Our Common Future Under Climate Change

International Scientific Conference 7-10 JULY 2015 Paris, France

Menu
  • Home
  • Zoom Interactive Programme
Cliquer pour fermer

Wednesday 8 July - 16:30-18:00 UNESCO Fontenoy - ROOM II

2219 (b) - Politics and numbers: Political and technical challenges in reducing emissions from forests with REDD+

Parallel Session

Lead Convener(s): C. Martius (Center for International Forestry Research (CIFOR), Bogor, Indonesia)

Convener(s): M. Deheza (CDC Climat Research, Paris, France)

16:30

Addressing uncertainty upstream and downstream of emissions reductions accounting from deforestation and forest degradation: examples from Panama and Zambia

J. Pelletier (Woods Hole Research Center, Falmouth, United States of America), S. Goetz (Woods hole research center, Falmouth, United States of America), N. Laporte, (NASA SERVIR project, Falmouth, Ma, United States of America), C. Potvin, (McGill University, Montréal , Canada), J. Busch (Center for Global Development, Washington, DC, United States of America)

Abstract details
Addressing uncertainty upstream and downstream of emissions reductions accounting from deforestation and forest degradation: examples from Panama and Zambia

J. Pelletier (1) ; S. Goetz (1) ; N. Laporte, (2) ; C. Potvin, (3) ; J. Busch (4)
(1) Woods Hole Research Center, Falmouth, United States of America; (2) NASA SERVIR project, Falmouth, Ma, United States of America; (3) McGill University, Biology, Montréal , Canada; (4) Center for Global Development, Washington, DC, United States of America

Abstract content

Uncertainty in emissions and emission changes estimates constitutes an unresolved issue for a future international climate agreement. Uncertainty can be addressed ‘upstream’ through improvements in the technologies or techniques used to measure, report, and verify (MRV) emission reductions, or ‘downstream’ through the application of discount factors to more uncertain reductions. Uncertainties is an important consideration in the REDD+ context since 1) the land-use/cover change and forestry (LULUCF) sector is identified as the sector where uncertainties are the largest, 2) many developing countries still lack capabilities for estimating stocks and flows from forests, 3) the financial mechanism planned for compensating developing countries in their successful efforts to slow, halt or revert forest cover change could include the use of offsets. The integration of REDD+ in the climate regime is providing a new impetus to deal with uncertainty as the offsetting of more certain emissions from fossil fuel with uncertain ER (Emission Reductions) from REDD+ remains an open question with possibly large consequences.

In the context of Reducing Emissions from Deforestation and forest Degradation (REDD+), we provide a diagnosis of the main sources of error to greenhouse gas estimation from land-cover change and overall error using Monte Carlo analysis, using data from Panama, in Central America, and from Zambia, in Southern Africa. For Zambia, we estimate the overall error in emissions from land-cover change and partition the different sources of error starting from national forest inventories. For Panama, we look at the effects of forest monitoring improvements on reductions in uncertainty. We also test five downstream proposals for discounting uncertainty of the potential credits received for reducing emissions. We compare the potential compensation received for these emission reductions to the cost of alternative upstream investments in forest monitoring capabilities. 

We show that upstream improvements can noticeably reduce the overall uncertainty in emission reductions. Furthermore, the costs of upstream investments in improved forest monitoring are relatively low compared to the potential benefits from carbon payments; they would allow Panama to receive higher financial compensation from more certain emission reductions. When uncertainty is discounted downstream, we find that the degree of conservativeness applied downstream has a major influence on both overall creditable emission reductions and on incentives for upstream forest monitoring improvements. Of the five downstream approaches that we analyze, only the Conservativeness Approach and the Risk Charge Approach provided consistent financial incentives to reduce uncertainty upstream. We provide policy relevant inputs for those countries to reduce uncertainty in estimates, taking into account natural variability and measurement errors. More generally, we provide recommendations on approaches to be use to address uncertainty upstream and downstream of accounting, especially relevant if REDD+ emission reductions are to be traded for emission reductions from other sectors.

16:40

Operational approaches for mapping tropical forest biomass and degradation patterns using optical and radar satellite data

N. Barbier (IRD-UMR AMAP, Montpellier, France), M. Schlund (Airbus Defence and Space, Immenstaad, Germany), P. Couteron (IRD-UMR AMAP, Montpellier, France), C. Mathian (Airbus Defence and Space, Toulouse, France), F. Von Poncet (Airbus Defence and Space, Immenstaad, Germany)

Abstract details
Operational approaches for mapping tropical forest biomass and degradation patterns using optical and radar satellite data

N. Barbier (1) ; M. Schlund (2) ; P. Couteron (1) ; C. Mathian (3) ; F. Von Poncet (2)
(1) IRD-UMR AMAP, Montpellier, France; (2) Airbus Defence and Space, Immenstaad, Germany; (3) Airbus Defence and Space, Toulouse, France

Abstract content

Tropical forest mapping is a basic need for a wide range of users, from fields as diverse as forestry, mining and conservation or for gaining carbon credits under practices related to the Reduced Emissions from Deforestation and forest Degradation (REDD+) framework.  Technical difficulties to validate cost-effective procedures to monitor forest structure from space are a hindrance to the increase in number and impacts of REDD+ projects, especially those that aim tackling forest degradation (i.e. REDD’s second D). Even punctual logging operations (e.g. selective logging) can lead to a persistent decline in forest goods and thus be a precursor to forest degradation. For projects quantifying and mapping dense tropical forest structure and dynamics at region to country level is a challenging, yet pressing need, notably but not exclusively for assessing carbon stocks and fluxes. The ‘Forest’ project, funded by EIT – Climate KIC (UE) aims at providing state-of-the-art mapping products for stakeholders requiring a reliable monitoring of tropical forest structure and derived variables. It is built on Airbus satellite fleet bearing optical and radar sensors. It refers to ground and airborne LiDAR data from Central Africa and South America as to calibrate signal inversion under high standards of quality control. The present contribution gathers the results of sensitivity studies carried out by the project team, using very high resolution optical data, as well as synthetic aperture radar data. We present the potential of these two complementary sources of data for assessing critical forest parameters such as above ground biomass density, canopy height and degradation levels.

Despite their broad availability, medium to high resolution optical imagery (e.g. Modis, Landsat 8 or Spot 5) exhibit important limitations when it comes to quantify forest structure variation in high biomass density environments. In contrast to the deforestation phenomena, signals over tropical forests are known to saturate early after canopy closure, bluring the measurement of degradation or biomass. Moreover, instrumental noise (BRDF, atmospheric pollution) is often higher than signal. It is therefore necessary to turn to other spaceborne biophysical observables to monitor forest parameters. In the optical domain, Fourier texture features obtained from very high spatial resolution (e.g. Pléiades) optical data provide non-saturating proxies for stand parameters, including above-ground biomass, within the highest standards of precision (RMSE < 15%) and accuracy achieved to date from remotely sensed data. The influence of acquisition geometry (sun-view angles), that usually hamper regional or multi-temporal studies combining multiple acquisitions can be handled as to avoid biases. These results, and the increasing availability of large swath VHR sensors (Spot 6-7), open the way to applications requiring operational large scale forest structure and degradation monitoring and mapping, except in very cloudy contexts.

Weather-independent TerraSAR-X / TanDEM-X radar satellite data offer an interesting alternative for degradation monitoring, especially when observation frequency is critical. Amplitude change detection of radar time series can indeed be used to automatically detect small scale selective logging or other disturbances. Interferometric SAR features (coherence and INSAR height) derived from TanDEM-X mission data have proved to be correlated to forest structure parameters, enabling consistent estimates of forest height and biomass for large areas at high resolution. In regions where accurate external terrain models are available, e.g. from airborne lidar LiDAR campaigns, the approach using INSAR heights can achieve high accuracies in tree height (LE90 of 7.5m) and biomass (RMSE below 10%) estimation.

Both SPOT6/7 and TanDEM-X data allow producing accurate LULUCF maps compliant with the six land categories mandatory for GHG Reporting as specified by IPCC guidelines. Thanks to information on forest structure, further stratification of the forest area into a range of forest types, including intact and degraded forests, can be achieved. These classes will represent the state of forests as a result of forest disturbance history and can be used to measure the performance and effectiveness of a wide array of mitigation actions as part of MRV schemes.  

16:55

Calibrating carbon cycle models to determine tropical ecosystem biomass stocks: A Tier 3 certification approach

B. Poulter (Montana State University, Bozeman, United States of America), N. Najdovski (Montana State University, Bozeman, United States of America), F. Maignan (Laboratoire des Sciences du Climat et de l'Environnment, Gif sur Yvette, France), P. Ciais (Laboratoire des Sciences du Climat et de l'Environnment, Gif sur Yvette, France), N. Barbier (IRD-UMR AMAP, Montpellier, France), J. Chave (CNRS, Toulouse, France), S. Luyssaert (Laboratoire des Sciences du Climat et de l'Environnment, Gif sur Yvette, France), F. Von Poncet (Airbus Defence and Space, 88090 Friedrichshafen, Germany)

Abstract details
Calibrating carbon cycle models to determine tropical ecosystem biomass stocks: A Tier 3 certification approach

B. Poulter (1) ; N. Najdovski (1) ; F. Maignan (2) ; P. Ciais (2) ; N. Barbier (3) ; J. Chave (4) ; S. Luyssaert (2) ; F. Von Poncet (5)
(1) Montana State University, Ecosystem dynamics lab, Bozeman, United States of America; (2) Laboratoire des Sciences du Climat et de l'Environnment, Gif sur Yvette, France; (3) IRD-UMR AMAP, Montpellier, France; (4) CNRS, Laboratoire evolution et diversité biologique (edb), umr 5174, Toulouse, France; (5) Airbus Defence and Space, Geo-intelligence, 88090 Friedrichshafen, Germany

Abstract content

According to the latest IPCC report land use and land-use change contributes to 9-11% of total anthropogenic greenhouse gas emissions each year and thus changes in land management could provide short-term solution to mitigating climate change.

Initiatives to protect and enhance carbon stored in forests are meant to be enhanced through the implementation of the Reducing Emissions from Deforestation and Forest Degradation (REDD+) mechanism and countries are supposed to take an active role in its implementation. REDD+ requires biomass to be accurately measured, reported and monitored so that carbon can be protected and efforts are remunerated through a result-based principle.

One key challenge is related to how to measure biomass with sufficient accuracy for reporting and in a cost-effective manner related to the pricing of carbon. Taking these constraints into consideration, REDD+ reporting of biomass has been categorized into:

  • high uncertainty estimates, i.e. Tier 1 using global biomass density estimates,
  • to moderate uncertainty, i.e. Tier 2 using country-level biomass densities,
  • and to low uncertainty, i.e. Tier 3, using carbon cycle models calibrated with forest inventory data.

Here we discuss a pathway for estimating Tier 3 carbon stocks in tropical regions using a carbon cycle model, ORCHIDEE, that is calibrated with forest inventory data and constrained by remote sensing data, to estimate project-level carbon stocks and fluxes in above- and belowground biomass pools. ORCHIDEE is in a class of ecosystem models known as Dynamic Global Vegetation Models, and simulates the establishment, growth, and mortality of trees using principles from eco-physiology.

Using tropical forest inventory data, the ORCHIDEE model is calibrated to reproduce growth rates of over and understory trees, and mortality, resulting in estimates of carbon and their change over time in biomass and in soil carbon pools. For project-level implementation, forest canopy height measured from either airborne LIDAR or space-based RADAR is assimilated to ORCHIDEE to extract corresponding carbon values.

Consequently, gridded maps of aboveground biomass, belowground biomass, soil carbon, leaf area index, net primary production, and net ecosystem exchange, at the corresponding spatial resolution, or mapping unit, to that of the canopy information, can be provided for REDD+ programs fulfilling the Tier 3 criteria. We demonstrate this processing chain for sites in French Guiana, Gabon and Cameroon.

17:05

Addressing emissions from agriculture and agriculture-driven deforestation: opportunities for land-sparing and climate-smart agriculture

S. Carter (Wageningen University, Wageningen, Netherlands), M. Herold (Wageningen University, Wageningen, Netherlands), M. Rufino (CIFOR / CCAFS, Nairobi, Kenya), K. Neumann, (Wageningen University, Wageningen, Netherlands), L. Kooistra (Wageningen University, Wageningen, Netherlands), L. Verchot (CIFOR, Bogor, Indonesia)

Abstract details
Addressing emissions from agriculture and agriculture-driven deforestation: opportunities for land-sparing and climate-smart agriculture

S. Carter (1) ; M. Herold (1) ; M. Rufino (2) ; K. Neumann, (1) ; L. Kooistra (1) ; L. Verchot (3)
(1) Wageningen University, Wageningen, Netherlands; (2) CIFOR / CCAFS, Nairobi, Kenya; (3) CIFOR, Bogor, Indonesia

Abstract content

Deforestation and forest degradation are major contributors of global greenhouse gas (GHG) emissions, accounting for a large proportion of many developing countries’ GHG emission budgets (Baumert et al., 2005). According to Hosonuma et al (2012), in 13 countries agricultural expansion is responsible for 100% of deforestation. In this study we consider both emissions from agriculture-driven deforestation, and emissions from existing agricultural land to assess the mitigation potential of the agriculture sector. The inclusion of emissions from agriculture-driven deforestation aligns with recent interest in including agriculture in REDD+ strategies, which aim to directly address the driver of deforestation.

Emissions from agriculture are available at the national level however despite recent efforts to map and quantify land use and land cover change (ESA, 2013; FAO & JRC, 2012; FAO, 2014; Hansen et al., 2013), we consider that these are not suitable to determine agriculture-driven deforestation. This is because deforestation in this case has a focus on land-use changes (from forest to agriculture), so deforestation data based on a forest land-use definition is required. Gross change data are required since, in the cases of China, India and Vietnam for example, large-scale afforestation projects will mean that gains to forest area will underestimate deforestation (FAO 2010). Differences in data coverage, spatial resolution of datasets and in particular definitions of agriculture and forests make comparisons and integration difficult. A new approach combining available datasets of deforestation will be presented which estimates that agriculture-driven deforestation results in 4.3 GtCO2e y-1 from loss of above and below-ground biomass. We use empirical data on forest loss and use data on deforestation drivers to convert this to area of agriculture-driven deforestation.

Using our data on emissions from both agriculture and agriculture-driven deforestation, we consider at the national level the mitigation potential. We demonstrate the use of a systematic framework, which considers sequentially (1) the level and main source of emissions, (2) mitigation potential, (3) enabling environment and (4) risks. The level and source of emissions (1) considers the main source of emissions from either agriculture or agriculture-driven deforestation. Mitigation potential (2) assesses the potential for using climate-smart interventions in the agriculture sector or land-sparing options to address agriculture-driven deforestation. Land-sparing options are supply side interventions which reduce the need for the expansion of agriculture land. The enabling environment (3) is assessed by considering the governance, or engagement in REDD+ which can support the implementation of interventions. Risk (4) assessments identify the potential vulnerability of communities to changes to the agricultural system, and we use food insecurity as an indicator. Findings estimate the mitigation potential of existing agricultural land can be up to 1 GtCO2e y-1. Land-sparing interventions which close the yield-gap, or rehabilitate degraded land offer opportunities to mitigate up to 1.3 GtCO2e y-1 in the tropics where there is potential to close the yield gap or utilize available land and there is a good enabling environment. We highlight countries which are likely to require increased support to implement mitigation initiatives and where safeguarding is required to avoid risks to livelihoods.

This research supports discussions on REDD+ which acknowledge the inherent link between REDD+ and agriculture due to competition for land. We draw on recent debates on land-sparing, and discuss supporting policies which will be required to ensure that negative feedbacks are not realized. Although there is a need to look beyond the interventions covered in this study, we find a large potential for mitigation from these options. This study gives a comprehensive overview of national emissions and mitigation priorities within the forest and agriculture sectors, which can guide decision making and investments at the international level.

17:15

Pathways for sustainable REDD+ policies in Brazil

A. Soterroni (National Institute for Space Research (INPE), Sao Jose dos Campos, Brazil), F. M. Ramos (National Institute for Space Research (INPE), Sao Jose dos Campos, Brazil), A. Mosnier, (International Institute of Applied Systems Analysis (IIASA), Laxenburg, Austria), A. X. Y. Carvalho (Institute for Applied Economic Research (IPEA), Brasilia, Brazil), R. C. M. Souza (National Institute for Space Research (INPE), Sao Jose dos Campos, Brazil), M. Buurman (Institute for Geoinformatics at the University of Münster (IFGI), Münster, Germany), G. Câmara, (National Institute for Space Research (INPE), Sao Jose dos Campos, Brazil), P. R. Andrade (National Institute for Space Research (INPE), Sao Jose dos Campos, Brazil), J. Pirker (International Institute of Applied Systems Analysis (IIASA), Laxenbourg, Austria), R. Mant (United Nations Environment Programme (UNEP), World Conservation Monitoring Centre (WCMC), Cambridge, United Kingdom), V. Kapos (United Nations Environment Programme (UNEP), World Conservation Monitoring Centre (WCMC), Cambridge, United Kingdom), M. Obersteiner (International Institute of Applied Systems Analysis (IIASA), Laxenbourg, Austria)

Abstract details
Pathways for sustainable REDD+ policies in Brazil

A. Soterroni (1) ; FM. Ramos (1) ; A. Mosnier, (2) ; AXY. Carvalho (3) ; RCM. Souza (1) ; M. Buurman (4) ; G. Câmara, (1) ; PR. Andrade (1) ; J. Pirker (5) ; R. Mant (6) ; V. Kapos (6) ; M. Obersteiner (5)
(1) National Institute for Space Research (INPE), Sao Jose dos Campos, Brazil; (2) International Institute of Applied Systems Analysis (IIASA), Ecosystems Services and Management, Laxenburg, Austria; (3) Institute for Applied Economic Research (IPEA), Brasilia, Brazil; (4) Institute for Geoinformatics at the University of Münster (IFGI), Münster, Germany; (5) International Institute of Applied Systems Analysis (IIASA), Laxenbourg, Austria; (6) United Nations Environment Programme (UNEP), World Conservation Monitoring Centre (WCMC), Cambridge, United Kingdom

Abstract content

Brazil’s forests constitute 13% of the global forest area and almost 30% of the tropical forest area. They account for a significant proportion of global terrestrial biodiversity and store about 20% of global above ground forest carbon. At the same time, Brazil’s has a vigorous and dynamic agribusiness, being the world’s largest producer and exporter of coffee, sugar, and orange juice, and is highly ranked in the production and export of soybean, corn, ethanol, pork, beef, and poultry chicken. Reconciling these two realities presents a major challenge to scientists and policy makers that will require up-to-date and accurate land cover/land use data and modeling tools. Within this context, the REDD-PAC project aims at providing a global forum for sharing and improving global data on forests and deforestation drivers and developing best practices for national REDD+ and land‐use planning (see http://www.redd-pac.org). One of the main outcomes of the REDD-PAC project is the development of the GLOBIOM-Brazil model, a partial equilibrium, spatially explicit global land-use model for Brazil’s agricultural, forestry and bioenergy sectors, over a 0.5 by 0.5 degrees spatial grid, adapted from IIASA’s GLOBIOM. Among other refinements, validated with 2010 IBGE agriculture census data and PRODES/INPE deforestation data, the model includes a detailed and up-to-date representation of Brazil’s land cover/land use data together with the main regulations of Brazil’s new Forest Code (FC).

Here we present GLOBIOM-Brazil simulation results in the assessment of the economic and environmental impacts, over the period of 2020-2050, of FC’s key new policies, namely, the restoration obligation (RO) of illegally deforested areas, the small farms amnesty (SFA) and the environment debt offset mechanism based on the CRA, a tradable legal title of forest surpluses. When compared to the non-additional policies scenario (NAPS), the FC has an overall positive impact on land cover changes and corresponding carbon emissions, with an increase of Brazil’s forested area. This is accomplished without much impact in food production, which continues to grow during the coming decades, but, mostly, by conversion of abandoned land and low productivity pastures, mainly in the Amazon.  We also observe that the RO policy is effective only to the extent that the ban on illegal deforestation is fully enforced, in order to avoid the trading between pristine and restored forests. As expected, both the SFA and the CRA mechanism have deleterious environmental impact, since they sensibly reduce the total forest deficit area to be restored by landowners. On the other hand, our simulations indicate that the CRA mechanism is effective in protecting areas of pristine forests, still rich in biodiversity, in regions subject to high economic and demographic pressure, such as the Atlantic Forest and the deforestation arch in the Legal Amazon.