the next three periods, the total converted areas were still significant, around 1.1 Mha, 1.09 Mha and 1.3 Mha respectively. The most notable change of LC in Sumatra peatlands is the reduction of natural forests. In 1990, 73% of peatlands in Sumatra was covered by forests. However, according to government data, in 2014/2015, there was only 22% of peatlands remained as forests as seen in Figure 3. The highest number of deforestations, around 1.2 Mha, was recorded in the second period (1995/1996-2000). After that, the trend of deforestation was going down even though the proportions were still considerably high (0.7 Mha (third period), 0.6 Mha (fourth period) and 0.5 Mha (last period)). On the other hand, the plantation area and agricultural land were escalated, from 0.9 Mha in 1990 to 3 Mha in 2015, which is equal to 48% of total peatlands in Sumatra. 1990 2014/2015 Figure 3. Spatial distribution of land cover types in Sumatra in 1990 and 2015. Another noticeable change is the expansion of the degraded lands (wet shrub, dry shrub, savanna, swamp) and the bare ground. Most of the degraded land was wet shrub (79-86% of total degraded lands). This type of land increased significantly by 55% in twenty years (1990-2009/2010) then decreased during last period by 17%. The similar trend is also found for dry shrub, savanna, and open swamp areas. Furthermore, the bare ground was also widened dramatically since 1990 even though it had decreased once during 2005/2006-2009/2010. Other LCs such as the fish pond, 23 port/harbour, urban and mining areas show no significant changes. These types of LC only took a small part of total peatland in Sumatra, around 0.6-0.8%. Kalimantan peatlands As seen in Table 6, Kalimantan island has approximately 4.9 Mha of peatlands. In twenty-five-year period (1990-2014/2015), parts of land covers in Kalimantan peatlands have been changed (Figure 4). It was recorded that the highest total peatland conversion (14% or 0.7 Mha of Kalimantan peatlands) is during the first period, 1990-1995/1996. This number is lower compared to the percentage of peatland cover change in Sumatra. Deforestation is still the main issue for Kalimantan peatlands. Natural forest covered approximately 80% of Kalimantan peatlands in 1990, then 32% (1,5 Mha) of them have been converted until 2014/2015. About 0.6 Mha of deforestation area was found during the first period, then the number decreased to 0.23-0.28 Mha per period in the next terms. On the other hand, the total degraded land (dry and wet shrub, open swamp) and the cleared area in Kalimantan peatlands were widened significantly from 13% (of total peatland) in 1990 to 35% in 2015, or around 1.1 Mha where 0.57 Mha of it was added during the first period. 1990 2014/2015 Figure 4. Spatial distribution of land cover types in Kalimantan in 1990 and 2015. Plantation and agricultural lands were also widespread in Kalimantan peatlands. In 1990, there were only 0.36 Mha of these types of LC (dominated by dry agricultural lands). However, this number was doubled to 0.77 Mha until 2015. Plantation and agricultural lands were expanded in every period. The highest changes were found during 2005/2006-2009/2010 for perennial crops, during 2010-2014/2015 for forest plantation and 1990-1995/1996 for agricultural lands. Looking at trend for other types of LCs, urban areas in Kalimantan peatlands increased during 1990-1995/1996 and then remained constant in the next period (0.02 Mha). Meanwhile, mining areas went up during the last period. 24 Table 5. Extent account of peatlands in Sumatra Total area (in 1000 ha) Undisturbed forest Forest plantation Disturbed forest Degraded peatland Dry agricultural Perennial crops Cloud coverage Bare ground Mining area Paddy field Fish pond Airport Water Urban Savanna and Peatland cover Total land Wet shrub Dry shrub swamps Grasses Sumatra Open Opening stock 1990 481 4159 106 617 7 38 33 7 378 317 192 5 30 6 1 0 1 6376 Additions 0 22 25 75 0 0 64 30 158 68 15 0 1 0 0 0 0 456 Reductions 30 357 2 37 0 0 1 6 0 20 5 0 0 0 0 0 0 456 1995/1996 450 3824 128 656 7 38 96 32 535 365 202 5 31 6 1 0 0 6376 Additions 0 2 49 608 59 26 131 16 421 91 10 0 4 1 6 0 3 1426 Reductions 73 1166 26 92 2 3 14 0 15 35 0 0 0 0 0 0 0 1426 2000 37 2659 150 1173 63 61 213 48 941 422 213 5 35 6 7 0 3 6376 Additions 50 94 19 292 0 0 344 225 81 7 2 0 0 0 1 0 0 1116 Reductions 26 673 17 267 6 0 91 12 15 8 0 0 0 0 0 0 1 1116 2005/2006 402 2081 152 1198 57 61 466 262 1007 421 214 4 35 7 8 0 2 6376 Additions 0 59 33 347 0 0 149 219 215 65 1 0 0 0 0 0 0 1087 Reductions 121 497 16 111 0 2 260 61 11 7 0 0 0 0 0 0 2 1087 2009/2010 281 1642 169 1434 57 60 355 420 1211 479 215 4 35 6 8 0 0 6376 Additions 0 50 64 166 0 10 183 493 201 154 3 0 0 0 0 0 0 1324 Reductions 56 435 134 405 6 22 158 50 13 20 25 0 0 0 0 0 0 1324 2014/2015 225 1257 100 1195 52 47 380 864 1398 612 192 4 34 7 8 0 0 6376 Table 6. Extent account of peatlands in Kalimantan Total area (in 1000 ha) Undisturbed forest Forest plantation Disturbed forest Degraded peatland Dry agricultural Perennial crops Cloud coverage Bare ground Mining area Paddy field Fish pond Airport Water Urban Savanna and Peatland cover Total land Wet shrub Dry shrub swamps Grasses Kalimantan Open Opening stock 1990 113 3790 86 398 0 105 27 0 59 231 68 6 16 1 0 0 0 4898 Additions 0 19 12 435 0 102 23 1 33 30 46 0 7 0 0 0 0 708 Reductions 33 575 25 20 0 11 6 0 19 19 1 0 0 0 0 0 1 708 1995/1996 80 3234 73 814 0 196 44 1 73 243 113 5 22 1 0 0 0 4898 Additions 0 9 7 251 0 1 5 0 11 12 2 0 0 0 0 0 0 299 Reductions 11 265 0 8 0 0 4 1 1 7 2 0 0 0 0 0 0 299 2000 68 2978 79 1058 0 198 44 0 83 248 113 5 22 1 0 0 0 4898 Additions 0 52 4 274 0 29 21 0 48 26 5 0 0 0 0 0 0 459 Reductions 6 231 3 107 0 99 3 0 0 6 3 0 0 0 0 0 0 459 2005/2006 62 2799 80 1225 0 127 62 0 131 268 115 5 22 1 0 0 0 4898 Additions 0 1 5 139 0 2 41 1 126 11 10 0 0 0 0 0 0 337 Reductions 6 236 11 66 0 1 9 0 0 9 0 0 0 0 0 0 0 337 2009/2010 58 2565 74 1298 0 128 94 1 256 270 126 5 22 2 0 0 0 4898 Additions 0 14 3 86 0 1 119 29 81 15 1 0 0 1 0 0 0 349 Reductions 7 271 1 52 0 5 9 0 1 1 1 0 0 0 0 0 0 349 2014/2015 50 2308 77 1332 0 123 203 30 336 284 126 6 22 3 0 0 0 4898 26 4.2 Ecosystem condition account 4.2.1 Vegetation biomass Table 7 summarises the total dry biomass matter in Sumatra and Kalimantan peatlands, estimated based on references in Annex 1. In a twenty-five-year period (1990 to 2015), 35% and 27% of total vegetation biomass was lost in Sumatra and Kalimantan, respectively. The loss of biomass was mainly caused by the LC changes as explained in the ecosystem extend account. Referring to data sources (Annex 1), the undisturbed forests have the highest biomass density (419 tdm/ha in Sumatra and 383 tdm/ha in Kalimantan), followed by disturbed forests (275 tdm/ha and 276 tdm/ha). Around 91% (Sumatra) and 95% (Kalimantan) of total biomass in 1990 was stored in the form of forests, but this number decreased to 46% and 76% in 2015. The converted areas such as plantation forest, perennial crops, agricultural and degraded lands only had a slight contribution to the additional biomass because of the low density of biomass. The loss from deforestation was much higher (Figure 5). Table 7. Total vegetation biomass in Sumatra and Kalimantan peatlands 1990-2015. Indicator Peatlands in 1990 1995/1996 2000 2005/2006 2009/2010 2014/2015 Vegetation Sumatra 1475 1409 1170 1079 991 965 biomass (Mtdm) Kalimantan 1148 1015 959 928 887 835 Mtdm: Million tonne of dry matter 1500 1500 Sumatra Kalimantan Biomass (Mtdm) 1000 1000 500 500 0 0 Figure 5. Total biomass (vegetation) in different land cover types of peatland in Sumatra and Kalimantan 1990-2015. 4.2.2 Groundwater level The estimated groundwater level is retrieved using interpolation of the (limited set of) available data points. The result (Figure 6) shows that the annual average of water level in 2013 varied from 0-117 cm in Sumatra and from 0-96 cm in Kalimantan. The deepest drainage was in the areas of perennial crop, plantation forest, bare land and degraded peat swamp forest in the distance less than 500 m from those areas. It was deeper in north-eastern parts of Sumatra. The value of coefficient of variation (CV) of root mean square error (RMSE) for the model are 0.15 (Sumatra) and 0.19 (Kalimantan). These numbers are close to 0 (range 0-1) which means that the models are sufficiently accurate, with regards to representing the data points (see details in Annex 13). Legend 0 - 20 cm 20 - 40 cm 40 - 60 cm 60- 80 cm 80 - 100 cm 100 - 117 cm Non peat Figure 6. Estimated water level map of Sumatra and Kalimantan peatlands in 2013. 4.2.3 Hotspots The hotspots are detected in different LC types (Figure 7). The highest percentage of total hotspots was in peatlands covered by wet shrub (33-45% in Sumatra, 46-57% in Kalimantan). Hotspots in Sumatra were also found in bare ground (10-22%), perennial crops (11-14%), disturbed forest (9-14%), and other LC types (below 10%). A significant increasing number of hotspots was showed in forest plantation from 1% in 2006 to 18% in 2014. Meanwhile in Kalimantan, other than in wet shrub, 17-25% of total hotspots was detected in disturbed forest, while other LCs experienced hotspots under 8% except the in perennial crops (14% in 2009 and 11% in 2014). 6000 6000 5000 5000 Number of hotspots 4000 4000 3000 3000 2000 2000 1000 1000 0 0 2006 2009 2014 2006 2009 2014 Sumatra Kalimantan Figure 7. Number of hotspots (with confidence value ≥ 80%) in different land cover types of peatland in Sumatra and Kalimantan for 2006, 2009 and 2014. MODIS reports uncertainty of hotspot measurements. The higher the applied confidence level, the lower the number of hotspots detected. Around 95-96% of total hotspots have minimal confidence 28 level 30, meanwhile 42-55% of hotspots was found in the higher level (confidence level ≥80). The total hotspots also vary temporally. In total, the least number of hotspots was during 2009, and the highest ones were in 2006 and 2014. A lower number in 2009 could be affected by a La Niña phenomenon in the previous year, where the climate condition of Indonesia is wetter than average (see details about ENSO years in Annex 11). The spatial distribution of hotspots can be seen in Annex 2. 4.3 Ecosystem service account Table 8 shows the physical values of five ES estimated based on MoEFRI land cover data. The values of three ES, timber production, CO2 sequestration and protected habitat decreased since 2000. Timber production in Sumatra and Kalimantan peatlands went down by 59% and 27%. The total CO2 sequestration declined by 40% (Sumatra) and 26% (Kalimantan) while the total protected habitat was narrowed by 6% (Sumatra) and 11% (Kalimantan). On the other hand, the expansion of forest plantation led to the increasing of total biomass (acacia) production dramatically, in particular Sumatra, from 1 Mt/year in 2000 to 18 Mt/year in 2015. Besides, the results show no notable change on paddy production. The production of paddy has slightly increased from 2000 to 2010, then remained constant (in Kalimantan) and has decreased (in Sumatra) in 2015. The spatial distribution of each ES can be seen in Annex 7. Table 8. Physical values of ecosystem services in Sumatra and Kalimantan peatlands. Physical value of ES ES specification Unit 2000 2005/2006 2009/2010 2014/2015 Timber production 1000 m3/year Sumatra 1893 1482 1094 777 Kalimantan 794 741 666 576 Biomass production 1000 t/year Sumatra 1011 5503 8833 18161 for pulp Kalimantan 0 2 24 624 Paddy production 1000 t/year Sumatra 620 625 627 561 Kalimantan 192 196 214 214 CO2 sequestration 1000 t/year Sumatra 7175 7629 5337 4282 Kalimantan 1299 1182 1099 958 Protected habitat 1000 ha Sumatra 442 451 423 416 Kalimantan 892 851 816 794 Table 9. Monetary values of ecosystem services in Sumatra and Kalimantan peatlands. Monetary value of ES (IDR billion/year) ES specification 2000 2005/2006 2009/2010 2014/2015 Timber production Sumatra 1278 1001 739 525 Kalimantan 536 500 450 389 Biomass production for Sumatra 95 518 831 1709 pulp Kalimantan 0 0 2 59 Paddy production Sumatra 1510 1522 1526 1365 Kalimantan 338 344 375 376 CO2 sequestration* Sumatra 2498 2656 1858 1491 Kalimantan 452 412 383 334 Protected habitat Sumatra - - - - Kalimantan - - - - 29 Comparing the monetary value of five ES in 2000, CO2 sequestration had the highest value, following by paddy production and timber production (Table 9). Note that if a restoration cost approach would be followed to value biodiversity based on the values presented in the previous section, then this service would have the highest value (reflecting the high costs of restoring peatlands). Due to the decreasing physical values of timber production and CO2 sequestration over time (related to land conversion), the monetary values of these ES went down as well. Timber production that provided IDR 1,814 billion throughout 2000 decreased to IDR 914 billion in 2015, the lowest value among five ES. Oil palm production Table 10 explains the total area of oil palm plantation in peatlands from several data sources. According to the references (Annex 3), there were only 0.6 Mha (Sumatra) and 840 ha (Kalimantan) of oil palm plantation in 2000. The areas of plantation in both islands expanded dramatically, and in 2014, they took 22% and 10% of total peatlands in Sumatra and Kalimantan. In total, there were 1.9 Mha of oil palm plantation in peatlands in 2014. Table 10. Total area of oil palm plantation in Sumatra and Kalimantan peatlands in 2000, 2005, 2010, and 2014 ES specification Total area in peatland (1000 ha) 2000 2005/2006 2009/2010 2014 Oil palm production Sumatra 621a 1007b 1210a 1414c Kalimantan 1a 131 b 256a 480 c a:estimated from Gunarso et al. 34, the values were adjusted from Wetlands International map 62, 63 with estimated peat area of 20.8 million ha to MoARI map (14.9 million ha of peatland);b: the total area of perennial crops of LC map of MoEFRI, we used this source because estimated oil palm plantation in peatland from Gunarso et al. 34 in 2005 and 2010 was larger, which could be located in other LC; c: estimated from plantation map from WRI and Transparent World 37, overlaid with peatland map from MoARI (see details in Annex 3). The expansion of oil palm plantation resulted to the increasing of total production. It is estimated that there were 32 Mt of FFB produced in 2014, triple times higher than the total production in 2000. The estimated economic value of the production reached IDR 4,336 billion in 2014, of which 93% was produced in Sumatra peatlands (Table 11). Table 11. Physical and monetary value of oil palm production in Sumatra and Kalimantan peatlands in 2000, 2005, 2010, and 2014 Physical value Monetary value of ES ES specification (1000 t FFB/year) (IDR billion/year) 2000 2005/ 2009/ 2014 2000 2005/ 2009/ 2014 2006 2010 2006 2010 Oil palm Sumatra 10389 16837 20242 23635 1764 2858 3436 4012 production Kalimantan 14 2185 4282 8022 1 88 173 324 30 4.4 Carbon account 4.4.1 Carbon stocks (vegetation) Table 12. Total carbon (CO2-eq) stocks as vegetation in Sumatra and Kalimantan peatlands 1990-2015. Deforestation and land use changes in peatland lead to the decrease of carbon stored in vegetation. Table 12 explains that vegetation of peatlands in 1990 stored around 4.8 Gt of CO2-eq, however, 31% (1.5 Gt of CO2-eq) of this stock was lost in 2015. Spatially, the decrease in carbon stocks was found in every province as seen in Figure 8. The extensive green areas (indicating high carbon stocks) in 1990 faded in next twenty-five years (2014/2015). The largest carbon loss was detected during the first period (1990-1995/1996) and second period (1995/1996-2000), Afterwards, the rate of total carbon loss decreased. Figure 8. Spatial distribution of carbon stocks (vegetation) in Sumatra and Kalimantan peatlands in 1990 and 2014. 31 1000 Social cost of carbon 750 (Trillion IDR) 500 250 0 2000 1990 1995/1996 2014 1990 2000 2014 2005/2006 2009/2010 1995/1996 2005/2006 2009/2010 Sumatra Kalimantan SCC with discount rate 3% SCC with discount rate 5% Figure 9. Social cost of carbon of carbon stocks (vegetation) in peatlands of Sumatra and Kalimantan 1990-2014/2015. Decreasing of carbon stocks in peatlands lead to the increasing of damage costs of releasing carbon as emissions by the vegetation in peatlands. Figure 9 explains that based on SCC, peatlands of both islands are valued 1676 Trillion IDR (discount rate 3%) or 512 Trillion IDR (discount rate 5%) in 1990. These numbers went down significantly to 1150 Trillion IDR (discount rate 3%) or 351 Trillion IDR (discount rate 3%) in 2014/2015. The estimated cost of damages from carbon stock loss during the period was about 526 Trillion IDR (discount rate 3%) or 161 Trillion IDR (discount rate 3%). 4.4.2 Carbon emissions Oxidation (estimate) Mt CO2/year Indicator Peatlands of 1990 1995/1996 2000 2005/2006 2009/2010 2014/2015 Sequestration Sumatra 9 9 7 8 5 4 Kalimantan 2 2 1 1 1 1 Emissions Sumatra -140 -155 -185 -202 -230 -276 Kalimantan -93 -96 -96 -100 -109 -116 Total Sumatra -131 -146 -178 -195 -225 -272 Kalimantan -91 -94 -95 -99 -108 -115 Table 13. Total CO2 sequestration and emissions (from oxidation) in peatlands 1990-2015. Table 13 displays the net carbon flux (sequestration and emission) from peatlands in Sumatra and Kalimantan based on referenced data (Annex 5). The estimated values also include the emission effected from drainage but not cover the emissions from peat fires. The total carbon fluxes were negative, indicating that CO2 emissions were higher than the sequestrated CO2. In total, net CO2 emissions from peatlands in Sumatra and Kalimantan from drainage were 222 MtCO2/year in 1990, then increased by 74% to 387 MtCO2/year in 2015. Around 70% of total CO2 emissions were released from Sumatran peatlands. Based on SCC, the estimated cost of damages of the carbon emissions released due to oxidation (Figure 10) in peatlands in 1990 was around 77 Trillion IDR (discount rate 3%) or 23.6 Trillion IDR (discount rate 5%) in total. In 2014/2014, the cost increased as the total emission went up by 74% or 57 Trillion IDR (discount rate 3%) or 17.5 Trillion IDR (discount rate 5%). 32 Sumatra Kalimantan Social cost of carbon (Trillion IDR) 2014/2015 1995/1996 2005/2006 2009/2010 1995/1996 2005/2006 2009/2010 2014/2015 1990 2000 1990 2000 0 -25 -50 -75 -100 SCC with discount rate 3% SCC with discount rate 5% Figure 10. Social cost of carbon of carbon (CO2-eq) emissions from oxidation in peatlands 1990- 2014/2015. sequestrations 10.0 10.0 emissions -40.0 -40.0 Net carbon flux (Mt CO2) -90.0 -90.0 -140.0 -140.0 -190.0 -190.0 -240.0 -240.0 -290.0 -290.0 -340.0 -340.0 1990 2000 2009/2010 1995/1996 2005/2006 2014/2015 1990 2000 2009/2010 1995/1996 2005/2006 2014/2015 Sumatra Kalimantan Figure 11. Trend of CO2 emissions and CO2 sequestrations in Sumatra and Kalimantan peatlands 1990-2015 based on land cover types. The bars depict the emission from each land cover type. These emissions are a function of drainage. Note that it is assumed that all plantations and dry agricultural land are drained. Disturbed forests and degraded peatlands only emit CO2 when they are drained. In Sumatra, perennial crops and forest plantations contributed to 43% and 25% of total CO2 emissions in 2015 (Figure 11). Note however that this assessment is based on the MoEFRI land cover map. In case the MoEFRI map underestimates the area covered with plantations, the contribution of plantations to these emissions may be higher. Depending upon accuracy of the land cover map, this is a very relevant policy finding – in particular that 32% of CO2 emissions from peatlands in Sumatra are not from oil palm, hevea or plantation forestry peatlands but from other land, in particular degraded land and drained forest edges close to plantations. These areas should be a first priority for rehabilitating peatlands since their rehabilitation would make a very 33 large contribution to reducing global CO2 emissions and not cause losses of crop production. The expansion of plantations in Kalimantan peatlands, in 2015, was not as significant as in Sumatra, and this sector (both perennial crops and forest plantation) contributed to 22% of CO2 emissions in 2015. Around 34% of CO2 emissions in Kalimantan was released from drained forests and 20% from drained, degraded lands. Restoring the hydrology of these areas should be a global priority in the mitigation of climate change, given the magnitude of emissions involved, the irreversibility of the effects of drainage, and the low costs of restoration compared to the carbon volumes involved. Peat fires Table 14 presents the total burned peatland areas and the carbon emission emitted from burned peat. Compared to the value of net carbon flux in the previous section, carbon emissions from peat fires were much higher, 704 Mt CO2-eq in 2006), 508 Mt CO2-eq in 2009), and 610Mt CO2-eq in 2014. Although the total peatland in Sumatra is larger than in Kalimantan, the results show that the carbon emissions from Kalimantan peatlands were higher due to larger burned areas every year. Table 14. Total burned peatlands and total carbon (CO2-eq) emissions from burned peat in Sumatra and Kalimantan in 2006, 2009, and 2014. Total burned peatlands Carbon emissions from peat fires* Peatlands in (1000 ha/year) (Mt CO2-eq)/year 2006 2009 2014 2006 2009 2014 Sumatra 345 198 310 318 183 286 Kalimantan 418 352 351 386 325 324 Method are taken from MoEFRI 48 *estimated 33-cm burned peat depth for all types of LCs. Based on SCC, high carbon emissions from burned peat cause loss which were estimated costed 184-256 Trillion IDR (discount rate 3%) or 56-78 Trillion IDR (discount rate 5%) each year in total (both islands, Sumatra and Kalimantan). Sumatra Kalimantan Social cost of carbon 2006 2009 2014 2006 2009 2014 (Trillion IDR) 0 -50 -100 -150 SCC with discount rate 3% SCC with discount rate 5% Figure 12. Social cost of carbon of carbon (CO2-eq) emissions from burned peat 1990-2014/2015 Figure 13 presents the CO2-eq emissions from different types of LC. Similar to the pattern of hotspot distribution, the highest emissions were from wet shrub 26-58% in Sumatra and 60-83% in Kalimantan. The details for the spatial distribution of burned scars can be seen in Annex 9. 34 400 400 Carbon emissions (CO2-eq) 300 300 200 200 100 100 0 0 2006 2009 2014 2006 2009 2014 Sumatra Kalimantan Figure 13. Carbon (CO2-eq) emissions from burned peat in several types of peatland covers in 2006, 2009, and 2014. Note that the emissions were from 33-cm burned peat, applied for all types of land cover. Note that the land cover classes were attributed based on the MoEFRI or KLHK land cover map. 35 Chapter 5. Discussions 5.1 Uncertainties and Limitations Peatland map and land cover maps There are several sources of uncertainty in this pilot ecosystem account. The first concerns the peatland area and land cover data. Several peatland maps have been published in the period 1952- 2016 by different authors (and using different assumptions and methods) with total areas reported varying from 13 Mha to 26.5 Mha 6,16. The account uses the peat map of the Indonesian government, BBSDLP, MoARI, which is the result of combined analyses of Landsat and ground survey data with 2,409 observation points for peat thickness and maturity 16. Comparing to the other versions of peat map from Wetland International (62, 63, 64) for example, the total peatlands in MoARI map are 5.7 Mha smaller. In addition, the same peat map is used for the whole period of analysis (1990-2015). This means that the assumption is that the peatland areas have remained the same and the changes of the spatial distribution and peat characteristics including peat depth are excluded 16. Also, as mentioned, the account is based on land cover data from MoEFRI. In this dataset, information on land cover for the year 2010 is partly based on remote sensing images of 2009; and information for the year 2015 is partly based on remote sensing images of 2014. Hence, the land cover changes are underestimated in this account (since there will in some areas have been further land use conversion from 2009 to 2010 and from 2014 to 2015). This underestimation may lead to an inaccuracy in the land cover data (affecting also the accounts for ecosystem services and carbon stocks and emissions). Second, there are uncertainties in land cover in peatlands, related to the accuracy of land cover maps from the MoEFRI. Over time these maps have been increasing in quality, with particularly for the years before 2000 a somewhat higher uncertainty due to a lack of ground truth data. With the improvement of methodology, the newest map (2015) reached overall accuracy 67% (forest) and 93% (non-forest) using stratified random sampling approach (18 as reviewed in the report of FAO 19). These uncertainties, as reported by the government of Indonesia, are confirmed when the peat map is compared with the land cover map. For instance, some (relatively small) areas are classified as mangrove in the land cover map and peat in the peat map, mostly in Papua (which is not ecologically consistent). In addition, some inconsistencies are found between the government maps and the maps of WRI and OFI (as explored in Annexe 8). In particular, some parts of plantation areas indicated in WRI and OFI maps are located outside the plantation area of MoEFRI maps. Satellite products Satellite products are commonly used due to large coverage area and data scarcity in the field. Satellite products are also applied in this study, in particular to estimate the number of fire hotspots and burned area in peatlands. The gridded 500m burned area products from MODIS were used in the carbon account to estimate burned peatlands. Hotspot data from MODIS have an accuracy of 64-84% in estimating fire occurrence in Indonesia (66% for peatlands), when compared to higher resolution maps such as SPOT-4 images and SPOT-5 65, 66. However, this accuracy is higher than another hotspot product, Indofire hotspot product 65. In response, this account, in line with government practices, uses only MODIS fire detection data with a high confidence value (≥80, around 45% of total hotspots). Note that this may lead to an underestimate of hotspots, since also areas with a lower confidence value may have been subject to fire 65, 67. 36 Further improvement of methods and data to estimate peat fire occurrence in Indonesia is needed. There is a new fire product from NASA, called VIIRS I-Band 375 that has a finer spatial resolution (375 m) than MODIS, available from 2012 to present 68. It is recommended to validate this data source in order to increase the accuracy of hotspot monitoring in Indonesia 69 70. Note also that it was found that burned area products from MODIS perform better than other products like L3JRC, even though MODIS product could underestimate the burned area in the forest 71. Data scarcity Lack of available ground truth information is also a main source of uncertainty. First, there was a lack of sufficient data points to estimate groundwater level. The groundwater level varies within and between land cover classes, in peat in particular as a function of drainage. More data are available on groundwater levels in peat (e.g. with Deltares and Hokkaido university) but in spite of repeated requests data holders were not willing to share any data for the peat account. Ultimately, the account has been based on 206 measurement points of groundwater levels in peat in combination with an extrapolation process that considers land cover and distance from (drained) plantations. Further improvements to the account are possible if access to more data are provided. Second, the ecosystem service (ES) account has uncertainties related to the flows of ecosystem services. This study tries to develop the account by using governmental data. However only data of market prices can be accessed in some governmental publications. Information regarding average production rate of provisioning services are also available, however as average value, not specifically from particular ecosystem type (peatland). Therefore, data from research articles were applied (Annex 4, 5, and 6). Still, the available data are generalised at island (Sumatra and Kalimantan) and national level, not at specific area (provincial or district level) due to the limited number of sources. Further refinements would be possible with inserting production data specific for peatlands. Uncertainty is also found in the combustion and emission factor for carbon emissions 71. It is assumed that all categories of LC have the same value, and that the emission factor is only a function of burned peat depth and standing biomass (not of the type of vegetation). However, this could affect the accuracy of the results because the depth of burned peat could vary between 25 and 85 cm 72 in different LC types. Compared to other sources, the total carbon emissions estimated in this account are lower number than the estimation in 10, 72 (value in 2006) but higher numbers than recent estimates by FAO (http://www.fao.org/faostat/en/#data)73. The estimation of this study and FAO used the same burned product MODIS, but different land cover data and emission factors 73, 59, 74, see details in Annex 12). Possible improvements This pilot study does not yet cover all aspects relevant for managing peat ecosystem, including for instance a lack of information on the hydrological ecosystem services provided by peatland, and a lack of information on the health effects of peat fires. In addition, more data would increase the quality of the accounts, in particular on fire occurrence and drainage depth. Additionally, more provisioning services such as rattan and sago cultivation could be included. Also, the habitat service could be mapped in more detail, for example by including habitats for two iconic species occurring in the peatlands, tiger and orangutan. The effects of seasonal flooding in subsided peatlands (already occurring in parts of Sumatra) and the effects of further soil subsidence on crop production should be included in an ecosystem asset account that would be an important 37 part of a peat account (reflecting the natural capital contained in the peat ecosystems, under current management). Furthermore, the accounts should be extended to Papua to inform land management (and rapid land use change) in this part of the country. Lastly, MoEFRI has published a new map of peat hydrology unit (called KHG in Indonesian) with finer resolution (1:50,000). This map could be applied (reanalysis) to improve the certainty of total peat area in Indonesia, therefore the quality of the results of peat accounts can be improved. 5.2 Policy applications By integrating and reporting on various aspects of environment and human activities in peat ecosystem, the pilot ecosystem accounts can support a range of policies in Indonesia. First, the accounts can support the rehabilitation of peatlands. Formed in 2016, the peat restoration agency (BRG) has a target to restore approximately 2.5 Mha degraded peatlands by 2020 (SK.05/BRG/Kpts/2016)15. Pilot ecosystem accounts for Indonesian peatlands can facilitate identifying areas that can be considered a priority for rehabilitation for example because they emit high amounts of carbon and do not have a productive use at present. This can help increase effectiveness and efficiency of the BRG work programme. Also, the accounts can support monitoring the peat restoration activities and the impacts of changes that have been implemented. In particular, the peat accounts can support the local, provincial governments and the BRG in monitoring the progress of the restoration program. This pilot account presents maps for the two islands, but the resolution is high (30m grid size) and local level maps can also be produced using the same data. These data can support activities at district or provincial level including monitoring and reporting land-cover change in degraded peatlands, vegetation rehabilitation, restoration of hydrology, and rehabilitation of carbon sequestration and storage13, 75. Second, the pilot ecosystem accounts can support projects of the MoFRI on financial reporting of natural resources use (policy under development), and of the MoEFRI on assessing the Carrying Capacity and Resilient Power in the Environment (DDDTLH) under regulation (UU) No.32/2009. These are basic information sets for planning and development control supporting related policies including the preparation or evaluation of spatial plans (RTRW), long-term and medium-term development plans (RPJP and RPJM) as well as policies, plans and/or programs that have potential to impact and/or risk the environment through a Strategic Environmental Assessment (KLHS). The pilot ES account in this study is consistent with the project since ecosystem services are the main method used in DDDTLH, and peatland is one of the targeted ecosystems in this policy instrument. The developed ES account can be used to facilitate further analysis concerning land use planning in diverse types of ecosystem. Comparing ES values between peatland and mineral soil, for example. It is found that oil palm plantation in peatland emits high CO2 emissions but produces less around 12-18 tFFB/ha/year, which is classified as the third class of oil palm production in mineral soil 77-79. Furthermore, the production costs of plantation in peatland are much higher compared to in mineral soil. This is mainly because there are additional costs for drainage and for constructing and maintaining infrastructure (for example roads need more regular upkeep in peat areas due to subsidence) 35, 36. Note also that the basic framework of ecosystem accounting is flexible. The list of ecosystem services can be adjusted depending on the priority of policy and location. Third, peatlands in Indonesia contribute substantially to the total carbon (CO2) emissions of the country. The carbon account in this study can support the National Action Plan to Reduce 38 Greenhouse Gas (GHG) Emissions as per Presidential Decree No.61/2011, by monitoring carbon emissions from peatlands. Furthermore, the carbon account facilitates monitoring not only the carbon emissions, but also the change of carbon stocks. Both carbon stocks and emissions are also related to the National forest reference emissions level for REDD+, in the Context of Decision 1/CP.16 Paragraph 70 UNFCCC 48. The carbon account provides very detailed and spatially explicit information by combining several sources of carbon stocks for all land cover classes of peatland, in combination with models for estimating carbon emissions from peat drainage and fire. Fourth, the peat account can support the design of fiscal transfer programs. The Government of Indonesia is currently considering if and how the sustainability of natural resource management could be an indicator to be considered as a basis for evaluating the performance of district governments. The peat account provides information on, for example, the sustainability of land use, fire occurrence or CO2 emissions by district (for those districts that have peat soils in their territory). Another relevant application is the payment for ecosystem services (PES) in Indonesia. There are at least fourteen regulations and several projects that are related to the application of PES in Indonesia 80. Ecosystem services implemented for PES projects in Indonesia are mostly related to preserving regulation and cultural services such as watershed protection, landscape/seascape beauty, recreation/biodiversity and carbon sequestration. The results (ES account) of this study are consistent with the potential application of PES in Indonesian peatlands, particularly for carbon sequestration, watershed protection and biodiversity habitat. The framework is flexible to be applied depending on the data availability and policy priority, not limited by the listed ecosystem services. The accounts can be used both to use the effectiveness of the payments (in terms of actual changes in ecosystems compared to nearby ecosystems) and to analyse co-benefits (e.g. of REDD+ projects). 39 Chapter 6. Conclusions The pilot ecosystem account for peatlands is developed to test the SEEA EEA framework for a specific, highly policy-relevant ecosystem. Data were available to produce four pilot ecosystem accounts (extent, condition, services, carbon) for peat ecosystems in Sumatra and Kalimantan for multiple years in the period 1990-2015. The ecosystem extent account summarises the trend of land cover changes spatially and temporally. The overlay of the peat distribution map and the land cover maps provide information on seventeen peatland land cover categories. The results show that the trends in land cover change in Sumatra and Kalimantan are different. The conversion of peat forests in Sumatra started earlier than in Kalimantan, and has been more thorough, to date. In the period 1990 to 2015, Sumatra has lost 47% of peat forest while in Kalimantan this was 32%. As a result, according to government data, in 2015, Sumatran peatlands were covered dominantly by plantation and agricultural land (48%). Meanwhile, only 16% of Kalimantan peatlands was covered by plantation (7%) and agricultural land (8%), in 2015. However, land cover change is still ongoing in Kalimantan whereas it has slowed (but not stopped) in Sumatra. The condition account comprises three key indicators for the state/health of peatland ecosystems (vegetation biomass, water level and fire hotspots). The condition account shows both temporal (1990-2015) and spatial patterns (30m grid) in these indicators. The account shows that land cover change evokes a reduction in total above ground biomass in the ecosystem – since annual and perennial crops and plantation forests have overall a much lower standing biomass than natural forests. The account shows that groundwater levels in peatlands in Indonesia range from 0 to 117 cm – for the points for which data were available. In undrained forests, the annual average groundwater table is close to the surface (allowing the peat to keep on accumulating carbon) and in plantations and drained, degraded lands the groundwater level varied from 30 to 117 cm. The deepest drained areas where the located in the north-eastern part of Sumatra. The recording of fire hotspots (based on MODIS) also provides a number of insights relevant for peat management. First, there were no fire hotspots in forests located at least 500m away from plantations where water tab les are still high. The number of fires was highest in drained, degraded (i.e. unused) peatlands followed by annual and perennial croplands and forestry plantations. Also, in the peatland zones close (<500 m) to the plantations, where forests are heavily degraded, and land is drained due to spill-over effects of drainage in plantations (near a drained area, the groundwater table reverts to close to the surface over a distance of around 500 meters), there are frequent fires. Draining and converting these lands to plantations, of course, does not solve the problem since this case the next strip of land to become drained and vulnerable to fire. Lastly, the number of hotspots varies considerably from one year to the next. It was highest in 1997 and 2017 (years not covered in the accounts) when there was an El Nino effect, and relatively low in 2015 which was a year in which the rainfall patterns were dominated by ENSO. The ecosystem service account includes the physical and monetary values of the main services provided by peat ecosystem in Indonesia. The expansion of plantation and agricultural areas has led to increased agricultural production of, in particular, oil palm and acacia biomass (for pulp). On the other hand, there was a decline in timber production, CO2 sequestration and in the area of protected peatland as biodiversity habitat. It is noted that the monetary values in the account do not cover all aspect of ecosystem services in peatlands. It is also not an indicator of sustainability since the externalities and ecosystem disservices were not quantified. 40 Lastly, the carbon account points to the contribution of Indonesian peatlands to greenhouse gas emissions in the country. The results indicate that peatland in Indonesia is a large source of carbon emissions. The total emissions from peat oxidation have increased over time while also fires in peatland release high carbon emissions every year, depending on the total burned area. In addition to this, the carbon account also shows that the total carbon stocks in above ground biomass (vegetation) have been decreasing. The pilot peatland account has numerous potential policy applications including the monitoring of ongoing land conversion in peatlands and the success of rehabilitation programs (in terms of both area and positive impacts achieved on CO2 emissions and ecosystem services). The accounts can also support local land use planning. In particular, the peatlands can identify areas that are of priority for rehabilitation because they lead to high CO2 emissions and have a low current productive use. It is also helpful for local governments including district and provincial level governments) to have access to the peat account for their jurisdiction, and at high resolution. This would assist local governments with better managing peat resources for example by supporting the allocation of permits to convert (or decisions to protect) land, and by allowing them monitoring of economic impacts of peat uses. Various projects indicate that local governments often have a major lack of information on peat (location, depth, CO2 emissions, biodiversity) and that their management decision are hampered by a lack of such information. It is also recommended to explore how the account (at high resolution) can be made available to local governments, potentially using a specific app/ tool with which information can be downloaded for specific areas of interest (e.g. a district, province, or part thereof). This because detailed information on land use, ecosystem services and natural capital is often lacking at lower government levels (provinces, districts) – where nevertheless much of the policy making and implementation on land use is conducted. Furthermore, the condition account could be expanded with the location of peat domes (an important aspect to be considered in peat management, data are available but have not yet been included in the account) and the location of canals (data are available for part of the country, including at BRG). In addition, it is recommended to update the account to a recent year, for example 2017 and potentially to add more ecosystem services (e.g. water regulation, control of fires and avoiding health effects from smoke, supply of non-timber forest products, and habitat for key species such as tiger and orangutan). 41 References 1. Sunderlin, W. D., Between danger and opportunity: Indonesia and forests in an era of economic crisis and political change. Society & Natural Resources 1999, 12, (6), 559-570. 2. Abood, S. A.; Lee, J. S. H.; Burivalova, Z.; Garcia‐Ulloa, J.; Koh, L. P., Relative contributions of the logging, fiber, oil palm, and mining industries to forest loss in Indonesia. Conservation Letters 2015, 8, (1), 58-67. 3. Koh, L. P.; Miettinen, J.; Liew, S. C.; Ghazoul, J., Remotely sensed evidence of tropical peatland conversion to oil palm. Proceedings of the National Academy of Sciences 2011, 108, (12), 5127-5132. 4. Broich, M.; Hansen, M.; Stolle, F.; Potapov, P.; Margono, B. A.; Adusei, B., Remotely sensed forest cover loss shows high spatial and temporal variation across Sumatra and Kalimantan, Indonesia 2000–2008. Environmental Research Letters 2011, 6, (1), 014010. 5. Margono, B. A.; Potapov, P. V.; Turubanova, S.; Stolle, F.; Hansen, M. C., Primary forest cover loss in Indonesia over 2000-2012. Nature Climate Change 2014, 4, (8), 730-735. 6. Uda, S. K.; Hein, L.; Sumarga, E., Towards sustainable management of Indonesian tropical peatlands. Wetlands Ecology and Management 2017, 1-19. 7. Agus, F.; Subiksa, I. M., Lahan Gambut: Potensi untuk pertanian dan aspek lingkungan. Balai Penelitian Tanah dan World Agroforestry Centre (ICRAF), Bogor 2008. 8. Agus, F.; Hairiah, K.; Mulyani, A., Measuring carbon stock in peat soils: practical guidelines. World Agroforestry Centre: 2010. 9. Hooijer, A.; Page, S.; Canadell, J.; Silvius, M.; Kwadijk, J.; Wösten, H.; Jauhiainen, J., Current and future CO2 emissions from drained peatlands in Southeast Asia. Biogeosciences 2010, 7, 1505-1514. 10. Hooijer, A.; Silvius, M.; Wösten, H.; Page, S.; Hooijer, A.; Silvius, M.; Wösten, H.; Page, S., PEAT-CO2. Assessment of CO2 emissions from drained peatlands in SE Asia, Delft Hydraulics report Q 2006, 3943. 11. Carlson, K. M.; Curran, L. M.; Ratnasari, D.; Pittman, A. M.; Soares-Filho, B. S.; Asner, G. P.; Trigg, S. N.; Gaveau, D. A.; Lawrence, D.; Rodrigues, H. O., Committed carbon emissions, deforestation, and community land conversion from oil palm plantation expansion in West Kalimantan, Indonesia. Proceedings of the National Academy of Sciences 2012, 109, (19), 7559-7564. 12. Miettinen, J.; Liew, S., Degradation and development of peatlands in Peninsular Malaysia and in the islands of Sumatra and Borneo since 1990. Land degradation & development 2010, 21, (3), 285-296. 13. Murdiyarso, D.; Dewi, S.; Lawrence, D.; Seymour, F., Indonesia’s forest moratorium: A stepping stone to better forest governance? Cifor: 2011. 14. MoJHRRI [Ministry of Justice and Human Right Republic of Indonesia], Peraturan Pemerintah No. 71/2014. In Direktorat Jenderal Peraturan Perundang-undangan: Jakarta, Indonesia, 2014. 15. BRG [Peat restoration agency] Quarterly Report of the Indonesian Agency for Restoration of Peatlands to the President of the Republic of Indonesia http://www.mongabay.co.id/wp-content/uploads/2016/11/Laporan- triwulan-program-BRG.pdf (12 April), 16. Ritung, S.; Wahyunto, N. K.; Sukarman, H.; Suparto, T. C., Peta Lahan Gambut Indonesia Skala 1: 250.000 (Indonesian peatland map at the scale 1: 250,000). Indonesian Center for Agricultural Land Resources Research and Development, Bogor 2011. 17. Wahyunto; Nugroho, K.; Ritung, S.; Agus, F. In Indonesian peatland map: Method, Certainty, and uses, Prosiding Seminar Nasional Pengelolaan Berkelanjutan Lahan Gambut Terdegradasi untuk Mitigasi Emisi GRK dan Peningkatan Nilai Ekonomi. Indonesian Agency or Agricultural Research and Development, Jakarta, Indonesia, 2014; 2014; pp 81-96. 18. Nugroho, S., the existing of Indonesia National Forest Monitoring System presenting its milestone, objectives, roadmap, process flow and method, accuracy assessment, FREL, burn scar monitoring system, and future improvement. In Training workshop: Land cover map accuracy and training workshop on FAO SEPAL remote sensing tool, FAO: Bogor, 2017. 19. FAO [Food and Agriculture Organization] Report on land cover map accuracy and training workshop on FAO SEPAL remote sensing tool Bogor, Indonesia, 2017. 20. SEEA Experimental Ecosystem Accounting: Technical Recommendations; 2017. 21. Metzger, M.; Rounsevell, M.; Acosta-Michlik, L.; Leemans, R.; Schröter, D., The vulnerability of ecosystem services to land use change. Agriculture, Ecosystems & Environment 2006, 114, (1), 69-85. 22. Sumarga, E.; Hein, L., Mapping ecosystem services for land use planning, the case of Central Kalimantan. Environmental management 2014, 54, (1), 84-97. 23. Indonesia, W., Deforestation, forest degradation, biodiversity loss and CO2 emissions in Riau, Sumatra, Indonesia. One Indonesian Province’s Forest and Peat Soil Carbon loss over a Quarter Cebtury and its Plans for the Future 2008. 24. Huan, P. L. K.; Lim, S. S.; Parish, F.; Suharto, R. Ikhtisar: buku panduan rspo untuk praktik pengelolaan terbaik (ppt) bagi budidaya kelapa sawit sedang berjalan di lahan gambut; RSPO: 2012. 25. Wösten, J.; Ismail, A.; Van Wijk, A., Peat subsidence and its practical implications: a case study in Malaysia. Geoderma 1997, 78, (1), 25-36. 26. Wösten, J.; Clymans, E.; Page, S.; Rieley, J.; Limin, S., Peat–water interrelationships in a tropical peatland ecosystem in Southeast Asia. Catena 2008, 73, (2), 212-224. 27. Rieley, J.; Wüst, R.; Jauhiainen, J.; Page, S.; Wösten, J.; Hooijer, A.; Siegert, E.; Limin, S.; Vasander, H.; Stahlhut, M., Tropical peatlands: carbon stores, carbon gas emissions and contribution to climate change processes. In Peatlands and climate change, International Peat Society: 2008; pp 148-181. 42 28. Agency for Forestry Research and Development, Center for Conservation and Rehabilitation Research and Development, RPI 5 Management of Peat Swamp Forest. In Ministry of Forestry Republic Indonesia: Jakarta, 2010. 29. Harrison, M. E.; Page, S. E.; Limin, S. H., The global impact of Indonesian forest fires. Biologist 2009, 56, (3), 156- 163. 30. Tacconi, L. Fires in Indonesia: causes, costs and policy implications; CIFOR, Bogor, Indonesia: 2003. 31. Van der Werf, G. R.; Randerson, J. T.; Giglio, L.; Collatz, G.; Mu, M.; Kasibhatla, P. S.; Morton, D. C.; DeFries, R.; Jin, Y. v.; van Leeuwen, T. T., Global fire emissions and the contribution of deforestation, savanna, forest, agricultural, and peat fires (1997–2009). Atmospheric Chemistry and Physics 2010, 10, (23), 11707-11735. 32. De Groot, R. S.; Wilson, M. A.; Boumans, R. M., A typology for the classification, description and valuation of ecosystem functions, goods and services. Ecological economics 2002, 41, (3), 393-408. 33. Hein, L. Biophysical Modelling and Analysis of Ecosystem Services in an Ecosystem Accounting Context ; 2014. 34. Gunarso, P.; Hartoyo, M.; Agus, F.; Killeen, T., Oil palm andland usechange in Indonesia, Malaysia and Papua New Guinea. Reports from the Technical Panels of the 2nd greenhouse gas working Group of the Roundtable on Sustainable Palm Oil (RSPO). 2013, 29-62. 35. Sumarga, E.; Hein, L.; Edens, B.; Suwarno, A., Mapping monetary values of ecosystem services in support of developing ecosystem accounts. Ecosystem Services 2015, 12, 71-83. 36. Suwarno, A.; Hein, L.; Sumarga, E., Who benefits from ecosystem services? A case study for Central Kalimantan, Indonesia. Environmental management 2016, 57, (2), 331-344. 37. WRI [World Resources Institute], Carbon emissions from peat drainage, Accessed through Global Forest Watch Climate [climate.globalforestwatch.org]: 2015. In. 38. The World Bank, World Bank Open Data. Availale online: http://data.worldbank.org/: 2016. In. 39. Statistics Indonesia, BPS-Statistics Indonesia. In 2015. 40. Aldrian, E.; Dwi Susanto, R., Identification of three dominant rainfall regions within Indonesia and their relationship to sea surface temperature. International Journal of Climatology 2003, 23, (12), 1435-1452. 41. MoARI [Ministry of Agriculture Republic of Indonesia], Indonesia peat lands, Accessed through Global Forest [www.globalforestwatch.org]. In. 42. MoEFRI [Ministry of Environment and Forestry Republic of Indonesia], Peraturan Direktur Jenderal Planologi Kehutanan Nomor: P.1/VII-IPSDH/2015 Tentang Pedoman Pemantauan Penutupan Lahan. In Ministry of Environment and Forestry Republic of Indonesia: Jakarta, 2015. 43. MoEFRI [Ministry of Environment and Forestry Republic of Indonesia] Peta penutupan lahan Indonesia. http://webgis.menlhk.go.id:8080/pl/pl.htm (15 March), 44. Krisnawati, H.; Imanuddin, R.; Adinugroho, W. C.; Hutabarat, S., Standard Methods for Estimating Greenhouse Gas Emissions from Forests and Peatlands in Indonesia (Version 2). In Research, Development and Innovation Agency of the Ministry of Environment and Forestry: Bogor, Indonesia, 2015. 45. Tosiani, A. Buku kegiatan serapan dan emisi karbon Directorate of inventory and monitoring of forest resources. Directorate general of forestry planology and environmental order. Ministry of Environment and Forestry Republic of Indonesia: Jakarta, 2015. 46. MODIS Collection 6 NRT Hotspot / Active Fire Detections MCD14DL. Available on-line [https://earthdata.nasa.gov/firms].DOI: 10.5067/FIRMS/MODIS/MCD14DL.NRT.006. In. 47. Giglio, L., MODIS Collection 6 Active Fire Product User’s Guide Revision A. Unpublished Manuscript, Department of Geographical Sciences, University of Maryland. Available online: https://lpdaac. usgs. gov/sites/default/files/public/product_documentation/mod14_user_guide. pdf (accessed on 22 February 2016) 2015. 48. MoEFRI [Ministry of Environment and Forestry Republic of Indonesia], National Forest Reference Emission Level for REDD+ In the Context of Decision 1/CP.16 In Directorate General of Climate Change. The Ministry of Environment and Forestry: Indonesia, 2015. 49. MRI Green House Gas Reduction Project through Forest Conservation in Peat land in Central Kalimantan ; Mitsubisi Rerearch Intititute, Inc: 2013. 50. Deputy for Remote Sensing LAPAN, Technical guide (v.01): Informasi titik panas (hotspot) kebakaran hutan/lahan In Center for Remote Sensing Utilization, LAPAN: 2016. 51. Statistics Indonesia, Indeks Harga Konsumen dan Inflasi Bulanan Indonesia. In Indonesia, S., Ed. Indonesia, 2017. 52. BI [Bank Indonesia], Foreign exchange reference rate. In Indonesia, 2017. 53. UN; EC; FAO; IMF, System of Environmental-Economic Accounting 2012: Central Framework. In 2014. 54. Edens, B.; Hein, L., Towards a consistent approach for ecosystem accounting. Ecological Economics 2013, 90, 41- 52. 55. Budiharta, S.; Meijaard, E.; Erskine, P. D.; Rondinini, C.; Pacifici, M.; Wilson, K. A., Restoring degraded tropical forests for carbon and biodiversity. Environmental Research Letters 2014, 9, (11), 114020. 56. MoFRI [Ministry of Forestry Republic of Indonesia], Regulation of the Directorate General of Watershed Management and Social Forestry No. P.6/V-SET/ 2012: Estimation of Standard Costs of Programs for Watershed Management for the Year 2013. In Ministry of Forestry: Jakarta, Indonesia, 2012. 57. Interagency Working Group on Social Cost of Greenhouse Gases, U. S. G., Technical Support Document: Technical Update of the Social Cost of Carbon for Regulatory Impact Analysis Under Executive Order 12866 In 2016. 58. MCD64A1, Modis Active Fire And Burned Area Products. Burned area products. http://modis- fire.umd.edu/pages/BurnedArea.php 59. Hiraishi, T.; Krug, T.; Tanabe, K.; Srivastava, N.; Baasansuren, J.; Fukuda, M.; Troxler, T., 2013 supplement to the 2006 IPCC guidelines for national greenhouse gas inventories: Wetlands. IPCC, Switzerland 2014. 43 60. Ballhorn, U.; Siegert, F.; Mason, M.; Limin, S., Derivation of burn scar depths and estimation of carbon emissions with LIDAR in Indonesian peatlands. Proceedings of the National Academy of Sciences 2009, 106, (50), 21213- 21218. 61. Mulyani, A.; Susanti, E.; Dariah, A.; Maswar, W.; Agus, F., Basis data karakteristik tanah gambut di Indonesia. Badan litbang Pertanian, Kementerian Pertanian Indonesia 2012. 62. Wahyunto; Ritung, S.; Subagjo, H., Peta luas sebaran lahan gambut dan kandungan karbon di Pulau Sumatra 1990– 2002. Bogor, Indonesia: Wetlands International, Indonesia Program and Rapid assessment in Tripa and Batang Toru‐117‐Wildlife Habitat Canada 2003. 63. Wahyunto; Ritung, S.; Subagjo, H., Peta Sebaran Lahan Gambut, Luas dan Kandungan Karbon di Kalimantan/Map of Peatland Distribution Area and Carbon Content in Kalimantan, 2000 –2002. Wetlands International Indonesia Programme and Wildlife Habitat Canada, Bogor, Indonesia 2004. 64. Wahyunto; Heryanto, B.; Bekti, H.; Widiastuti, F., Peta-Peta Sebaran Lahan Gambut, Luas dan Kandungan Karbon di Papua/Maps of Peatland Distribution, Area and Carbon Content in Papua, 2000-2001. Wetlands International– Indonesia Programme & Wildlife Habitat Canada (WHC) 2006. 65. Zubaidah, A.; Vetrita, Y.; Khomarudin, M. R., Validasi Hotspot MODIS di Wilayah Sumatra dan Kalimantan Berdasarkan Data Penginderaan Jauh SPOT-4 Tahun 2012. Jurnal Penginderaan Jauh dan Analisa Citra Digital 2014, 11, (1-14), 114. 66. Zubaidah, A.; Vetrita, Y.; Priyatna, M.; Ayu, K., Analisis pemanfaatan dan validasi hotspot viirs nightfire untuk identifikasi kebakaran hutan dan lahan di indonesia (analysis of use and validation viirs nightfire hotspot for identification of forest and land fire in indonesia). Jurnal Penginderaan Jauh dan Pengolahan Data Citra Digital 2015, 12, (1). 67. Tansey, K.; Beston, J.; Hoscilo, A.; Page, S.; Paredes Hernández, C., Relationship between MODIS fire hot spot count and burned area in a degraded tropical peat swamp forest in Central Kalimantan, Indonesia. Journal of Geophysical Research: Atmospheres 2008, 113, (D23). 68. NRT VIIRS 375 m Active Fire product VNP14IMGT. Available on-line https://earthdata.nasa.gov/firms]. DOI:10.5067/FIRMS/VIIRS/VNP14IMGT.NRT.001. In. 69. MoEFRI [Ministry of Environment and Forestry Republic of Indonesia], National Forest Reference Emission Level for REDD+ In the Context of Decision 1/CP.16 Paragraph 70 UNFCCC. In Directorate General of Climate Change: Indonesia, 2015. 70. Roy, D. P.; Boschetti, L., Southern Africa validation of the MODIS, L3JRC, and GlobCarbon burned-area products. IEEE transactions on Geoscience and Remote Sensing 2009, 47, (4), 1032-1044. 71. Chang, D.; Song, Y., Estimates of biomass burning emissions in tropical Asia based on satellite-derived data. Atmospheric Chemistry and Physics 2010, 10, (5), 2335-2351. 72. Page, S. E.; Siegert, F.; Rieley, J. O.; Boehm, H.-D. V.; Jaya, A.; Limin, S., The amount of carbon released from peat and forest fires in Indonesia during 1997. Nature 2002, 420, (6911), 61-65. 73. FAO [Food and Agriculture Organization], FAOSTAT Emissions Database. In FAO. 74. Rossi, S.; Tubiello, F. N.; Prosperi, P.; Salvatore, M.; Jacobs, H.; Biancalani, R.; House, J. I.; Boschetti, L., FAOSTAT estimates of greenhouse gas emissions from biomass and peat fires. Climatic Change 2016, 135, (3), 699-711. 75. Page, S.; Hosciło, A.; Wösten, H.; Jauhiainen, J.; Silvius, M.; Rieley, J.; Ritzema, H.; Tansey, K.; Graham, L.; Vasander, H., Restoration ecology of lowland tropical peatlands in Southeast Asia: current knowledge and future research directions. Ecosystems 2009, 12, (6), 888-905. 76. Jaenicke, J.; Wösten, H.; Budiman, A.; Siegert, F., Planning hydrological restoration of peatlands in Indonesia to mitigate carbon dioxide emissions. Mitigation and adaptation strategies for global change 2010, 15, (3), 223-239. 77. Lubis, A. U., Bull. Pusat Penelitian Marihat 1990, 10, (2). 78. Risza, I. S., Kelapa sawit, upaya peningkatan produktivitas . Kanisius: 1994. 79. Saragih, J. M., Pengelolaan Lahan Gambut di Perkebunan Kelapa Sawit di Kebun Teluk Bakau, PT Bhumireksa Nusa Sejati, Minamas Plantation, Riau. 2014. 80. Amalia, M.; Syahril, S. Overview of the Payment for the Environmental Services Implementation in Indonesia; Economic and Envrionmental Program for Sout-east Asia-WorldFish: Laguna, Philippines, 2016. 81. Pusat Penelitian dan Pengembangan Perubahan Iklim dan Kebijakan, Cadangan Karbon Pada Berbagai Tipe Hutan dan Jenis Tanaman di Indonesia Seri 2. In Badan Litbang Kehutanan, Kementerian Kehutanan: Bogor, 2014. 82. GOFC-Gold A source book of methods and procedures for monitoring and reporting anthropogenic greenhouse gas emissions and removals caused by deforestation, gains and losses of carbon stocks in forests remaining forests, and forestation.; GOFC-GOLD Project office Natural Resources Canada: Alberta, Canda, 2009. 83. Verwer, C.; Van der Meer, P. Carbon pools in tropical peat forest: towards a reference value for forest biomass carbon in relatively undisturbed peat swamp forests in Southeast Asia; 1566-7197; Alterra: 2010. 84. Delaney, M.; Brown, S.; Lugo, A.; Torres-Lezama, A.; Quintero, N. B., The distribution of organic carbon in major components of forests located in five life zones of Venezuela. Journal of tropical ecology 1997, 13, (5), 697-708. 85. Brown, I. F.; Martinelli, L. A.; Thomas, W. W.; Moreira, M. Z.; Ferreira, C. C.; Victoria, R. A., Uncertainty in the biomass of Amazonian forests: an example from Rondonia, Brazil. Forest Ecology and Management 1995, 75, (1-3), 175- 189. 86. Yoneda, T.; Tamin, R.; Ogino, K., Dynamics of aboveground big woody organs in a foothill dipterocarp forest, West Sumatra, Indonesia. Ecological Research 1990, 5, (1), 111-130. 87. Syahrinudin, The potential of oil palm and forest plantations for carbon sequestration on degraded land in Indonesia: to 25; Pages: 26 to 50; Pages: 51 to 75; Pages: 76 to 100; Pages: 101 to 125 . Cuvillier Verlag: 2005. 44 88. Miettinen, J.; Hooijer, A.; Shi, C.; Tollenaar, D.; Vernimmen, R.; Liew, S.; Page, S., Extent of industrial plantations on Southeast Asian peatlands in 2010 with analysis of historical expansion and future projections. GCB Bioenergy 2012, 4, (6), 908–918. 89. Miettinen, J.; Shi, C.; Liew, S., Land cover distribution in the peatlands of Peninsular Malaysia, Sumatra and Borneo in 2015 with changes since 1990. Global Ecol Conserv 2016, 6, 67–78. 90. Statistics Indonesia, Forestry. In 2004-2014, Jakarta, Indonesia, 2017. 91. Salwati, R.; Purnamayani; Firdaus; Endrizal Respone of oil palm planted on peatland to ameliorants : A case study in Arang-arang, Jambi Balai Pengkajian Teknologi Pertanian Jambi: 2014. 92. Rochmayanto, Y., Perubahan kandungan karbon dan nilai ekonominya pada konversi hutan rawa gambut menjadi hutan tanaman industri pulp. 2009. 93. Ananto, E. E.; Pasandaran, E. Pengelolaan lahan gambut di provinsi Sumatra Selatan; Badan Penelitian dan Pengembangan Pertanian: 2007; pp 194-211. 94. Ananto, E. E.; Supriyo, A.; Soentoro; Hermanto; Sulaeman, Y.; S, I. W.; Nuryanto., B., Pengembangan usaha pertanian lahan pasang surut Sumatra Selatan: Mendukung Ketahanan Pangan dan Pengembangan Agribisnis. In Badan Litbang Pertanian: 2000; p 166. 95. Suzuki, S.; Ishida, T.; Nagano, T.; Waijaroen, S., Influences of deforestation on carbon balance in a natural tropical peat swampforest in Thailand. Environ Control Biol 1999, 37, 115–128. 96. Jauhiainen, J.; Hooijer, A.; Page, S., Carbon dioxide emissions from an Acacia plantation on peatland in Sumatra, Indonesia. Biogeosciences. 2012. 97. Hirano, T.; Segah, H.; Harada, T.; Limin, S.; June, T.; Hirata, R.; Osaki, M., Carbon dioxide balance of a tropical peat swamp forest in Kalimantan, Indonesia. Global Change Biol 2007, 13, (2), 412–425. 98. MoTRI [Ministry of Trade Republic of Indonesia], Ministry of Trade regulation (Permendag) No. 22 year 2012 about the amendment to the Regulation of Minister of Trade No. 12/M-DAG/PER/3/2012 on pricing of forest product benchmark for the calculation of forest resources provision. In Ministry of Trade Republic of Indonesia: Indonesia, 2012. 99. Directorate General of Estate Crops, Tree crop estate statistics of Indonesia 2013-2015: Oil palm. In Ministry of Agriculture: Jakarta, Indonesia, 2014. 100. Statistics Indonesia, Producer Price Indices. In 2017. 101. TRMM, Tropical Rainfall Measuring Mission (TRMM) (TMPA/3B43) Rainfall Estimate L3 1 month 0.25 degree x 0.25 degree V7, Greenbelt, MD, Goddard Earth Sciences Data and Information Services Center (GES DISC). In 2011. 102. Climate Prediction Center / NCEP, ENSO: recent evolution, current status and predictions. In 2017. 103. Sofan, P., Karakteristik pola curah hujan di wilayah Indonesia berdasarkan data satelit TRMM (1998-2011). Bunga Rampai Penginderaan Jauh Indonesia 2013 2015, 2-20. 104. Pebesma, E.; Graeler, B.; Pebesma, M. E., Package ‘gstat’. 2018. 105. Isaaks, E. H.; Srivastava, M. R., Applied Geostatistics. 1989. 106. Furukawa, Y.; Inubushi, K.; Ali, M.; Itang, A.; Tsuruta, H., Effect of changing groundwater levels caused by land-use changes on greenhouse gas fluxes from tropical peat lands. Nutrient Cycling in Agroecosystems 2005, 71, (1), 81- 91. 45 Annexes Annex 1. Estimated carbon stocks in vegetation by land cover type Carbon stocks (tC) in Land cover AGB Root Dead wood Litter Primary peatland forest Sumatra 157.1a 27.7 a 1.95 a 22.75 a Kalimantan 125.95 a 34.55 a 2.45 a 28.4 a Secondary peatland forest Sumatra 96.15 a 19.2 a 1.95 a 20.1 a Kalimantan 91.65 a 21.65 a 2.2 a 22.65 a Plantation forest Sumatra 76.7 b 18.2 d n.a. 2.9 e Kalimantan 54.7 b 18.2 d n.a. 2.9 e Perennial crops 63 c 11 d 5.2 f 1.8 e Settlement 4c 1d n.a n.a Transmigration areas 10 c 1d n.a n.a Bare ground 2.5 c 4g n.a n.a Savannah 4c 2d n.a n.a Dry agricultural land 10 c 1.0 d n.a n.a Dry agricultural land (mixed) 30 c 1.0 d n.a n.a Paddy field 2c 1.0 d n.a n.a Open water/mining area/open swamp 0c n.a n.a n.a Shrub 30 c 3.9 d 1.8 f 0.6 e Notes: n.a: data are not available. a: 44; 48. b: 81 as reviewed in 45. c: Juknis PEP RAD GRK (2013) as reviewed in 45. d: Estimated from above ground carbon based on the ratio from GOFC-Gold 82 as reviewed in 83. e: Estimated from above ground carbon based on the average ratio from 84 and 85 as reviewed in 83. f: Estimated from above ground carbon based on the ratio from 86 as reviewed in 83. g : 87 46 Annex 2. Hotspots 2.A Spatial distribution of hotspots (CV ≥ 80%) in 2006. 2009 and 2014 Data sources: peatland map 41 and fire product MODIS 46 47 2.B Number of hotspots in Sumatra and Kalimantan peatlands 2006, 2009, and 2014 estimated from the MODIS fire product Number of At minimum confidence level (≥) hotspots in 0 30 80 peatlands 2006 2009 2014 2006 2009 2014 2006 2009 2014 Sumatra 8638 5856 10246 8236 5538 9777 4035 2448 5663 Kalimantan 8456 5904 5778 8005 5638 5527 3879 2619 2635 Total 17094 11760 16024 16241 11176 15304 7914 5067 8298 Note: hotspots in the same or nearest location (approximately 1 km resolution) detected in the same year are assumed as one hotspot. Annex 3. Estimated total area of oil palm plantation in Sumatra and Kalimantan peatlands. Oil palm plantation (ha) in Year Estimated total peatlands Source Sumatra Kalimantan Papua Total 2000* Indonesia (20.8 Mha). 700000 1000 0 701000 34 Sumatra (7.21 Mha). Kalimantan (5.83 Mha). Papua (7.8 Mha) Sumatra (7.23 Mha). 512341 15982 n.a 528323 88 Kalimantan (5.77 Mha) 2005* Indonesia (20.8 Mha). 1200000 50000 1500 1251500 34 Sumatra (7.21 Mha). Kalimantan (5.83 Mha). Papua (7.8 Mha) 2010* Indonesia (20.8 Mha). 1400000 308000 1700 1709700 34 Sumatra (7.21 Mha). Kalimantan (5.83 Mha). Papua (7.8 Mha) Sumatra (7.23 Mha). 1026922 258299 n.a 1285221 88 Kalimantan (5.77 Mha) 2014 Indonesia (14.9 Mha). 1413628 479816 1850 1895294 37 Sumatra (6.4 Mha). Kalimantan (4.9 Mha). Papua (3.6 Mha) 2015* Sumatra (7.23 Mha). 1315830 730750 n.a 2046580 89 Kalimantan (5.78 Mha) *data were taken from 6 48 Annex 4. Annual production rate of ecosystem services in peatlands of Sumatra and Kalimantan Peatlands in ES specification Unit Source Sumatra Kalimantan Timber production m3/ha 0.73 * 0.37* Statistics Indonesia 90 Oil palm production ton FFB/ha 16.7 14.8 79, 35, 91 Biomass production ton biomass/ha 21.03 21.03 92 for pulp Paddy production ton paddy/ha 2.92 1.7 36, 93, 94 CO2 sequestration ton CO2/ha 19 19 95 as reviewed in 6 *average value 2004-2015 Note: the values might vary at provincial or district level Annex 5. Estimated net carbon flux by land cover type Land cover Net carbon (CO2) flux Source Undisturbed natural forest 19 9, 10, 95-97 Disturbed natural forest -17 Forest plantation (Acacia) -81 Oil palm. other types of plantation -85 Dry land agriculture -48 Paddy field -48 Open water 0 Degraded land. bare ground -15 Other land uses -15 Note: data are taken from 6 Annex 6. Production cost and market price of commodities ES Indicator Unit Value Value at Source Timber Price 1000 IDR/m3 1658.5a 2012 98 production Production cost 1000 IDR/m3 991 2010 36 Oil palm Price 1000 IDR/tFFB 1102 2013 99 production Production cost 1000 IDR/ha 15424 2010 35 Acacia biomass Price 1000 IDR/t 350 2009 92 production Production cost 1000 IDR/ha 5459 2009 92 Paddy production Price 1000 IDR/ton 3.3 2013 100 Production cost 1000 IDR/ha 2652 2010 35 CO2 sequestration Social cost of US$/tonCO2 36b 2007 57 carbon Protected habitat Restoration cost US$/ha 1054 2014 55 a: b: average timber (of several types) price. value with discount rate at 3% Note: the values, particularly the production cost, might vary at provincial or district level. In the calculation of monetary values, all data were converted to the values at 2017 (considering inflation rates 6.96 in 2010; 3.79 in 2011; 4.3 in 2012; 8.38 in 2013; 8.36 in 2014; 3.35 in 2015; 3.02 in 2016 51, and exchange rates based on Bank Indonesia 52. 49 Annex 7. Spatial distribution of ecosystem services (physical values) in Sumatra and Kalimantan peatlands for 2000 and 2014 50 51 Annex 8. Comparison of total area and spatial distribution of plantation in peatlands 8.1 Plantation in peatlands of Sumatra and Kalimantan 2014 a a b b a: plantation map from MoEFRI land cover 2014 (forest plantation and perennial crops) b: plantation map 2013/2014 from WRI, accessed through Global Forest Watch 37 Table 15. Total plantation area in peatlands 2014 (MoEFRI and WRI maps) Total area (1000 ha) in Source Sumatra Kalimantan MoEFRI land cover map 2014* 2262 366 WRI37 3307 749 *for forest plantation and perennial crops 52 8.2 Plantation (oil palm) in peatlands of Central Kalimantan 2012 a b a: plantation map from MoEFRI land cover 2012 (perennial crops) b: oil palm plantation map 2012 from OFI (Orangutan Foundation Indonesia) Table 16. Total plantation area in peatlands of Central Kalimantan 2012 Total area (1000 ha) Source Plantation (perennial crops) 112 MoEFRI land cover map 2012* Plantation (oil palm) 290 OFI (2012) *land cover of perennial crops 53 Annex 9. Burned peatlands in Sumatra and Kalimantan for 2006. 2009 and 2014 Data sources: peatland map 41 and burned area product MODIS 58 54 Annex 10. Precipitation Introduction: Peat swamp forest is one of the main water storages for the surrounding environment. The water content in peatland has various functions for the ecosystem such as preventing land subsidence, flood and fire 26. Water supply is one of the relevant indicators that can be added to the ecosystem condition account. In this annex, the precipitation is presented as a basic information to monitor the main element (water supply) of the hydrological system in peatlands. Monitoring precipitation is also essential to understand the rainfall pattern in the context of climate variability (e.g. ENSO and IOD), as well as to understand the changes in fire risk. However, at the same time, it needs to be recognised that precipitation is not dependent upon human management of peatlands or reflecting the environmental state of peatlands. It is therefore an indicator of the overall environment (indicating impacts of climate change on peatlands) rather than a specific peatland related indicator. Method: The Tropical Rainfall Measuring Mission (TRMM) data organised by NASA and Japan National Space Development Agency were used for identifying the annual rainfall of Indonesian peatlands. Due to research need and data availability. this study extracted monthly precipitation data from TRMM 3B43. a precipitation estimate resolution 0.25 degree V7 101. This data has been available from January 1998 until present. The rate of precipitation from TRMM product was extracted and accumulated to estimate the total precipitation annually using spatial analysis (ArcMap 10.5) tools. Firstly. data (.NetCDF) were converted into raster. Then, to cover the whole study area that is distributed in a small resolution (compared to the resolution of TRMM products). the precipitation raster map was resampled to 100 metres gridded raster. The resample raster then was extracted by mask with peatland distribution maps from 41. Raster calculator tool and zonal statistic tool were applied to get the annual precipitation of peatlands for every province. The average of annual precipitation was calculated using Equation 1. ∑( × ) = ∑ Equation 5 Results: The average of annual precipitation (at island level) in peatlands is presented in Table 17. following the same time-period as in the ecosystem extent account. From these data. it is estimated that Indonesian peatlands received precipitation varied between 1849-2941 mm/year in Sumatra and between 2610-4031 mm/year in Kalimantan. The year of 2015. known as one of El Niño years, a phenomenon resulting in warmer climate condition in Indonesia 102 (see details in Annex 11), is recorded as a dry year with the lower precipitation. Meanwhile. peatlands received a high amount of rainfall in 2010 when La Niña, an event resulting in cooler climate condition in Indonesia (mainly in seven last months), was detected. Table 17. The average of annual precipitation in Sumatra and Kalimantan peatlands . Indicator Peatlands in 1990 1996 2000 2006 2009 2010 2014 2015 Average Sumatra n.a n.a 2633 2571 2520 2941 2119 1849 precipitation Kalimantan n.a n.a 3056 2610 2740 4031 2757 2804 Data source: TRMM 101. n.a: data are not available It is also noticed that the average of rainfall is significantly varied in spatial distribution. As seen in Figure 11, for example, peatlands in Sumatra during 2014, received less rainfall below 2000 mm/year particularly in the eastern part (Riau, Jambi, South Sumatra, Riau islands, and Bangka- 55 Belitung provinces). On the other hand. most of the peat areas in Kalimantan and western Sumatra had quite high precipitation which was more than 3000 mm/year. Figure 14. Spatial distribution of precipitation in Sumatra and Kalimantan peatlands in 2014. estimated from TRMM (2011). Data validation process is vital to be considered since there would be some differences in values when they are compared to observation or survey data. The coefficient correlation between TRMM precipitation product for Indonesia and the observation data from BMKG (Meteorological, Climatological, and Geophysical Agency) is 0.8 103. This number is relatively high although there are still 20% of possibility for error estimation. 56 Annex 11. Oceanic Niño index 1950-2017 Data source: NCEP 102. Warm and cold periods based on a threshold of +/- 0.50C for the Oceanic Niño Index (ONI) [3 month running mean of ERSST.v5 SST anomalies in the Niño 3.4 region (5 0N-50S, 1200W-1700W)], based on centred 30-year base periods updated every 5 years. Annex 12. Estimated emission factor and biomass consumption of burned peat Data taken from Indicator MoEFRI 48 IPCC supplement 2013 as reviewed in 74 Mass of fuel available for combustion 504.9a 353b (tdm/ha) Emission factor (CO2 kg/ton) 1828.2 1703 aEstimated as 33-cm burned peat depth in forest bEstimated for drained peatland 57 Annex 13. Kriging interpolation method A kriging interpolation technique was applied to map peat drainage depth in certain land uses (perennial crop area, degraded peat swamp forest in the distance less than 500 m from perennial crop area, plantation forest, and bare land), based on geostatistical analysis of available drainage depth data. Geostatistics is a general term for a spatial model of spatial structure that considers the spatial autocorrelation between values in the sampled locations, where these models can then be used to predict the values of unsampled locations. The procedures of the krigring interpolation technique is briefly described in Figure 12. The procedures were separately applied for Sumatra and Kalimantan. Figure 15. Flow chart of kriging interpolation technique At the first step, the spatial structure of the drainage depth data was analyzed using variogram analysis. Variogram is a scatter plot describing the relationship between semivariance and distance. Ideally, the increase of separation between point-pairs will also be followed by the increase of semivariance until a certain separation called the range. In total 166 drainage data from Sumatra and 40 data from Kalimantan were analyzed. The gstat package of “R” 104 was used for this analysis. After fitting some variogram models to the sample variogram, the parameters of the best model (partial sill, range, and nugget) were used for kriging interpolation. Figure 13 and 14 show the selected variogram model for Sumatra and Kalimantan. Ordinary kriging was selected as interpolation method. Ordinary kriging is a best linear unbiased predictor since it tries to have a mean error of zero and to minimize the variance error 105. An interpolation tool in spatial analysis tools of ArcMap was used for the interpolation to produce the map of peat drainage depth. Finally, a cross validation (leave-one-out method) was applied to analyze model accuracy by calculating coefficient of variation (CV) of root mean square error (RMSE). This value represents the deviation of the prediction error from the mean of input data, whose value ranges from 0 to 1. A CV of RMSE of 0 indicates a perfect accuracy. The cross validation gave a CV of RMSE of 15.1% for Sumatra and 19.1% for Kalimantan. 58 185 600 71 500 5 10 400 semivariance 883 300 200 396 1 8 1147 525 280 100 1684 86 1753 429 0.5 1.0 1.5 2.0 2.5 distance Figure 16. Selected variogram model for Sumatra Figure 17. Selected variogram model for Kalimantan with parameters: partial sill 187.35, range 1.9 and with parameters: partial sill 457.15, range 1.1 and nugget 56.39 nugget 85.8 Annex 14. Selected water table depth data Land cover Water level (m) Undisturbed natural forest 0* Disturbed natural forest in the distance more than 500 m from 0* perennial crops or plantation forests Mixed cropland; small-scale 0.6a agriculture Shrubland 0.33a Lowland paddy field 0.05b Others (water, savanna, etc.) 0* Note: * 0 means undrained Sources: a: 9, b: 106 59