Kenya National REDD+ Programme. Roadmap for the establishment of. reference levels. and. national forest monitoring system

Transcription

1 Kenya National REDD+ Programme Roadmap for the establishment of reference levels and national forest monitoring system

2 This document was prepared by Patrick Van Laake International Consultant on Forests and Climate Change For the Ministry of Forests and Wildlife / National REDD+ Coordinating Office With financial support from the Miti Mingi Maisha Bora project

3 Table of Contents List of Acronyms... ii 1 Introduction Objectives Principles Development approach Task details Key documents on the national forest monitoring system and reference levels Official documents from the UNFCCC Other relevant documents on specific issues Kenyan context Technical background IPCC Guidelines key concepts REDD+ activities and sampling Data quality assessment Roadmap Preliminaries National forest monitoring system Reference Levels Official status Terminology IPCC methodologies related to reference levels Proposals for the establishment of reference levels Conclusion Appendix: Land use area matrices and sub-matrices i

4 List of Acronyms AFOLU AGB BAU BGB CATIE CFA CH 4 CO 2 CO 2 e CoP DAF DBH DOM DRSRS FAO GHG GWP H HWP IPCC KEFRI KENCAS KFS KNBS MFW NGO N 2 O NO x NRCO PCM PDF PFM PSP RCMRD REDD+ REL RL SiHA SO 2 SOM UNEP UNFCCC WMO Agriculture, Forestry and Other Land Use scope of IPCC reporting Above-ground biomass Business-as-usual Below-ground biomass Center for Tropical Agronomical Research and Training Community Forest Association Methane Carbon dioxide CO 2 -equivalent Conference of the Parties to the UNFCCC Development adjustment factor Diameter at breast height Dead organic matter Department of Resource Survey and Remote Sensing Food and Agriculture Organization of the United Nations Greenhouse gas Global Warming Potential Height of the bole of a tree Harvested wood products Intergovernmental Panel on Climate Change Kenya Forestry Research Institute Kenya National Carbon Accounting System Kenya Forest Service Kenya National Bureau of Statistics Ministry of Forestry and Wildlife Non-governmental organization Nitrous oxide Nitrogen oxides National REDD+ Coordination Office Participatory carbon monitoring Probability density function Participatory forest monitoring Permanent sampling plot Regional Center for Mapping of Resources for Development Reducing emissions from deforestation and forest degradation and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks in developing countries Reference emission level Reference level Simple Historical Approach Sulfur dioxide Soil organic matter United Nations Environment Programme United Nations Framework Convention on Climate Change World Meteorological Organization ii

5 1 Introduction 1.1 Objectives The Government of Kenya is in the process of establishing a National REDD+ Programme. Two critical elements of the National REDD+ Programme are forest reference levels and a national forest monitoring system. This document provides the first comprehensive steps towards establishing the design of the national forest monitoring system and reference levels, through an enumeration and detailed description of required tasks. 1.2 Principles Activities under the roadmap will be developed and implemented to be consistent with a number of key principles that comply with Kenya s legal and customary systems and with Decisions from the Conference of the Parties (CoP) to the United Nations Framework Convention on Climate Change (UNFCCC): Inclusive of all relevant aspects of a National REDD+ Programme. The roadmap will consider how the national forest monitoring system and reference levels fit into a larger picture of REDD+ implementation and national accounting of greenhouse gas emissions. Alignment with UNFCCC decisions. The roadmap recognizes all relevant decisions of the CoP with regards to the national forest monitoring system and reference levels. Consultation with all relevant stakeholders. Consultations on the development and implementation of this roadmap will be held within the government agencies responsible for REDD+ management, but also with stakeholders in other areas of government, lower administrative levels in forestry and other areas of government, universities and research institutions, non-governmental organizations, private sector parties and local communities living in or near forests. Recognizing cross-cutting issues. Governance in forestry is considered a crucial element for the successful implementation of a National REDD+ Programme and it permeates all areas of intervention. Gender equality will be respected and fostered throughout the development and implementation of this roadmap and the National REDD+ Programme itself. Recognizing secondary benefits. A successful National REDD+ Programme will recognize the need to sustain the livelihoods of the communities living in or near the forest, as well as the overall economic development of Kenya. Many secondary benefits in timber and non-timber forest products can be had from the results of the National REDD+ Programme that supplement the core benefits that will accrue to the local communities for their efforts in implementing REDD+. Other, non-tangible or non-monetary benefits can be had as well, such as enhanced protection of water towers and the preservation and enhancement of biodiversity. The national forest monitoring system must facilitate the monitoring and quantification of these secondary benefits. 1

6 1.3 Development approach The development of the roadmap was coordinated by the Ministry of Forestry and Wildlife, specifically the National REDD+ Coordination Office. In the development of the Roadmap the following key considerations will be adhered to: Strategic. This roadmap is a document that drives further design, development and implementation activities. It is intended to draw the outlines of what needs to be done, by whom, when and how. It will recognize the new Constitution of Kenya and the process of restructuring the administration of the forestry sector. Cooperative. The Government of Kenya will work closely together with all relevant stakeholders to develop and implement the roadmap. These stakeholders include the international community, multi-lateral agencies and the donor community at the international level and all interested parties in Kenya domestically. Multi-phased. The roadmap and its activities will be developed and implemented over three key phases, in concert with the decision of the Conference of the Parties to the UNFCCC (Decision 1/CP.16, paragraph 73): 1. Formulating policies for the national forest monitoring system and reference levels and building capacity for designing and implementing these. 2. Implementation of the national forest monitoring system and construction of the reference levels and capacity building for operators at all levels of stakeholders. 3. Full operation of the national forest management system. Action-oriented. Activities will be designed to deliver concrete and tangible outcomes and will be supported with capacity building initiatives to ensure effective on-the-ground implementation. Realistic. Activities will be designed and implemented in a practical and realistic manner that reflects the national priorities and institutional capacity of Kenya. Time-bound. Each activity will be given an indicative timeframe, where possible, to track and monitor the implementation of the roadmap. 1.4 Task details The current document is a first enumeration of tasks related to reference levels and the national forest monitoring system, with comments from relevant stakeholders integrated. Comments have been collected following a distribution of the first draft to all relevant stakeholders and during a workshop with all relevant stakeholders on 18 July This version of the roadmap is the result of this consultation round. Further work will be conducted during the remainder of 2012, particularly focusing on practical issues of undertaking design and implementation of the identified tasks. The tasks are described in detail in the remainder of this document. Given that this is the first elaboration of the roadmap, some of the details for the tasks are still quite uncertain. This related primarily to the list of stakeholders, the implementation schedule and the required budget. 2

7 Stakeholders. The list of stakeholders is indicative only and it needs to be based on a comprehensive stakeholder analysis for each task. It is recommended that each task is assigned a lead agency, responsible for maintaining progress in the detailed design of the task, coordination with other stakeholders, and implementation. Contact and discussions with third parties e.g. potential donors or sponsors will be coordinated with the National REDD+ Coordination Office. Operation of the task, once it is implemented, may be assigned to the same lead agency and group of stakeholders or an alternative arrangement may emerge from the design and implementation, to be decided by the National REDD+ Coordinating Office and other competent authorities, in consultation with the relevant stakeholders. Schedule. The schedule is indicative only, as implementation is dependent upon external support for design and implementation. Of more interest than the specific timing of individual tasks is the sequence in which certain tasks need to be implemented. As an example, the definition of forest (task 0.1) and the preparation of a set of standards and protocols for data collection, processing and management (task 0.2) are critical for the design and implementation of many other tasks. These should therefore be undertaken with priority and completed before designing and implementing dependent tasks. Budget. For design and especially implementation of many of the tasks it is difficult to determine the required budget, as this is dependent upon design decisions and a capacity building needs assessment of the stakeholders implementing and operating the task. For most tasks there is currently no or limited funding available. It may be expected that this roadmap will provide a framework to secure (additional) funding for the design and implementation of the tasks. Operation of the tasks is expected to have its separate budgets later on, specifically from the National REDD+ Programme once it is fully operational; the task details do not indicate any such operational budgets but these obviously need to be budgeted for e.g. staff time, depreciation, maintenance and replacement of equipment and other inputs, consumables, insurance, etc. 1.5 Key documents on the national forest monitoring system and reference levels Official documents from the UNFCCC United Nations Framework Convention on Climate Change. Official text of the Convention, with details on reporting in Article 4. Document FCCC/INFORMAL/84. Conference of the Parties to the UNFCCC, eleventh session, December 2005: Reducing emissions from deforestation in developing countries: approaches to stimulate action (the first submission on REDD to the UNFCCC); submission by Papua New Guinea and Costa Rica on behalf of the Coalition for Rainforest Nations. Document FCCC/CP/2005/MISC.1. Decisions of the Conference of the Parties to the UNFCCC on REDD+: o Decision 1/CP.13: Bali Action Plan, with reference to all five eligible activities for REDD+. Document FCCC/CP/2007/6/Add.1. 3

8 o o o o Decision 2/CP.13: Reducing emissions from deforestation in developing countries: approaches to stimulate action. First decision on REDD, calling for demonstration activities. Document FCCC/CP/2007/6/Add.1. Decision 4/CP.15: Methodological guidance for activities relating to reducing emissions from deforestation and forest degradation and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks in developing countries. Specific references to the national forest monitoring system. Document FCCC/CP/2009/11/Add.1. Decision 1/CP.16, section C: Policy approaches and positive incentives on issues relating to reducing emissions from deforestation and forest degradation in developing countries; and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks in developing countries. Introduction of safeguards, reiteration of requirements for the national forest monitoring system. Document FCCC/CP/2010/7/Add.1. Decision 12/CP.17: Guidance on systems for providing information on how safeguards are addressed and respected and modalities relating to forest reference emission levels and forest reference levels as referred to in decision 1/CP.16. Details on preparation and submission of reference levels and guidance on providing information on safeguards. Document FCCC/CP/2011/9/Add.2. There are many unofficial documents as well, such as submissions by observers on specific issues. All of these are available from the UNFCCC web site Other relevant documents on specific issues Meridian Institute: Modalities for REDD+ Reference Levels: Technical and Procedural Issues. Available for download at Meridian Institute: Guidelines for REDD+ Reference Levels: Principles and Recommendations. Available for download at UN-REDD guidance on governance in forestry. World Bank SESA methodology Kenyan context Constitution of 2010, with devolution of authority to lower levels of government and engagement of local communities in resource management. Vision 2030, the long-term strategy of the Government of Kenya. The new Forest Bill, currently under development The Kenya Climate Change Response Strategy Kenya REDD+ Readiness Preparation Proposal 4

9 2 Technical background All reports on greenhouse gas emissions and removals that countries submit to the UNFCCC have to follow the methodology adopted by the IPCC. Like the UNFCCC itself, IPCC is an international body made up of representatives of Parties to the Panel virtually all the countries in the world. The IPCC was established under the auspices of the World Meteorological Organization (WMO) and the United Nations Environment Programme (UNEP) to assess the extent of climate change, its causes and its effects. Additionally, it publishes guidelines and guidance for the reporting of anthropogenic greenhouse gas emissions and removals. The UNFCCC has formally adopted the 1996 IPCC Guidelines for National Greenhouse Gas Inventories for the reporting of National Communications. In more recent Decisions, the UNFCCC has encouraged to use the most recent Intergovernmental Panel on Climate Change guidance and guidelines, as adopted or encouraged by the Conference of the Parties 1, being the 2006 IPCC Guidelines for National Greenhouse Gas Emissions. The guidelines can be downloaded for free from the internet 2. All reporting for REDD+ will in all likelihood be done through the National Communication of the individual countries participating in the REDD+ mechanism. Even if a separate reporting mechanism is eventually mandated, the IPCC methodology will in all likelihood apply. Of particular interest are Volume 1 on general issues and Volume 4: Agriculture, Forestry and Other Land Uses (AFOLU). 2.1 IPCC Guidelines key concepts This section describes the key concepts that the IPCC Guidelines use in the estimation of emissions and removals of greenhouse gases from anthropogenic origin. Adherence to these concepts is important in order to apply the same principles in REDD+, guaranteeing consistency with the national forest monitoring system for REDD+ and the National Communications to the UNFCCC. Tiers In order to accommodate different levels of information, knowledge and inventory capacity in different countries, the IPCC defines three Tiers for the estimation of emissions and removals of greenhouse gases; see the below Box. Increasing Tiers imply increasing accuracy of the estimates due to an increased use of local data and more evolved estimation methods. Estimates based on higher Tier methods and parameters have a higher (implied) confidence and are thus able to generate higher amounts of results-based benefits. Tier 1 estimates using default parameter values given in the IPCC Guidelines should only be used if there is no other option. For REDD+ this will typically be the case for the soil organic matter and below-ground biomass carbon pools. Application of Tier 1 default parameters will result in a conservative estimate of the size of the carbon pool meaning that the size of the pool is estimated in such a way that the actual, but unknown, amount of the carbon pool is likely to be larger. 1 UNFCCC Decision 4/CP.15 Methodological guidance for activities relating to reducing emissions from deforestation and forest degradation and the role of conservation, sustainable management of forests and enhancement of forest carbon stocks in developing countries, paragraph 1(c). UNFCCC document FCCC/CP/2009/11/Add

10 Box 2-1: Framework of Tier structure for AFOLU methods. Tier 1 methods are designed to be the simplest to use, for which equations and default parameter values (e.g. emission factors) are provided in the IPCC Guidelines. Country-specific activity data are needed, but for Tier 1 there are often globally available sources of activity data estimates (e.g. deforestation rates, global land cover maps, fertilizer use, etc.), although these data are usually spatially coarse. Tier 2 can use the same methodological approach as Tier 1, but applies emission factors that are based on country- or region-specific data, for the most important land-use categories. Country-defined emission factors are more appropriate for the climatic regions and land-use systems in the country. Higher temporal and spatial resolution and more disaggregated activity data are typically used in Tier 2 to correspond with country-defined coefficients for specific regions and specialized land-use categories. At Tier 3, higher order methods are used, including models and inventory measurement systems tailored to address national circumstances, repeated over time, and driven by high-resolution activity data and disaggregated at sub-national level. These higher order methods provide estimates of greater certainty than lower tiers. Such systems may include comprehensive field sampling repeated at regular time intervals and/or GIS-based systems of age, soils data, and land-use and management activity data, integrating several types of monitoring. Pieces of land where a land-use change occurs can usually be tracked over time, at least statistically. In most cases, these systems have a climate dependency and thus provide source estimates with interannual variability. Models should undergo quality checks, audits, and validations and be thoroughly documented. Adapted from: 2006 IPCC Guidelines, Volume 4, Chapter 1, Box 1.1 Tier 2 estimates apply nationally-appropriate methods and/or parameters. This is the recommended minimum Tier for REDD+. It implies that there is detailed information on the forest land and the forest types in the country and that at least basic information is available about species composition and merchantable timber volume for the different forest types. Countries should aim for this Tier in the estimation of at least above-ground biomass and preferable also for dead wood and litter carbon pools; if soil organic matter is an important carbon pool under specific conditions such as in mangroves use of Tier 2 estimation methods in those circumstances is recommended too. Tier 3 methods are the most complex in terms of monitoring of status and processes (e.g. changes in land use in a spatially-explicit manner), knowledge of biophysical processes and required data to make the estimates. The IPCC Guidelines do not present any methods for Tier 3 estimation, only good practice guidance. For REDD+ purposes, a Tier 3 method could be the development and application of an allometric equation for the estimation of one or more carbon pool(s) on the basis of the measurement of a few key properties of a tree or a forest stand. As such, Tier 3 methods are closely aligned with participatory approaches involving a large number of local communities: the method is derived for local conditions and the many measurements provide a detailed spatial resolution of the estimate. Despite the cost and organizational burden of implementing Tier 3 6

11 estimation, this is the recommended target as it yields the best estimate with the highest confidence; the additional cost is likely to be more than offset by the higher estimate compared to lower Tier estimates (which are conservative) and the higher benefits that such estimates may generate in the international market. Estimation methods from different tiers may be freely combined within a reporting period or over time, as long as proper documentation of estimation methods is provided. Land-use categories The IPCC identifies six land-use categories for estimation of emissions and removals of greenhouse gases of anthropogenic origin: 1. Forest Land Land that supports tree cover following the national definition of forest, which is typically expressed in terms of a minimum area, minimum height at maturity and minimum crown cover. This also includes land that is temporarily unstocked (e.g. after harvesting or a major disturbance), but which is expected to revert to tree cover under the definition of forest. Depending on the national definition, land covered with certain non-tree species such as bamboo or palms may be considered as forest. Some types of tree production systems such as coffee or rubber plantations may also be counted as forest. 2. Cropland Land for the production of agricultural crops. This includes agroforestry systems under the IPCC description, but land under agroforestry use that is dominated by tree cover may also be counted as Forest Land. 3. Grassland Pastures and rangeland that is not Cropland. Land with sparse woody vegetation (e.g. canopy cover is too low to be considered Forest Land) is part of this category. 4. Wetlands Land that is perennially or seasonally inundated and which is not accounted under any of the above categories. It includes natural and artificial water courses and bodies (e.g. rivers, lakes and reservoirs). 5. Settlements Land which supports human habitation (cities, villages) and transportation and other infrastructure, according to national definitions. 6. Other Land Land which does not fall into any of the above categories bare (degraded) land, bare rocks, desert, snow and ice. The land-use categories may be further subdivided for estimation of emissions and removals and reporting purposes. IPCC does not prescribe or require any subdivisions, but it does suggest that these may be applied in order to facilitate the estimation and reporting. Examples of such subdivisions related to forest and tree cover are: Land-use category Forest Land Cropland Grassland Settlements Sub-category Cattle grazing Plantations Charcoal burning Agroforestry Perennial crops Woodland Tree-lined road 7

12 The area reported for these six land-use categories, including any sub-categories, should sum up to the total area of the country. Care should be taken therefore not to double-count any land use, such as a tea plantation as a tree crop under Forest Land and as an agricultural production system under Cropland. The exact definition of the land-use categories should be left to the assigned national authority to report on greenhouse gas emissions and removals from all sectors in the country, the Ministry of Environment and Natural Resources in Kenya. Estimation of emissions and removals of greenhouses gases is only done for areas of managed land, that land which is subject to human impacts for production, ecological or social functions. That implies that REDD+ can only be applied to those areas of forest land which are classified as managed (otherwise there is no estimation and reporting and thus no basis for generating results-based benefits). In general, any forest which is generally accessible to people can be classified as managed management is not restricted to formal types of management under control or license of the state or by private ownership. Both managed and unmanaged land should be accounted for in National Communications, even if only as a validity check on the areas reported. In Kenya it is more than likely that all Forest Land can be classified as managed. An example of how to construct matrices for reporting on changes between land use categories is given in the Appendix to this document. Stratification Land-use (sub) categories can be stratified in order to obtain more homogeneous units for which separate estimation methodologies or conversion factors exist. The IPCC suggests a stratification of the Forest Land category along ecological zones and climatic domains (between which one may expect a large correlation). Further stratifications can be made on forest type e.g. broad-leaf, needle leaf, deciduous, evergreen, mangrove, plantation. Stratification should strike a careful balance between the overhead costs of having to estimate emissions and removals of greenhouses gases from every stratum, with its concomitant burden of establishing an estimation method and/or set of conversion parameters for each stratum, and the (assumed) higher accuracy of the estimate that is obtained. In general, stratification should at least be applied to the forest type and preferably also on ecological zone if the country has distinct zones based on topography or climate. Stratification is considered good practice by the IPCC. In Kenya stratification can be based ecological and floristic characteristics, using as a basis for instance the Kenya Vegetation Map of Care must be taken not to confuse actual vegetation from potential vegetation that is represented in an ecological map. The intention should be to assess what the climax vegetation would be if there are no anthropogenic disturbances. Land-use conversion Over time, land use can change from one category to another. The IPCC identifies separate methodologies for land staying within a certain land-use category between reporting periods and land converted to another land-use category during a reporting period. Whenever possible, these conversions should be made explicit e.g. Grassland Converted to Forest Land instead of generic e.g. Land Converted to Forest Land. For REDD+ purposes one land-use category and two land-use conversions are of interest: 8

13 Forest Land Remaining Forest Land This should be the case for most of the managed forest land in Kenya. Most of the REDD+ eligible activities can be considered under this scenario. Land Converted to Forest Land In case of natural regeneration upon abandonment of Cropland or Grassland or planned conversion to forest. Whenever possible, the origin land-use category prior to it becoming Forest Land should be made explicit. In case of natural regeneration the REDD+ eligible activity of enhancement of forest carbon stocks could be applied, while for planned reforestation it could be sustainable management of forest or enhancement of forest carbon stocks. Forest Land Converted to other Land (any of the other land-use categories) When forest is lost due to deforestation or some other disturbance. This is not a REDD+ activity of course, but the country has to report on all the managed Forest Land. If REDD+ activities had been developed on the land being converted, then these estimates may be used to estimate the emission from the conversion. Note that land that remains assigned for forestry purposes after harvesting is still considered forest the temporarily unstocked situation. After a conversion the area is counted under the new land-use category, but in a so-called transition period during which specific estimation methods for the carbon pools are used to properly estimate the carbon dynamics until a new equilibrium is reached. This period lasts for 20 years (IPCC default). Estimation of area for each land-use category The IPCC identifies three Approaches to estimate the area in each land-use category: 1. Total area of the land-use category In this case only national statistics on the total area for each land-use category is available. No details are given on conversions between land-use categories. This might be sub-divided between administrative units, but no specific locations or processes are known. 2. Total area of the land-use category and conversions between land-use categories In addition to the information from the previous category, changes between land-use categories are known, for instance from detailed inventories at sub-national level, but their locations are not. This information can be represented in a so-called conversion matrix, which gives areas for each combination of original land-use category and current land-use category. 3. Spatially-explicit land-use conversion data In this case the precise locations of changes in landuse category are known, for instance from analysis of satellite imagery or detailed field surveys. As in Approach 2, a conversion matrix can be constructed. The IPCC does not differentiate strictly in the reporting between Approaches. Countries are encouraged to use Approach 2 or 3 and a suggestion is made that countries can mix Approaches within a reporting period and over time as the need arises and knowledge and information increase. Approaches for the estimation of areas in each land-use category are not to be confused with Tiers for estimating amounts of carbon pools. It is fair to suggest, however, that higher Approaches and Tiers should be adopted in step. It would be very odd to apply a Tier 3 estimation method for a carbon pool on an Approach 1 area estimate, for instance. Inversely, a Tier 1 estimation method could 9

14 conceivably be applied on an Approach 3 area estimate as in the case of soil organic carbon, which is not likely to be measured with sufficient spatial resolution to provide tailored estimates for all landuse categories and conversions between them. Greenhouses gases In the 1996 Guidelines the IPCC identifies a number of greenhouse gases, from the well-known carbon-dioxide (CO 2 ) to complex molecules such as the family of halocarbons (used in refrigerators, air conditioners and power transformers). All of these gases have a so-called Global Warming Potential (GWP) in the atmosphere, which is expressed in CO 2 -equivalent units (CO 2 e).for the AFOLU sector and thus REDD+ carbon dioxide is the principal greenhouse gas of consideration, with a few notable exceptions, which must be estimated and reported: In mangroves and seasonally flooded forests, a non-negligible amount of methane (CH 4 ) and nitrous oxide (N 2 O) may be formed in and released from the soil 3. If soils in plantations are being fertilized, nitrogen oxides (NO x ) will be released, which are a precursor to the greenhouse gas nitrous oxide. Forest fires release nitrous-oxides and sulfur-dioxide (SO 2 ). Carbon pools The IPCC recognizes five carbon pools for the AFOLU sector; see the table below. Countries do have the option to deviate from these definitions, but this is not recommendable as all methods provided in the Guidelines use these carbon pools consistently. It is possible, however, to adapt the pool details to the current practice of forest survey and inventory, such as in the definition of class sizes, of which some examples are given in the Table below. Table 2-1: Definitions for carbon pools used in AFOLU for each land-use category. Pool Biomass Dead organic matter (DOM) Above-ground biomass (AGB) Below-ground biomass (BGB) Dead wood Description All biomass of living vegetation, both woody and herbaceous, above the soil, including stems, stumps, branches, bark, seeds, and foliage. Note: In cases where forest understory is a relatively small component of the aboveground biomass carbon pool, it is acceptable for the methodologies and associated data used in some tiers to exclude it, provided the tiers are used in a consistent manner throughout the inventory time series. All biomass of live roots. Fine roots of less than (suggested) 2mm diameter are often excluded because these often cannot be distinguished empirically from soil organic matter or litter. Includes all non-living woody biomass not contained in the litter, either standing, lying on the ground, or in the soil. Dead wood includes wood lying on the surface, dead roots, and stumps, larger than or equal to 10 cm in diameter (or the diameter specified by the country). 3 The IPCC Guidelines provide estimation methodologies for these greenhouse gases under the land-use category of Wetlands. In many countries, however, mangroves and particularly seasonally flooded forests are accounted as forests and reported as Forest Land in the National Communications. The Guidelines allows for such deviations from the categories listed, but these need to be clearly documented and applied consistently in subsequent National Communications. In practice this means that the methodology for Wetlands can be applied to estimate emissions for Forest Land. 10

15 Pool Soils Litter Soil organic matter (SOM) Adapted from: IPCC 2006 Guidelines, Volume 4, Chapter 1, Table 1.1. Description Litter Includes all non-living biomass with a size greater than the limit for soil organic matter (suggested 2 mm) and less than the minimum diameter chosen for dead wood (e.g. 10 cm), lying dead, in various states of decomposition above or within the mineral or organic soil. This includes the litter layer as usually defined in soil typologies. Live fine roots above the mineral or organic soil (of less than the minimum diameter limit chosen for below-ground biomass) are included in litter where they cannot be distinguished from it empirically. Includes organic carbon in mineral soils to a specified depth chosen by the country and applied consistently through the time series. Live and dead fine roots and DOM within the soil, which are less than the minimum diameter limit (suggested 2 mm) for roots and DOM, are included with soil organic matter where they cannot be distinguished from it empirically. The default for soil depth is 30 cm. The 2006 IPCC Guidelines saw the addition of a new carbon pool: Harvested Wood Products (HWP). HWPs are those components of trees that are converted into timber, board, paper/pulp, etc. which store carbon after removal from the forest. It is fundamentally different from the other carbon pools in several respects: The carbon is not stored on-site. HWPs have a life-cycle that is different from that of the forest. For every product a halflife period can be defined, which is used to determine the amount of product still remaining after a certain period of time; the rest is assumed to have oxidized and thus emitted into the atmosphere as CO 2. The half-life has a default value of 2 years for paper and 30 years for solidwood products. HWPs are not reported together with the other carbon pools for a certain area. Instead they are reported at the national level, in order to account for them together with imports and exports of wood-based materials. For application in REDD+, where local stakeholders have an interest in a locally consistent accounting structure such that results-based benefits may be determined, HWPs are not relevant, particularly not when those stakeholders accrue the benefits from raw material being converted into HWP e.g. harvesting of trees for pulp or timber by smallholders. If, however, harvesting of trees is managed and controlled by large companies or state-owned organizations, it is imperative that the local stakeholders share in the benefits of the harvest, because the conversion of trees into HWPs will impact the carbon balance of the forest and thus the amount of results-based benefits that can be claimed to support monitoring for REDD+. Key categories In order to reduce the effort in compiling National Communications and focusing resources on those areas where the majority of emissions and removals take place, the IPCC identifies key categories. A key category is one that is prioritised within the national inventory system because its estimate has a significant influence on a country s total inventory of greenhouse gases in terms of the absolute level, the trend, or the uncertainty in emissions and removals. As such, key category analysis is a task for the national authority that compiles the National Communications, to be applied on each sector individually. Key category analysis aims to identify those categories within the sectoral scope such as AFOLU that have a significant influence on the inventory. For Forest Land this implies that the following questions need to be answered: 11

16 Which management activities are significant? Which CO 2 emissions or removals from various carbon pools are significant? Which non-co 2 gases are significant? Typical key categories in forestry are: AGB for all forests that have undergone or are undergoing human impacts. For managed forests that have not suffered such impacts such as National Parks or wildlife reserves AGB does not necessarily have to be estimated because the carbon dynamics from human impacts will be minimal, unless there are special management actions such as fire management. If there is a significant amount of deforestation: SOC and dead wood in Forest Land converted to some other category (BGB is instantly converted to dead wood upon deforestation). If there is a significant amount of forest regeneration: SOC, AGB and BGB on Land Converted to Forest Land. SOC in mangroves and peatlands. Other key categories may be identified based on national circumstances. The key categories should jointly comprise a substantial part of emissions and removals 95% as established by the IPCC. They should be identified using quantitative data if such data are available in the country (e.g. forest inventory data); in the absence of such data a qualitative assessment of key categories may be made. Key categories should be estimated with the highest tier method possible. Other categories may be estimated with a Tier 1 or Tier 2 method. In most developing countries, estimation of at least some carbon pools on Forest Land will be classified as key category, especially when the country participates in the REDD+ mechanism. This will, in all likelihood, include AGB, which can be included in a PCM protocol. Litter and dead wood may also be collected through PCM as the sampling and measurement are straightforward even if they are not classified as a key category. Even if they are classified as a key category, SOC and BGB are not amenable for inclusion in PCM due to the complexities of sampling and measurement these carbon pools are best left to professional forest survey staff or researchers. Two basic approaches to calculation: stock-difference and gain-loss The IPCC Guidelines provides methods and parameters for Tier 1 estimation of emissions and removals. These same methods can be applied for Tier 2 estimation, if country-specific parameters are available. For the estimation of biomass in the land-use category of Forest Land Remaining Forest Land two different methods are provided, the default gain-loss method and the stock-difference method. The gain-loss method is a process-based method in which all changes in the carbon pools in the forest are accounted for: growth of trees, transfers from one pool to another (e.g. leaf shedding, falling branches, decomposition) and removals (harvesting of timber and non-timber products). In its basic form the gain-loss method is quite simple: (1) 12

17 where ΔC is the annual carbon stock change in the forest and the subscripts refer to gains and losses, respectively. This formula is deceptively simple and the devil is in the detail. In order to apply this formula, all the gains and losses have to be individually identified and quantified. It is unlikely that such data is available with sufficient detail in Kenya. For instance, estimating the annual gain in biomass requires knowledge of the average net annual increment for specific vegetation type for every climatological and ecological zone (IPCC, 2006; section ). For losses, separate estimates have to be made for wood removals (timber harvesting), fuel wood collection (as whole trees and as tree parts) and disturbances (due to fire, by fraction of biomass affected). The stock-difference method, on the other hand, appears more complex but is more compatible with the general context of forest inventory: (2) where C * refers to the carbon stock in the forest at time t 2 and t 1, as indicated. This method is quite easily implemented with the available forest inventory data. This does require, though, that the inventory data are of sufficient quality to be acceptable to the international community for the purpose of estimating carbon fluxes. Both of these methods are considered to be Tier 1 methods when applied with standard data: using default values for certain forest properties and assumed to have large uncertainties. If the stockdifference approach is applied with fine-grained, locally measured data it can be considered to be of a higher tier with lower associated uncertainty. IPCC suggests using higher tiers where possible, with higher tiers requiring more locally measured data and more advanced methods, such as speciesspecific allometric equations or dynamic models. Figure 1 provides a decision tree to select the appropriate tier for the assessment of the carbon pools. Assuming that accurate biomass data can be extracted from the forest inventory data but without advanced methods (e.g. locally derived, speciesspecific allometric equations for the various types of forest), Kenya should be able to report biomass at Tier 2 level. IPCC explicitly indicates that the methods are for managed forests, i.e. forests subject to some form of intervention that typically requires or yields information on forest resources (e.g. harvested volume, mean annual increment). However, the methods are Figure 1: Tier selection tree (IPCC, 2006). 13

18 defined so generically 4 that they can also be applied to unmanaged forests, if some estimate of the required parameters can be made. For DOM, dead wood and litter, similar formulations exist. For SOM a separate formulation is developed. For Land Converted to Forest Land and Forest Land converted to some other land-use category similar formulations exist, but with specific attention to transfers between carbon pools such as BGB becoming DOM upon deforestation. Spatially-explicit data is usually only available for the above-ground biomass. Other carbon pools then have to be estimated using a conservative approach, which for reducing deforestation and forest degradation means that the other carbon pools are assumed to have a constant value, a Tier 1 approach. Measurement of especially DOM is advisable because that will increase the confidence in the estimate of total carbon in the forest while they are relatively easily collected in the field. Belowground biomass and SOM may be determined by professional forest inventory specialists on a coarse sampling basis. This is particularly important in those forest ecosystems where there is a large pool of carbon in the soil e.g. mangroves. Both equations are conceptually simple, but they are quite different in implementation: The stock-difference method requires periodic assessments of the standing biomass and DOM. For the biomass equation the independent variable of interest is the merchantable growing stock volume which is converted into biomass using a number of conversion factors that are either supplied by the IPCC Guidelines (Tier 1 estimation) or specific to the local conditions (Tier 2 estimation). The volume itself is typically calculated from measurements of diameterat-breast-height (DBH) and height of the bole (H), in combination with the so-called form factor which can be taken from published values for specific tree species. DOM can be calculated from weights of dead wood and litter. All of these values are easily assessed from simple measurements and samples in the field. The gain-loss method requires continuous assessment of removals and disturbances, with a variety of parameters to be measured or estimated. These parameters include things like mean annual biomass growth for each forest type for the calculation of gain and volume of timber removed through harvesting, volume of fuel-wood removed, area burnt, etc. On a more fundamental level, the gain-loss method is an accounting approach, while the stockdifference method is an inventory approach. In the gain-loss method gains are added and losses are subtracted from the previous balance in order to arrive at the current balance. Errors in the measurements or estimations will persist in the balance and these tend to grow over time as subsequent measurements or estimations of gains and losses add their errors to the balance. There may also be bias in the measurements or estimates, resulting from the nature of the parameters. As an example: How much does a bundle of fuel wood weigh? Where exactly does it come from? How many bundles are extracted in a year? The stock-difference method, on the other hand, makes a fresh, unrelated measurement periodically and any errors of previous measurements are cancelled out over time. From the perspective of PCM the stock-difference method is preferable as it relies on simple measurements which can be collected without using complicated methods or tools and the measurements can be expected to be unbiased between locations and over time. 4 They are defined to be applicable to any of the defined land use categories. 14

19 Countries are actually encouraged to move to a Tier 3 approach, different from the two methods presented above, as more data and knowledge become available. For biomass assessments (AGB and BGB) this typically implies the use of a set of allometric equations that each describe the relationship between biomass and some easily measured tree or forest stand properties for all the dominant (e.g. contained in a key category) forest strata. The use of such allometric equations is more attuned with the stock-difference method, as it requires a few simple measurements that are collected through field surveys, rather than assessment of gains and losses. Allometric equations are available for a large number of commercially interesting tree species, but more parameterizations are necessary for typical use at a national scale in tropical countries. Such additional parameterizations are best developed by the national forest research institutes. Another Tier 3 approach would be the development of a set of models of forest growth, utilization and disturbance, for each forest stratum and management regime with management referring to all human interaction with the forest. These models are typically driven by the kind of data that is collected for the gain-loss method. Some models exist, but they tend to be not very accurate in their estimation, primarily due to over-simplifications in the models and lack of adequate data to run the models. 2.2 REDD+ activities and sampling If a country decides to participate in the REDD+ mechanism of the UNFCCC, all the managed Forest Land in that country needs to be included in REDD+ accounting. For Kenya this may involve measuring or estimating emissions or removals on many thousands of hectares of forest of various types and under various forms of management and REDD+ activities. In order not to over-burden the forestry authorities with this overhead a sampling approach to the estimation of emissions and removals should be employed. Number of samples, accuracy and confidence interval The goal of sampling is to achieve an estimate of emissions and removals with a stated accuracy and confidence interval, using the least number of measurements. To determine the number of samples, n, given a certain confidence interval and maximum error, one can apply the following formula: (3) where z* is the distribution critical value at a certain confidence interval (published in any textbook on statistics), σ is the standard deviation, e is the maximum allowable error, and μ is the average biomass in the forest stratum. For a forest where μ is 400 t/ha with σ 65 t/ha, if you want to have an error of at most 5%, with 90% confidence interval (z* = 1.645): 15

20 For a 95% confidence interval (z* = 1.960; IPCC good practice value): In all cases the average biomass in the forest stratum μ and its standard deviation σ need to be established first. This is best done by professional foresters, using generally accepted techniques for sampling. In practice this implies a minimum of 30 randomly located samples per forest stratum. Protocols regarding confidence interval and maximum error are likely to be established nationally. The number of samples required to reach that confidence interval given a certain maximum error for each forest (type) should be determined by a professional organization, e.g. the Forest Survey Department, using accepted statistical practice. It can be reduced by careful stratification of forest ecosystem / type, because that will reduce the standard deviation of the samples in each stratum, and this is strongly recommended. Sampling design Sampling design is typically undertaken in function of the expected information to be extracted from the measurements. In the case of carbon assessment for REDD+, the number of samples per forest stratum should be based on the guidance given above. Where the assessment is made for and with communities implementing REDD+ activities, sampling should be designed such that all communities have regular measurements in the forest land that they manage for the national REDD+ program. This typically leads to a sampling density that is much higher than would be necessary following the equations above. There are a variety of reasons to support such a higher sampling density: Communities need to be able to self-assess their performance in reducing emissions and enhancing removals. This helps them in negotiating benefits that the community should receive from implementing REDD+ activities and the reduced emissions and enhanced removals achieved. Communities can only assess their performance if measurements are made in the forest that they manage, since it is unlikely that the state agencies will be able to assess the dynamics in the forest for each participating community from a national forest monitoring system using remote sensing and standard forest inventory programs alone. Areas where no REDD+ activities are implemented, or where implementing is lackluster or inadequate will negatively impact the achievements of efficient implementers, thus reducing their interest for lack of benefits. If measurements are made regularly, they will be made with fewer mistakes than if protocols are used only sporadically, both for forestry staff and community members. A higher sampling density will allow for error checking. In practice, a single community may be managing from a few hundred to a few thousand hectares of forest land. This area will likely have several distinct regions in it, based on forest type, condition (e.g. pristine, degraded; on a steep slope, riparian buffer) or management regime (e.g. plantation, regeneration, sustainable harvesting). Each of these regions will require measurement of tree and 16

21 forest properties to estimate the carbon balance. Given that the reporting to the UNFCCC takes place bi-annually, the measurement needs to be repeated with at least a bi-annual frequency. Plot design Plot design is a function of the parameter that needs to be measured. The plot should be large enough to capture enough specimens (trees, litter traps, etc.) to enable the calculation of mean value and variance, but not so large that the measurement becomes burdensome. The table on the next page can be used as a first guidance for the assessment of trees and bamboo. Other carbon pools or vegetation are typically assessed using sub-plots. Litter and herbs are typically collected using 1m 2 sub-plots, of which several may be laid out on a single sample plot. As with most guidance from the IPCC, there is scope for adaptation of sampling methodologies, so long as they are properly documented and consistently applied. Table 2-2: Sample plot sizes for different forest conditions. Type Condition Size (radius) Natural 10 or more trees with DBH >6cm 500m 2 (12.62m) Less than 10 trees with DBH >6cm No trees with DBH >6cm Bamboo 25 or more trees with DBH <6cm Less than 25 trees with DBH <6cm m 2 (12.62m) 1,000m 2 (17.84m) 500m 2 (12.62m) 100m 2 (5.64m) Mixed forest Tree bamboo 500m 2 (12.62m) Bamboo tree 100m 2 (5.64m) Plantation Irregularly spaced As above for trees or bamboo Regularly spaced 5 rows x 5 trees Source: Technical Manual for Participatory Carbon Monitoring, UN-REDD Viet Nam Programme. It is preferable to make Permanent Sampling Plots (PSP) plots that are revisited at regular intervals. Use of PSPs does have some drawbacks people might avoid degrading activities there because they know it will negatively impact the benefits from REDD+, they have to be locatable after a year, etc. but their use does allow for a more robust statistical analysis of the measurements. 2.3 Data quality assessment All the information that is reported to the UNFCCC in the National Communications needs to be assessed for its quality and an estimate of accuracy and associated uncertainty is highly encouraged. Accuracy is defined by the IPCC as agreement between the true value and the average of repeated measured observations or estimates of a variable. An accurate measurement or prediction lacks bias or, equivalently, systematic error and uncertainty as lack of knowledge of the true value of a variable that can be described as a probability density function (PDF) characterising the range and likelihood of possible values. Uncertainty depends on the analyst s state of knowledge, which in turn depends on the quality and quantity of applicable data as well as knowledge of underlying processes and inference methods. In the section on sampling design equations are given to compute the minimum number of samples to obtain an estimate with a known error, given the average and standard deviation of the parameter under consideration (biomass, litter, dead wood, etc.) and an established confidence

22 interval. The equation can also be inverted to compute, given a certain number of samples, average and standard deviation and the confidence interval, the expected error: (4) When applied to the characteristics of the specific area of forest that the measurements refer to average μ and standard deviation σ this equation can be used as the first-order assessment of the quality of the data. Other straightforward assessment methods meaning those can be easily implemented and applied include the identification of outliers (values that are unlikely given other measurements) and comparison to previously measured values. Assessment of data quality can take place in the temporal domain cross-reference repeated measurements over time and in the spatial domain compare estimates with those from other REDD+ areas in the same forest stratum. Observed inaccuracies may have more causes than just measurement errors, such as: Improper stratification of the forest. Differences in efficiency of intervention to reduce emissions and enhance removals. (Use of) Instrument error (e.g. use of a regular tape measure gives circumference of the tree which might be recorded as DBH). Data quality analysis should therefore not be a last step before data is inserted into national accounts of greenhouse gases; instead it should be used to provide feed-back to practitioners of REDD+ on potentials improvements in procedures and implementation. 18

23 3 Roadmap 3.1 Preliminaries There are a number of issues that are not specific to the National REDD+ Programme, the national forest monitoring system or reference Levels, but which nonetheless need to be considered in the development and implementation of this roadmap. These issues are presented here also see the previous chapter. Typically, they require coordination and collaboration with a variety of stakeholders outside of the forestry sector. 0.1 Forest definition Description Kenya has to decide upon a definition of forest for reporting to the UNFCCC. This definition automatically defines how to asses Forest Land one of the six IPCC land use categories. This definition has to be developed in consultation with stakeholders in other sectors who are responsible for the assessment of the other five IPCC land use categories in order to arrive at a consistent land use category classification that encompasses the entire territory of Kenya. Different definitions would be applicable to forests in different ecological regions. Height at maturity, for instance, is very different in forests in the semi-arid zone from that in the southwest. Unfortunately, the UNFCCC does not seem to allow for multiple definitions for a single country. The dry woodlands are preferably included in the definition of forest greatly enhancing the scope for implementing REDD+ activities in Kenya but considering the potential for sequestering carbon and options for managing the land. Mangroves should be classified as forest. It is critical to differentiate between the mandate to set and implement policy for land use and reporting to the UNFCCC. As an example, a coffee estate may qualify as Forest Land and reported as such to the UNFCCC while the governing mandate is with the Ministry of Agriculture. Remarks The UNFCCC has established a generic definition of forest for use under the Kyoto Protocol in Decision 16/CMP.1, Annex, paragraph 1(a): Forest is a minimum area of land of hectare with tree crown cover (or equivalent stocking level) of more than per cent with trees with the potential to reach a minimum height of 2 5 metres at maturity in situ. A forest may consist either of closed forest formations where trees of various storeys and undergrowth cover a high proportion of the ground or open forest. Young natural stands and all plantations which have yet to reach a crown density of per cent or tree height 19

24 0.1 Forest definition of 2 5 metres are included under forest, as are areas normally forming part of the forest area which are temporarily unstocked as a result of human intervention such as harvesting or natural causes but which are expected to revert to forest. Kenya has to establish the exact values of the three criteria minimum area, minimum tree crown cover and minimum tree height at maturity. An additional criterion may be the minimum width of a strip of trees to be considered forest (typically 20 meter). The criteria should be set in such a way that woodlands in the semi-arid areas of Kenya would still qualify as Forest Land. Palms and bamboo are typically included in the definition, but this has to be specifically indicated. Of special consideration are tree-based agricultural production systems, such as coffee, mango, cashew, macadamia, citrus, coconut and mixed home-stead trees or agroforestry systems. If included, these areas should be reported to the UNFCCC as such and they become eligible for REDD+ interventions and benefits. Their inclusion should be explicitly described, especially because it may be expected that these agricultural systems are reported as agricultural land in national accounts and statistics, thus creating a discrepancy between national accounts and international reports. Setting the minimum area and height at maturity at the low threshold such that small-holder tree groves and dry woodlands would classify as Forest Land would make assessment of forest cover with current remote sensing techniques very difficult and error-prone. Alternative assessment methods or extensive field monitoring for use in satellite image classification would become necessary for these environments. Stakeholders Ministry of Environment and Mineral Resources Kenya Forest Service Kenya Wildlife Service Ministry of Agriculture Ministry of Local Authorities National Environment Management Authority Ministry of Planning, KNBS UN Agencies in Nairobi KEFRI DRSRS National Land Commission National Universities WWF IUCN 20

25 0.1 Forest definition Schedule Budget Current status and actions required This definition is crucial to all subsequent activities and should be established as soon as possible. None. Definition of forest used for CDM projects. Consensus agreed on broad parameters to define a forest under REDD+ (minimum 2m height, minimum 15% forest cover, minimum 0.5 ha). Need to hold further consultations to agree on a definition that addresses the additional possible REDD+ areas. Description 0.2 Mapping, classification and data management All mapping needs to be done in conformance with national standards. This includes properties such as scale, resolution, accuracy, coordinate reference system, data quality assessment and control, but also standards related to data collection, verification, analysis and processing. While coordinate system, data quality assessment and control are elements whose decision can be determined easily, issues of scale, resolution, and accuracy need to be decided against the overall objective of the Programme, the current and future and historical availability of data as well as current and future budget allocation. The decision has to be realistic to these facts. All data needs to be stored in a secure environment such that it will remain available for later analysis and re-processing as required. Security needs to be in place to safeguard both the integrity of the data and the National REDD+ Programme built on top of it. The Government Data Center may provide this service to the National REDD+ Programme. Remarks Initiatives like the Kenya Biodiversity Information System (KENBIF) have initiated a process on mobilization of biodiversity information, including data sharing issues, versioning, updating of data, and aggregation of distributed databases, interoperability standards and infrastructure. Similarly, data standards or protocols for secondary data sources such as livelihoods and economic activities should be identified and considered. The UNFCCC has published standards for data management and transmission. Stakeholders Government Data Center Kenya National Bureau of Statistics Department of Resource Survey and Remote Sensing Survey of Kenya Kenya Forest Service 21

26 Schedule 0.2 Mapping, classification and data management Kenya Wildlife Service Ministry of Local Authorities National Museum of Kenya National Universities Inventory has to be made of existing standards and protocols. This should be completed prior to any mapping and data collection. Budget Probably less than $10,000. Current status and actions required Existing standards for mapping exist with the Survey of Kenya. Description 0.3 Land use category mapping The 2006 IPCC Guidelines require that Agriculture, Forestry and Other Land Use (AFOLU) are reported in six land use categories: Forest Land, Cropland, Grassland, Wetlands, Settlements and Other Land. As specified in the Cancun Agreements UNFCCC Decision 1/CP.16, paragraph 60 (b) Kenya has to provide a new National Communication to the UNFCCC every four years and thus the land use category mapping has to be repeated every four years, with a biannual update. This therefore requires integrating this task in the standing mandate of an organization. The areas reported in the land use category mapping needs to be consistent with other national accounts and statistics. This is particularly relevant where land use categories cross over into distinct sectors of government and reporting e.g. classification of tree-based production systems as Forest Land but accounting them in national statistics as agriculture. REDD+ can be applied only in those areas classified as Forest Land. To benefit farmers who have planted trees on farms, these woodlands/trees should be classified as Forest Land (e.g. private forests). Remarks The areas reported in the land use category mapping needs to be consistent with other national accounts and statistics. This requires the collaboration with or preferably the involvement of the Kenya National Bureau of Statistics. Land use category mapping is typically performed using satellite imagery. Secondary data sets should be included to the extent possible. Typical data sets are: areas leased for specific activities e.g. tree-based crops, irrigated land, boundaries of National Parks, gazetted forest areas, boundaries of urban areas, 22

27 0.3 Land use category mapping road network. Stakeholders Schedule Budget Current status and actions required Ministry of Environment and Mineral Resources Kenya National Bureau of Statistics Department of Resource Surveys and Remote Sensing Ministry of Forestry and Wildlife Ministry of Lands Ministry of Agriculture Survey of Kenya National Land Commission Recurrent task. The structures to implement this task need to be established as soon as possible as many other tasks on REDD+ depend on it. Extensive capacity building is required to build sufficient human skills and institutional capacity to successfully implement this task. Initial establishment of the capacity, approximately $200,000 for design and implementation of protocols, procedures and a data management system, and training of staff. Kenya Forest Service, with support from DRSRS, RCMRD and the Survey of Kenya are mapping the country in these six categories for three historic periods on the basis of wall-to-wall satellite imagery. Good documentation of procedures and outputs is needed, as well as consultation with specialists in this area, plus quarterly, mid and final evaluation of quality and accuracy assessment before completing the wall-to-wall mapping. The sensitivity of the exercise and the value of the final output can only be safeguarded if such measures are put in place well before the team submits the final output. The choice of technique is critical to ensure comparability over time, as some techniques require definition of thresholds used to separate cover categories. Since these thresholds are arrived at by the individual analysts using visual interpretation of the satellite images, then to some extent there should be some internal control/peer review among the team members. The KENCAS project, supported by the Clinton Climate Initiative with financial support from the Government of Australia, has as objective to implement this task. Design of this task needs to be completed and consolidated with strategies and procedures in the Ministry of Environment and Mineral Resources. 23

28 Description 0.4 Ecological stratification Land-use categories can be stratified into sub-categories in order to obtain more homogeneous units for which separate estimation methodologies or conversion factors exist. The IPCC suggests stratification along ecological zones and climatic domains (between which one may expect a large correlation). On Forest Land, further stratifications can be made first on ecological zone and then on forest type e.g. savannah woodland, miombo, broadleaf, mangrove, plantation. The stratification should not map current vegetative cover, but rather make a map of potential vegetation through bio-geographical analysis e.g. what vegetative cover would the land support if there was no human interference. This is useful not only for accounting purposes, but also for determining intervention strategies and the gap between current and potential carbon content and thus the potential for enhancing removals of greenhouse gases. For REDD+, stratification of Forest Land based on ecological criteria is most relevant and considered good practice by the IPCC. Ecological stratification is key to determining carbon contents in Kenya. For example, montane forests have very similar characteristics and a sampling done on Mt Kenya would not deviate much from one done in a similar forest in Mau. Similarly the woodlands of Kenya (which will qualify as forests if the definition of forest includes trees of 2 metres) have similar characteristics. Remarks Stratification should strike a careful balance between the overhead costs of having to estimate emissions and removals of greenhouses gases from every stratum, with its concomitant burden of establishing an estimation method and/or set of conversion parameters for each stratum, and the (assumed) higher accuracy of the estimate that is obtained. In general, on Forest Land stratification should at least be applied to the ecological zone based on topography and climate. While every additional stratification parameter has the potential to substantially increase the number of strata (as an example, 3 topo-climatic zones and 5 forest types give 15 potential combinations; an additional 4 forest management types yields a total of 60 potential combinations), in practice the number will be smaller (there is no mangrove in high mountain areas) and there may be common estimation methods or sets of conversion parameters between different strata. Stratification is considered good practice by the IPCC. The stratification is typically done only once, given that the ecological criteria underlying vegetation patterns are quite stable. It would be possible, however, to make similar mappings assuming certain patterns of climate change, in particular temperature rise and changes in precipitation patterns. Such simulations may be very useful in establishing reference levels and developing intervention 24

29 0.4 Ecological stratification (adaptation) strategies. Ecological stratification should be undertaken by research organizations, bringing together expertise from all relevant fields in bio-geography e.g. ecology, soil science, climate science. The stratification should be subjected to rigourous peer review and be endorsed by the relevant authorities e.g. the Ministries dealing with land cover and land use prior to being published and applied. Several approaches are available for stratification and some efforts have been made for Kenya in the past. Ecological mapping by F. White (around ) may be used as a basis for a current mapping. Holdridge life zone mapping and the LCCS system by FAO/Africover, may be options for suitable methodologies. Stakeholders Schedule Kenya Forest Service Kenya Forest Research Institute Kenya Wildlife Service Ecology research organizations NMK Moi and Jomo Kenyatta Universities NEMA Ministry of Agriculture Ecological stratification of the Kenyan territory is a key element of accounting of reduced emissions and enhanced removals on Forest Land and possibly also other land use categories, but also of planning interventions and possibly benefit distribution, and it should thus be undertaken with priority. This is a one-off activity, although updates may be applied on a long-term basis e.g. once every 10 years. Scenarios simulating ecological patterns as a function of climate change may be developed separately. These are not critical to the National REDD+ Programme, but they are useful inputs in a number of other processes. Useful temporal horizons are 2050 and 2100, as these years are typically used in climate change simulations, but other reference years for which climate change scenarios for Kenya are available may be used. The temporal horizons for simulations could specifically be based on the National Carbon Emissions discussions as the factors to consider in such simulations would be cross-sectoral and based on hypothetical development scenarios based on trend lines, which are beyond the REDD+ scope. Budget Approximately $200,000 for an initial ecological stratification, involving experts from a number of fields in natural resources and ecology. Approximately $100,000 for two simulations of scenarios of climate change. 25

30 Current status and actions required 0.4 Ecological stratification Description 0.5 Forest Land sub-category mapping The land use categories should be subdivided into sub-categories, which is considered good practice by the IPCC. For Forest Land the sub-categories should be based in the first instance on the ecological strata (see section 0.4) and then on forest type e.g. evergreen, needle-leaf, mangrove. A third subdivision should be forest management e.g. protection, plantation forestry, agro-forestry. These environmental conditions determine to a large amount how a forest is monitored and how it will respond to interventions and climate change; but also how it is being exploited and which drivers of deforestation apply on it. There can be as many sub-categories as are deemed necessary, but every new stratum does incur a cost in establishing a new monitoring and accounting protocol although they will typically be very similar and in the determination of a new set of allometric equations and/or conversion parameters. On the positive side, more homogenous strata will require less effort in estimating emissions and removals with low uncertainty, offsetting the cost of stratification. Remarks The criteria for land use category mapping (section 0.3) apply to Forest Land subcategory mapping as well. The mapping has to be repeated every four years for every new National Communication to the UNFCCC. Forest Land sub-categories have a direct relationship to the implementation of REDD+ eligible activities. The results of the implementation needs to be reported using the same basis i.e. the land use (sub-)category mapping as the National Communications. The Forest Land sub-categories should thus always have specific strata for specific implementations of REDD+ activities. The implementation of REDD+ activities should therefore lead to an update of the Forest Land subcategories. This can be in the form both of activity data where are the REDD+ activities implemented and in the form of emission factors the measurements of forest properties used to determine the carbon dynamics in the carbon pools in the area. Forest Land sub-category mapping should use data collected in the field by Forest Stations and Zonal Offices for added accuracy over satellite image interpretation. Especially indication of areas with specific forestry management practices that translate into sub-categories should be mapped from field data (for instance using 26

31 0.5 Forest Land sub-category mapping GPS equipment). Such areas include at least plantations, but any forest area for which field data is available should be included in the mapping. Implementation of sub-category mapping on forestry management criteria should use information sourced from community-level forest management players. Clearly, sub-division at this level will be influenced by the implementation of Forest Act Community Forest Associations (CFA) together with forest managers are developing respective management plans, wherein are defined management zones such as protection areas, utilization zones, intervention zones (usually the 5km buffer round the forest). Such information, whenever available, as well as projection in areas where it is not yet available, will be critical. Therefore, the involvement of CFAs or their representative alliance will be important. Other land use sectors should implement a similar mechanism for establishing subcategories of land use. This is important as different land use transfers conversion from one land use sub-category to another might have specific methods to estimate emissions or removals of greenhouse gases. While this is outside of the scope of the National REDD+ Programme, the Coordination Office should initiate a dialog with the Ministry of Environment and Mineral Resources on how other sectors should implement this task. Stakeholders Schedule Kenya Forest Service Kenya Wildlife Service Kenya Forestry Research Institute Survey of Kenya Department of Resource Surveys and Remote Sensing Community Forest Associations NGOs National Universities Initial mapping and linking with secondary datasets should be undertaken as soon as the ecological stratification is completed (section 0.4) or when an alternative subdivision on ecological criteria is available. Recurrent activity. Basic mapping can be integrated with the ongoing mapping of DRSRS. At later stages, when new mapping has to be undertaken, relevant information from REDD+ projects under the National REDD+ Programme must be incorporated. Budget Initial mapping of sub-categories approximately $250,000. Initially the mapping will be primarily based upon interpretation of satellite imagery, with incorporation of existing secondary information that is known to be accurate e.g. plantations, National Parks, other gazetted areas. Use of secondary data sources has to be 27

32 0.5 Forest Land sub-category mapping considered in the design of this task. Current status and actions required KIFCON data may be used as a basis. Description 0.6 Modeling and upscaling of estimates Even with the definition of strata wherein will be the formulation of representative allometric equations, there still will be required the point-based information to large-area information representing the whole strata or defined forest land. Simple algebraic equations may be explored, or where need be, defining stratumspecific relationship between satellite image values and the parameters of the point-based allometric equation. Implementation of such scheme(s) needs to be agreed, tested and validated. Stakeholders Schedule Kenya Forestry Research Institute Kenya Forest Service Kenya Wildlife Service National Universities Not critical initially. Can be designed and implemented at a later stage. Budget Current status and actions required 3.2 National forest monitoring system The national forest monitoring system is the key component in the management of information on the National REDD+ Programme. The UNFCCC has requested all Parties establishing a National REDD+ Programme to establish a national forest monitoring system to support the functions of measurement, reporting and verification of actions and achievements of the implementation of REDD+ activities. A fully functional monitoring system has to go beyond the requirements posted by the UNFCCC, to include issues such as a registry of projects and participants and evaluation of Programme achievements and policy effectiveness. The national forest monitoring system will operate at all levels of forestry management. Of particular interest, in addition to the national-level functions, are functions performed at the level of 28

33 Zonal Offices and how they relate to the functions of other stakeholders at the national (vertical integration) and sub-national (horizontal integration) levels. Following the upcoming devolution of government, the implementation arrangements need to be revisited, specifically looking at the role of the county government. Initial activities should be piloted first before large-scale roll-out across the country. Description 1.1 Information system All the information pertaining to the National REDD+ Programme needs to be securely stored in one or more information systems. For the national forest monitoring system the requirements of the information system are extremely high. Not only will there be data from a variety of sources and on a variety of topics, but the information in the system is critical to the functioning of the National REDD+ Programme. Given that the information is also very likely of financial value benefits from the National REDD+ Programme or commercial interests from information on forest resources security is a principal concern, not only in data storage and access, but also in transmission. A professionally designed integrated system is therefore necessary. The information should be accessible to all relevant stakeholders, as per the UNFCCC Decisions made in 2009 and This accessibility can best be achieved through a web-based interface. This includes the Forest Conservancies, county governments and Zonal Offices. Also, it should be integrated with other information systems in use in the forestry sector. Care must be taken to provide the right kind of access to every type of stakeholder. Participants in a REDD+ project need to have full access to all data on their project area but not to that of other REDD+ projects, for instance, while the general public should have access to generalized information which cannot be attributed to a single REDD+ project or participant. This requires careful design of the information system and security profiles. Remarks The UNFCCC has issued directives on the design of information systems for the management of transactions under the Kyoto Protocol, the so-called International Transaction Log; see php for a more detailed explanation. Full details can be found in the document referenced on that page. While this information applies to the ITL only, it may be expected that future transactions under a REDD+ mechanism will have similar requirements for information management and transaction processing. Given the sensitive nature of the information system, full security must be ensured at all levels, from data ingest to storage and back-up, fail-over redundant hardware, back-up power supply and physical system security. Given these rather strict requirements it may be advisable to physically establish the information system in an existing secure location, such as the Government Data Center, with 29

34 1.1 Information system off-site storage of back-up data. Access to the system will be through applications based on standard interfaces over the internet. Stakeholders Schedule Government Data Center Kenya national Bureau of Statistics Department of Resource Survey and Remote Sensing National REDD+ Coordinating Office Ministry of Environment and Mineral Resources NEMA KFS Establishing the information system will probably take quite some time and may be undertaken in stages. In a first stage the overall design and choices for a hardware and software environment are made. Critical elements of the National REDD+ Programme are implemented such as the connection with KENCAS and support for the Forest Land sub-categories, and the registry of projects, participants and implementing partners. In a second phase support for measurement protocols can be added. The measurements can be made by participants in REDD+ activities or by professional surveyors through the National Forest Inventory. The measurements need to be checked for quality. A feed-back mechanism needs to be included to allow uploaders to assess their data. After initial quality checking and first-level aggregation the data is uploaded to KENCAS for further processing and reporting. In a third phase support for evaluation of the implementation can be added. Which interventions are successful and which are not? This should also involve monitoring of safeguards as established by the UNFCCC and any other environmental issues or priorities such as protection of water towers, registration of damage from wildlife, impact on poverty reduction or resilience of rural livelihoods, etc. The second and third stage may be implemented simultaneously or consecutively. Budget First phase: Approximately $200,000 for design and $500,000 for hardware and communication equipment. Second phase: Approximately $100,000 for design and $200,000 for capacity building. Third phase: Approximately $50,000 for design and $100,000 for capacity building. 30

35 Current status and actions required 1.1 Information system The Finnish supported MMMB project under KFS is currently developing a Forestry Information System which may be considered when developing the information system. Description 1.2 Registry of projects, participants and implementing partners The National REDD+ Programme has to maintain a registry of all activities taking place to implement REDD+. This has to include at least three targets: 1. Registry of all areas covered under some REDD+ project. This is required for accounting GHG emissions (activity data will fall under separate Forest Land sub-categories for each eligible REDD+ activity), determining benefits, establishing land tenure rights, etc. The areas have to be registered including full spatial data such that analysis in a GIS environment can be supported. 2. Registry of all people involved in a REDD+ project. Only people that are registered will be eligible to receive training, support and benefits from the National REDD+ Programme. A REDD+ project may be set up at various spatial scales, but most likely it will be covering larger areas involving, for instance, community groups jointly managing an area of forest. 3. Functions of the National REDD+ Programme will be implemented by third parties such as local government agencies, NGOs or private contractors. All of these implementing partners need to be instructed by the national coordinating office on their tasks and how to implement it. There could be a training and certification program for specific tasks. It is important that such requirements be set and be made public to avoid situation of creating a club of just a few. The registry of third parties should be made accessible to the public. Remarks The registry will be implemented in the information system (1.1). The registry can be used to support the provisioning of information on certain safeguards and other requirements from the UNFCCC. These include issues like participation of local communities and Indigenous Peoples, gender equality, risk of reversals, etc. Stakeholders National REDD+ Coordinating Office Ministry of Forestry and Wildlife Ministry of Environment and Mineral Resources NEMA Private sector, NGOs 31

36 1.2 Registry of projects, participants and implementing partners Schedule Part of the first phase of the information system (1.1). Budget Approximately $20,000 for design of the registry. Implementation as part of the information system (1.1). Approximately $20,000 for training and deployment. Current status and actions required Training, certification and registration of qualifying third parties akin to the NEMA EIA/EA procedures. However the NRCO will need to track and steer the process to ensure quality of service. The National REDD+ Coordination Office plans to undertake initial activities in the development and implementation of the registry. Description 1.3 Participatory Carbon Monitoring Local communities including indigenous peoples have demonstrated in many ways that they are effective custodians and managers of the forest that makes up part of their daily life and livelihoods. They are often knowledgeable about the forest in aspects that are not covered by professional forest inventory and, through their physical and socio-economic condition, well placed to manage the forest on a daily basis. In Participatory Carbon Monitoring (PCM) the National REDD+ Coordinating Office, circle forestry agencies, forest owners and local communities collaborate to collect data on the carbon stored in the forest. This data collection will be relatively straightforward, focusing on easily measured properties of the forest such as species name, DBH and density (number of trees per sampling plot). Local communities need to be trained on measurement protocols, data recording and management and analysis (review) of results. Initial establishment of the sampling plots needs to be supported by professional foresters or a certified implementing partner. Remarks PCM could be presented as part of the broader Participatory Forest Management. KFS has to develop procedures for measurement of forest properties with the local communities or owners taking primary responsibility and KFS Inventory Section providing support to this process and undertaking more complicated tasks. The data that is thus collected should be available to the communities/owners and feedback should be given on quality of the data and amount of carbon stored. There are some activities under REDD+ that do not directly relate to measurement of trees, but which would require the involvement of communities, such as patrolling for encroachment, fires, illegal grazing and illegal logging; assessment of 32

37 1.3 Participatory Carbon Monitoring biodiversity and other ecosystem properties. Such activities should be an integral part of PCM/PFM and likewise be recorded in the information system. PCM has proven to be successful in many countries e.g. Tanzania, Senegal, Nepal, India, Mexico, Brazil and various manuals are available to help set up measurement protocols and mechanisms to collaborate with local communities. Stakeholders Schedule National REDD+ Coordinating Office National NGOs working with communities and/or forestry NMK Universities A mechanism for PCM can be set up relatively quickly with the guidance from successful implementation in other countries. Information management needs to be integrated with the information system (1.1) and the registry (1.2). Budget Limited budget for setting up the mechanism. Training and certification program for implementing partners, approximately $50,000. Measurement equipment and data recording equipment, approximately $200,000. Current status and actions required Description 1.4 Integration with National Forest Inventory PCM can provide large amounts of relatively basic information on forest properties where REDD+ project are being undertaken. This basic data needs to be converted to above-ground biomass contained in Forest Land for reporting in the National GHG Inventory. This conversion requires either conversion factors such as wood density, biomass expansion factor, and root-to-shoot ratio or an allometric equation. These have to be based on professional assessment of forests and the associated analysis and assessment of uncertainty using generally accepted statistical procedures. A National Forest Inventory can provide the formal framework to collect forest data and derive the conversion factors or allometric equations. The results from the National Forest Inventory need to be available in the information system where the PCM data are stored in order to facilitate 33

38 1.4 Integration with National Forest Inventory processing of the PCM data. Remarks The National Forest Inventory of Kenya is currently defunct. Under the support program of the Government of Finland to the Government of Kenya an allocation is made to the Finnish Forest Research Programme (METLA) to re-establish the National Forest Inventory, with specific reference to supporting the National REDD+ Programme. While the National Forest Inventory is being re-established, use can be made of Tier 1 IPCC conversion factors and global allometric equations. Once the practice of PCM and data collection protocol are adequately in-grained in the structures practicing PCM, then the latter can form a solid foundation for longterm and sustainable National Forest Inventory. Because with good grass-root data collection structure in the name of PCM, continuation of the National Forest Inventory can be assured. Stakeholders Schedule Budget Ministry of Forests and Wildlife Kenya Forest Service Kenya Forest Research Institute Moi University, other Universities National REDD+ Coordination Office NMK DRSRS NEMA KWS The METLA support program is scheduled to start in the second half of 2012 with an initial planning phase. Establishing the NFI may take several years to complete and provision of conversion factors and/or allometric equations may therefore take some time to materialize. Covered under the METLA support program. Current status and actions required Description 1.5 Information analysis and internal verification The Government of Kenya has to report regularly on its National GHG Inventory to the UNFCCC. As agreed in December 2011, a new report has to be submitted every four years. Information on Forest Land will in all likelihood come from the National REDD+ Programme. The information reported to the UNFCCC has to be accompanied by an estimate of 34

39 1.5 Information analysis and internal verification the uncertainty where possible. For Forest Land this can be accomplished by analyzing all the data submitted from REDD+ projects throughout Kenya, data from the National Forest Inventory and data from other sources e.g. harvesting from plantations, use of wood in various economic activities. The UNFCCC will establish procedures to verify the information on the National GHG Inventory that all countries submit. Kenya can apply these same procedures prior to submitting any information to the UNFCCC to ensure that the independent verification will validate the information submitted by Kenya. Information analysis and verification of data from REDD+ projects may be conducted in a variety of ways: Statistical analysis of submitted data to filter out extraneous data. The analysis can be in the temporal domain compare successive measurements from individual sampling plots or REDD+ projects and in the spatial domain compare measurements from independent REDD+ projects that have similar characteristics. Comparison of submitted data to an independent data source, such as data from the National Forest Inventory or trade information. High-resolution satellite imagery e.g. from Quickbird or Ikonos sensors or airborne photography may be used to establish the accuracy of activity data and verification of the plausibility of submitted data on biomass. Remarks Information analysis and internal verification should apply the procedures for QA/QC established by the IPCC, as these are most likely the same procedures that will be applied by external validators. Information analysis should be undertaken by a party who has no stake in the result of the analysis, such as a research institute or a university. Similarly, internal verification should be undertaken by a party without any involvement in other aspects of the National REDD+ Programme, in particular planning and benefit distribution. Extraneous data may have several causes, ranging from erroneous stratification of Forest Land sub-categories, inadequate training, measurement error, reporting error, to fraud. Extraneous data can be used to improve the efficiency of the National REDD+ Programme e.g. by identifying where additional data is necessary or where procedures need to be modified and should be used as such. Stakeholders Kenya Forest Research Institute Universities Kenya National Bureau of Statistics 35

40 Schedule 1.5 Information analysis and internal verification KARI NEMA Information analysis and internal verification are continuous activities to be started up as soon as REDD+ projects are starting. Budget Information analysis should be a continuous activity and thus be integrated in the planning and budget of implementing agencies. Internal verification is best performed as an ad-hoc activity such that predictability is reduced but it should have a secured institutional mandate and budget. Current status and actions required Description 1.6 Accounting and reporting The Ministry of Environment and Mineral Resources is the responsible government agency for reporting the National GHG Inventory to the UNFCCC. As per the UNFCCC agreement from December 2011, a new National Communication will have to be submitted every four years, with a biannual update on important developments. The information from the National REDD+ Programme will be included in the reporting on the Agriculture, Forestry and Other Land Use (AFOLU) sector. Remarks The Government of Australia is providing support to the Government of Kenya in accounting and reporting for the AFOLU sector through the KENCAS program. Reporting for the various elements of the National Greenhouse Gas Inventory should be carefully considered. The KENCAS program is aiming to implement reporting for the AFOLU sector; reporting on Forest Land will come from the National REDD+ Programme. Such dependencies need to be carefully considered and accounted for in the design of the accounting and reporting system. 36

41 1.6 Accounting and reporting Stakeholders Schedule Budget Ministry of Environment and Mineral Resources NEMA Kenya National Bureau of Statistics National REDD+ Coordinating Office Ministry of Forestry and Wildlife Dependent on KENCAS planning. The KENCAS program will provide funding for the development of the capacity. The task has to be integrated in an institutional mandate and budget made available, for continuous implementation. This will in all likelihood be arranged through the Ministry of Environment and Mineral Resources. Current status and actions required Description 1.7 Providing information on safeguards The UNFCCC requires all countries implementing the REDD+ mechanism to provide information on how safeguards are being implemented. These safeguards are: (a) That actions complement or are consistent with the objectives of national forest programmes and relevant international conventions and agreements; (b) Transparent and effective national forest governance structures, taking into account national legislation and sovereignty; 37