5th iLEAPS Science Conference Abstracts - D4

Abstracts – Session D4

Where are the greatest uncertainties in the Global terrestrial Carbon Budgets?


Philippe Ciais1

1 Laboratoire des Sciences du Climat et de l'Environnement (LSCE), Gif-sur-Yvette, France

D401 - ORAL-0112: An empirical spatiotemporal description of the global surface-atmosphere carbon fluxes: opportunities and data limitations

Jakob Zscheischler, Miguel Mahecha, Valerio Avitabile, Leonardo Calle, Nuno Carvalhais, Philippe Ciais, Fabian Gans, Nicolas Gruber, Jens Hartmann, Martin Herold, Kazuhito Ichii, Martin Jung, Peter Landschützer, Goulven Laruelle, Ronny Lauerwald, Dario Papale, Philippe Peylin, Benjamin Poulter, Deepak Ray, Pierre Regnier, Christian Rödenbeck, Rosa Roman-Cuesta, Christopher Schwalm, Gianluca Tramontana, Alexandra Tyukavina, Riccardo Valentini, Guido van der Werf, Tristram West, Julie Wolf, Markus Reichstein
Understanding the global carbon (C) cycle requires detailed information on spatiotemporal patterns of surface-atmosphere fluxes. Here we adopt a data-driven approach to synthesize a wide range of observation-based spatially explicit surface-atmosphere CO2 fluxes from 2001 to 2010, to identify the state of today’s observational opportunities and data limitations. The considered fluxes include net C exchange of open oceans, continental shelves, estuaries, rivers, and lakes, as well as CO2 fluxes related to net ecosystem productivity, fire emissions, loss of tropical aboveground C, harvested wood and crops, as well as fossil fuel and cement emissions. Spatially explicit CO2 fluxes are obtained through geostatistical and/or remote sensing-based upscaling, minimizing assumptions encoded in process-based models. In many regions, our NCE estimates agree well with independent estimates from other sources. This holds for Europe (mean±1 SD: 0.8±0.1 PgC/yr, positive numbers denote land-atmosphere fluxes), Russia (0.1±0.4 PgC/yr), East Asia (1.6±0.3 PgC/yr), South Asia (0.3±0.1 PgC/yr), Australia (0.2±0.3 PgC/yr) and most of the Ocean regions. Our NCE estimates give a likely too large CO2 sink in tropical areas such as the Amazon, Congo and Indonesia. Overall, our global bottom-up NCE amounts to a net sink of -5.4±2.0 PgC/yr. The comparison with the atmospheric growth rate of CO2 over 2001-2010 leads to a mismatch of nearly 10 PgC/yr and highlights observational gaps and limitations of data-driven models in tropical lands, but also in North America and the Southern Ocean. Our uncertainty assessment provides guidance where to increase carbon observations in the future, and could serve as prior knowledge in multi-criteria optimization schemes and atmospheric inversions. In the future, data-driven estimates of vertical CO2 exchange could be aggregated at national scale to compare with the official national inventories of CO2 fluxes in the land use, land use change and forestry sector, upon which future emission reductions are proposed.

D402 ORAL-0283: Inter-annual variation of Amazon greenhouse balances 2010-2014: nature and causes

Emanuel Gloor1, Luciana Vanni Gatti2, John Miller3, Caroline Alden, David Galbraith1, Lucas Domingues4, Michelle Kalamandeen1, Caio Correia2, Viviane Borges2

1School of Geography, University of Leeds, Leeds, United Kingdom 2INPE, CCST, Sao Jose dos Campos, Brazil 3NOAA ERSL, Boulder, Colorado, The United States of America 4INPE CCST, Sao Jose dos Campos, Brazil

Net carbon exchange between tropical land and the atmosphere is potentially important because tropical land holds large amounts of carbon in forests and soils which can be released on short time-scales. One tool to better understand land atmosphere carbon fluxes over large scales is via the atmospheric greenhouse gas concentration patterns they cause. However the tropics, and particularly tropical land regions, have traditionally been poorly observed. Amongst the land regions in the tropics of particular importance for the global carbon cycle is the Amazon, by far the largest forested region, hosting ~200 PgC carbon in vegetation and soils. Human pressure on the forests is strong and climate conditions are shifting quite rapidly. From 2010 onwards we have extended an earlier greenhouse gas measurement program to include regular vertical profiles of CO2, CH4, CO, SF6, from the ground up to 4.5 km height at four sites along the main air-stream over the basin. We will report here what these data tell us about greenhouse gas balances and their controls over the period from 2010-2014 (possibly including 2015/16). For the specific period we will discuss the year 2010 was anomalously dry, followed by 4 wet years wet (2011, 2012, 2013 and 2014) and another dry year (2015/16), typical for an El Nino year, with drought centred on the Eastern Amazon. These years thus provide an interesting contrast of climatic conditions in a warming world with increasing human pressures. We will analyse the effect of this variability in climate conditions and also discuss what our results suggest for the role of the tropics for the global carbon balance.

D403 - ORAL-0372: High Resolution Net Carbon Balance of the Brazilian Amazon

David Galbraith1, Marcos Adami2, Alessandra Gomes2, Yunxia Wang1, Oliver Phillips1, Roel Brienen1, Guy Ziv1

1School of Geography, University of Leeds, Leeds, United Kingdom 2Instituto Nacional de Pesquisas Espaciais - Centro Regional da Amazonia, Belem, Brazil

Despite its importance globally, the carbon cycle of the Brazilian Amazon is still poorly understood, with different studies yielding fundamentally different conclusions about the overall sign of the net carbon balance.  In this study, we use a new high resolution land cover dataset (TERRACLASS) to evaluate the dynamics of vegetation carbon stocks over a 10-year period (2004-2014).  Our approach explicitly considered spatial variation in aboveground biomass stocks, old growth forest sink strength and secondary forest age.  We find that the net carbon balance of the Brazilian Amazon was an insignificant source over this period, without considering the effects of forest degradation.  Although carbon losses from deforestation declined over our study period, this was somewhat offset by a declining old growth forest sink.   Regrowth of secondary forests made only a modest contribution to the overall carbon balance, being partially negated by re-clearing activities.  We also find that the carbon sequestration potential of secondary forests, despite still being substantial, is considerably lower than other recently published estimates. The continued development of the TERRACLASS product represents an important resource for fine-scale evaluation of the Amazonian carbon balance, allowing for state-level and municipal-level assessments.  The assumed slope of the carbon sink decline in old growth forest was found to be a major source of uncertainty in our study.

D404 ORAL-0038: Constraining the terrestrial carbon sink in the high latitudes of the Northern Hemisphere with large-scale observations

Guy Schurgers1, Anders Ahlström3, 2, Rasmus Fensholt1, Thomas Friborg1, Torbern Tagesson1, 3, Jing Tang1

1University of Copenhagen, Copenhagen, Denmark 2Stanford University, Stanford, The United States of America 3Lund University, Lund, Sweden

Over the last five decades, the terrestrial biosphere has taken up approximately 100 Pg C, equalling nearly 30% of the total anthropogenic emissions for the same period. Model simulations reveal that more than half of this uptake has taken place in latitudes north of 40°N.
Whereas the global terrestrial carbon sink is fairly well-constrained with global numbers for other sources and sinks of CO2, it is challenging to validate the spatial distribution of this sink with large-scale observations. Here, we applied observed CO2 concentrations and earth observation-based estimates of leaf area index (LAI) for the period 1982-2015 to do so.
In contrast to model-data comparison studies that apply variations in the global annual CO2 growth rate only, we use the variations in the seasonal dynamics of observed atmospheric CO2 concentrations from multiple stations, which enables the investigation of regional impacts. Together with the earth observation-based estimates of phenology (GIMMS LAI3g), this provides opportunities to constrain the spatial distribution of the carbon sink in the high latitudes of the Northern Hemisphere.
A series of factorial simulations with the dynamic vegetation model LPJ-GUESS was performed in which the impacts of different drivers of carbon cycle changes (temperature, precipitation, radiation, atmospheric CO2, nitrogen deposition and land use) were applied. These were compared with the observations, and the model results were used to determine typical seasonal patterns of the changes in net ecosystem exchange and LAI induced by these drivers. Because of the difference between these patterns (e.g., temperature-induced lengthening of the growing season results in the strongest impact in spring and autumn, whereas CO2 fertilization results primarily in mid-season changes), observed changes could be attributed to different drivers.

D405 ORAL-0108: Crop yield data can constrain the European carbon budget

Marie Combe1, Allard de Wit2, Jordi Vilà-Guerau de Arellano3, Michiel1 van der Molen3, Vincenzo Magliulo4, Wouter Peters3, 5

1Ghent University, Ghent, Belgium 2Alterra Wageningen UR, Wageningen, The Netherlands 3Wageningen University, Wageningen, The Netherlands 4CNR ISAFoM, Ercolano, Italy 5University of Groningen, Groningen, The Netherlands

Carbon exchange over croplands plays a major role in the European carbon cycle over daily to seasonal time scales. A better description of this exchange in terrestrial biosphere models - most of which currently treat crops as unmanaged grasslands - is needed to improve model estimates of the land carbon sink. In the framework we present here, we model gross European cropland CO2 fluxes with a crop growth model constrained by grain yield observations. Our approach follows a two-step procedure. In the first step, we calculate day-to-day crop carbon fluxes and pools with the WOrld FOod STudies (WOFOST) model. A scaling factor of crop growth is optimised regionally by minimizing the final grain carbon pool difference to crop yield observations from the Statistical Office of the European Union (EUROSTAT). In a second step, we re-run our WOFOST model for the full European 25 x 25 km gridded domain using the optimized scaling factors. We combine our optimized crop CO2 fluxes with a simple soil respiration model to obtain the net cropland CO2 exchange. In this presentation, we assess our model's ability to represent cropland CO2 exchange using 40 years of observations at 7 European FluxNet sites and compare it with carbon fluxes produced by a typical terrestrial biosphere model. We conclude that our new model framework provides a more realistic and strongly observation-driven estimate of carbon exchange over European croplands. Our framework will serve as a new cropland component in the CarbonTracker Europe inverse model to reduce the uncertainty of its terrestrial carbon sink. Its product will be made available to the scientific community through the ICOS Carbon Portal.

D406 ORAL-0300: Is the apportioning methodology used for carbon stock accounting in agricultural soils correct?

Mohammad Ibrahim Khalil1, Bruce A. Osborne1

1UCD School of Biology & Environmental Science and Earth Institute, University College Dublin, Dublin 4, Ireland

The Paris Agreement emphasises the need for enhanced greenhouse gas (GHG) mitigation measures, a reduction in assessment uncertainties, better quantified sinks, and the tailored use of different offsetting mechanisms to keep global temperature <2°C. A precise, verifiable estimation of soil organic carbon (SOC), and its variation at field scales has become pivotal for achieving these goals, and for facilitating carbon offset/credit benefits. For SOC measurements, often reported as a percentage, the absence of consistent sampling protocols and other factors particularly soil moisture and bulk density makes it difficult to assess SOC density/stock and its changes. For accurate estimations, that require no bulk density measurements, the determination of SOC ‘mass by volume’ on an equal soil mass basis, in a defined but adjustable soil layer could be an appropriate approach. The Intergovernmental Panel on Climate Change (IPCC) also proposes using proportional (%) values for a stock change factor (SCF) for application across key agricultural land uses, management practices with variable inputs. Based on SCF factors, methodologies  developed to improve their estimations for national reporting indicate that higher spatial resolution databases, coupled with empirical modelling (bi-exponential) and GIS approaches, have the potential to provide robust estimates (Tier-2 development). However, the SCF factors (as a proportional  gain or loss) used to estimate any change in SOC density/stock resulted in highly variable values, depending on the amount of carbon present in a particular soil varying, for instance, from <10% in mineral soil to >10-<20% in organo-mineral soil. This highlights the importance of replacing the apportioning approach, even for the 4‰ concept, by a ‘mass by area (depth-specific)’ one for more precise estimations. This includes the sub-categorization of mineral and organic soils, calculation of country-specific SCFs for individual land uses/management practices, and the estimation of weighting factors for backward and forward projections where required.

D407 - ORAL-0266: Important omissions in the quantification of the global carbon cycle

Miko Kirschbaum

The main components of the global carbon cycle are the emissions from burning fossil fuels and from land-use change, and the CO2 uptake by the oceans. In addition, there is a “residual uptake” by the terrestrial biosphere that is largely inferred from the differences between the other, more readily quantifiable terms. The size of that residual uptake is often used to constrain global earth-system models, and, more broadly, it guides us about expectations of natural changes in global carbon stocks that may currently and in future add to, or subtract from, anthropogenic CO2 emissions.
However, a global carbon budget with only these four recognised terms omits a number of important additional terms. They are the fluxes into pools of (1) wood products in daily use and (2) wood products in land-fills, and fluxes to the oceans via (3) charcoal transport, (4) wind erosion and (5) non-CO2 break-down products of methane degradation. In addition, there is (6) uncertainty about the way that carbon flux to the oceans by river transport has been incorporated in the overall budget. Each one of these terms is relatively small compared to the dominant fluxes, but together, they can constitute an important additional component that would significantly reduce the size of the inferred “residual uptake”. Incorrect quantification of this flux has potentially serious consequences for our overall understanding of the global carbon cycle, and it can lead to wrong inferences for the future management of the global carbon budget.

D408 ORAL-0169: Quantifying the global carbon sink in secondary forest

Tom Pugh1, 2, Mats Lindeskog3, Benjamin Smith3, Ben Poulter4, Almut Arneth2, Vanessa Haverd5

1University of Birmingham, Birmingham, United Kingdom 2Karlsruhe Institute of Technology, IMK-IFU, Garmisch-Partenkirchen, Germany 3Lund University, Lund, Sweden 4Montana State University, Bozeman, The United States of America 5CSIRO Marine and Atmospheric Research, Canberra, Australia

Although the magnitude of the global terrestrial carbon sink is well constrained, attribution of this sink to its causes remains highly uncertain. A major role has been hypothesised to be played by regrowing forest on secondary land, especially in the northern hemisphere, however this has not yet been consistently quantified at the global scale. We use a global compilation of forest age observations, combined with a global terrestrial biosphere model with explicit modelling of forest regrowth, to calculate the contribution of secondary forest to global terrestrial carbon uptake over the last three decades. We find that the total carbon sink in secondary forests is approximately equivalent to that in largely-undisturbed primary forests, with more than half of this secondary sink originating directly from regrowth. The regrowth sink is concentrated in northern temperate and boreal forests, and some regions show sign of the sink having already peaked. Extrapolating our simulations into the future, we calculate the size of the unrealised secondary forest carbon sink based on current forest structure. Contrasting our data-constrained results with those from the widely-used LUH2 dataset highlights substantial discrepancies in the size and location of secondary forest carbon sinks. Our results suggest that it is not possible to understand the current global terrestrial carbon sink without accounting for the major sink due to regrowth in northern secondary forests. They also imply that a large portion of the current terrestrial carbon sink is strictly transient in nature.

D409 - ORAL-0092: Potential strong contribution of future anthropogenic land-use and land-cover change to the terrestrial carbon cycle

Benjamin Quesada1, Almut Arneth1, Nathalie De Noblet-Ducoudré2, Eddy Robertson3

1Karlsruhe Institute of Technology, Institute of Meteorology and Climate research, Atmospheric Environmental Research, Garmisch-Partenkirchen, Germany 2Laboratoire des Sciences du Climat et de l'Environnement (LSCE), Gif-sur-Yvette, France 3Met Office, Exeter, United Kingdom

Anthropogenic land-use and land cover changes (LULCC) can affect the global terrestrial carbon (C) cycle through changes in climate, fluxes of greenhouse gases and carbon sequestration. However, few studies have quantified the sensitivity of the future terrestrial carbon cycle to LULCC. Here, using Earth System Models (ESM) simulations performed with and without future LULCC, under the RCP8.5 scenario, we find that in response to future LULCC, the carbon cycle is substantially weakened: browning, decreased productivity, higher C loss by disturbances, lower turnover rates and lower ecosystem carbon stocks are simulated. On average, projected global greening and land carbon storage are dampened by ~30% and projected C loss by disturbances enhanced by ~40% when LULCC are taken into account.
LULCC is found to be a predominant driver of future C changes in regions like South America and Southern part of Africa. At the edges of the Amazon and African rainforest, reversals of projected increased carbon sinks and greening are simulated when simulations with are compared to simulations without LULCC. Finally, in most carbon cycle responses, direct removal of C dominates above the indirect CO2 fertilization due to LULCC. Offline simulations with dynamic vegetation model LPJ-GUESS confirm the substantial impact of future LULCC on terrestrial carbon cycle.

D410 ORAL-0374: Who claims the forest sink: The urgent need for IPCC and UNFCCC to reconcile different approaches to assess the

anthropogenic forest greenhouse gas flux

Jo House, Giacomo Grassi, Richard Houghton, Sandro Federici

A recent study (Grassi, House et al. 2017) found that forests are expected to contribute about a quarter of counties’ pledged future mitigation efforts under the Paris Agreement. The study also highlighted a large discrepancy in greenhouse gas flux estimates from the land sector (Land Use, Land Use Change and Forestry) between country reports to UNFCCC and from the IPCC/Global Carbon Project, such that courtiers report a small global net anthropogenic sink while IPCC/GCP find a net source. A different perception of “what is anthropogenic” was speculated as a likely key reason for such difference. This difference may seriously undermine any future assessment of carbon budgets and the gap between current action (by Parties) and the efforts needed to reach the targets of the Paris Agreement. It also undermines transparency and confidence in estimates ahead of the Global Stocktake mandated under Paris.
Here we:
  1. Present the conceptual differences between country reports and global modeling approaches summarized in IPCC AR5 with regard to what is “anthropogenic removal”; specifically, we discuss the differences in (a) the treatment of effects of environmental change and (b) the area of forest considered, with particular reference to what is considered “managed” or anthropogenic forest.
  2. Provide a preliminary quantification of these differences, by comparing the latest forest greenhouse gas flux estimates from individual developed countries UNFCCC submissions, from Houghton (2017) and from TRENDY model results; 
  3. Discuss the implications for the upcoming work by the IPCC and by UNFCCC negotiators, and explore possible solutions.

D411 ORAL-0260: The ecosystem carbon balance under elevated CO2

Christian Körner

Three issues need to be resolved empirically: (1) What controls the carbon residence time in terrestrial ecosystems? Acceleration of carbon fixation for what ever reason (climatic warming, nitrogen deposition, elevated CO2) should accelerate carbon returns, that is, the carbon cycle, not what is needed to sequester C away from the atmosphere. A rise in net primary production most commonly reduces residence time and thus, C-stocks.  This issue is related to tree longevity controls and soil organic matter (SOM). (2) The incorporation of C into biomass or SOM requires free soil nutrients.  How can the nutrient provision keep pace with the rising availability of CO2 for photosynthesis? The four existing long term CO2 enrichment experiments with natural, mature forests show no growth stimulation by elevated CO2, presumably due to stoichiometric constraints. Because photosynthesis rises, where does the extra C go? (3) Related to two, what controls the fraction of SOM that is cycling and thus, is providing the nutrients required to build biomass. This is the most critical question, because C fixation is ultimately controlled by the mineral nutrient cycle. I will illustrate these three aspects of the terrestrial carbon cycle.

D412 ORAL-0121: Climate-related large-scale variation in forest carbon turnover rate - Evaluating global vegetation models using

remote sensing products of biomass and NPP

Martin Thurner2, 1, Christian Beer2, 1, Nuno Carvalhais3, 4, Philippe Ciais5, Matthias Forkel6, Andrew Friend7, Akihiko Ito8, Axel Kleidon3, Mark Lomas9, Shaun Quegan9, Tim Tito Rademacher7, Maurizio Santoro10, Sibyll Schaphoff11, Christiane Schmullius12, Markus Tum13, Andy Wiltshire14

1Stockholm University, Stockholm, Sweden 2Bolin Centre for Climate Research, Stockholm, Sweden 3Max Planck Institute for Biogeochemistry, Jena, Germany 4CENSE, Departamento de Ciências e Engenharia do Ambiente, Faculdade de Ciências e Tecnologia, Universidade NOVA de Lisboa, Caparica, Portugal 5Laboratoire des Sciences du Climat et de l’Environnement (LSCE), Gif-sur-Yvette, France 6Department of Geodesy and Geoinformation, Technische Universität Wien, Wien, Austria 7Department of Geography, University of Cambridge, Cambridge, United Kingdom 8National Institute for Environmental Studies, Tsukuba, Japan 9School of Mathematics and Statistics, University of Sheffield, Sheffield, United Kingdom 10Gamma Remote Sensing, Gümligen, Switzerland 11Potsdam Institute for Climate Impact Research, Potsdam, Germany 12Department of Earth Observation, Friedrich Schiller University, Jena, Germany 13German Aerospace Center (DLR), German Remote Sensing Data Center (DFD), Wessling, Germany 14Met Office Hadley Centre, Exeter, United Kingdom

Vegetation carbon turnover, in terms of its spatial variation and its response to climate change, is one of the most important, but also most uncertain carbon fluxes in terrestrial ecosystems. Its measurement is hardly possible by inventory studies alone, due to several reasons: First, vegetation carbon turnover involves a variety of processes, including litterfall, background mortality, and mortality by all kinds of disturbances, affecting single biomass compartments, individual trees or even whole ecosystems. Second, these processes act on very different timescales, involving short-term extreme events and long-term responses, and spatial scales, from local extremes to global impacts. In order to capture this variety of processes, spatial scales and timescales, here we estimate forest carbon turnover rate from novel remote sensing products of NPP and biomass. We observe an increase in turnover rate with colder and longer winters in boreal forests, whereas in temperate forests the spatial gradients in turnover rate are related to the length of both warm and dry periods. An evaluation of a set of global vegetation models (GVMs) participating in the Inter-Sectoral Impact Model Intercomparison Project (ISI-MIP; including HYBRID4, JeDi, JULES, LPJml, ORCHIDEE, SDGVM, VISIT) shows that those models are able to reproduce the observation-based spatial relationships only to a limited extent. Deviations from the observation-based turnover rates can be mostly attributed to severe overestimations of biomass, however also important differences in the simulated spatial patterns in NPP become visible. These results highlight that the representations of mortality mechanisms in GVMs need to be improved in order to better match observed carbon stock spatial patterns and finally enable a better informed prediction of future land carbon cycle feedbacks to climate change. Current shortcomings are expected to lead to underestimation of climate change induced intensification in forest mortality and of the resulting positive feedback to climate change.

D413 - ORAL-0376: Advancing mechanistic representation of Photosynthesis and Respiration in the GFDL LM4 land model: implications for the uncertainty in the future carbon cycling

Paul Gauthier1, Elena Shevliakova2, Sergey Malyshev2, Stephen Pacala1

1Princeton University, Princeton, NJ, The United States of America 2Geophysical Fluid Dynamics Laboratory, Princeton, NJ, The United States of America

Plant respiration is a key factor determining World ecosystems productivity and functioning and a major flux of terrestrial CO2 to atmosphere. Under optimal conditions, 20-80% (~ 60 GtC/year) of gross primary production is released back to the atmosphere by plant respiration. However, in the present global terrestrial carbon cycling models and Earth system models the representation of respiration remains very simplistic. Typically, it is assumed to be just a consumption of assimilated carbon (i.e. a fraction of gross photosynthesis). With this oversimplification of both respiration and photosynthesis, models are likely to underestimate uncertainty in global Carbon budget and to unrealistically project ecosystems’ productivity in high latitude. By improving photosynthesis and respiration parameterizations of GFDL land model LM4, a demonstration of the sensitivity of the model on these parameterizations will be presented and interpreted in the context of light use efficiency within the canopy for Tropical, Temperate, Boreal and Arctic ecosystems.

D414 ORAL-0339: Precipitation, CO2, and the savanna - forest transition in Amazonia

Anders Ahlström1, 2, Minchao Wu1, Josep G. Canadell3, Guy Schurgers4, Benjamin Smith1, Joeseph A. Berry5, Kaiyu Guan6, Rob B. Jackson2

1Lund University, Lund, Sweden 2Stanford University, Stanford, The United States of America 3Global Carbon Projetc, Canberra, Australia 4University of Copenhagen, copenhagen, Denmark 5Carnegie Institution for Science, Stanford, The United States of America 6University of Illinois at Urbana Champaign, Urbana, The United States of America

The Amazon rainforest has a disproportionate significance for global CO2 storage and biodiversity. Earth system models (ESMs) that estimate climate and vegetation show little agreement in simulations of the Amazon rain forest. Here we show that evapotranspiration (ET), gross primary productivity (GPP) and above ground biomass (AGB) in both models and empirical data align on a functional relationship with annual precipitation along basin scale gradients. ESMs show generally poor evaluation against empirical datasets, but good agreement when examined with respect to the simulated annual precipitation. Our analysis points to biases in internally generated climate as the cause of the poor performance of ESMs in simulating vegetation structure, carbon and water fluxes. Both ESMs and empirical datasets show a clear breakpoint at ~2000 mm annual precipitation, where the system transitions between water and radiation limitation of annual ET. Future elevated CO2 may increase plant water use efficiency and shift GPP upwards, but the ESMs predict no significant changes in the location of the breakpoint. A detailed regional ESM that include an individual based ecosystem model at high spatial resolution is used to investigate the future stability of the breakpoint in detail, and how it may respond to climate change, CO2, land use and their various feedbacks.

D415 POSTER-0109: Nutrients limitation of the tropical forest carbon cycle

Marie Combe1, Daniel Goll2, Marijn Bauters1, Philippe Ciais2, Ivan Janssens3, Pascal Boeckx1, Hans Verbeeck1

1Ghent University, Ghent, Belgium 2Laboratoire des Sciences du Climat et de l'Environnement (LSCE), Gif-sur-Yvette, France 3University of Antwerp, Antwerpen, Belgium

Tropical forests contribute to about half of the global forest carbon (C) sink, however little is still known about the impact of nutrient limitation on this tropical land sink. A better understanding of the nitrogen (N) and phosphorus (P)-limitation in the tropics could thus help reduce uncertainties of the global land C sink. Since the industrial revolution, human activities have significantly increased the availability of C and reactive N due to fossil fuel burning and the industrial/agricultural fixation of atmospheric N2, while the availability of P is hardly affected in many natural ecosystems. The stoichiometric disequilibrium between C:P and N:P is now unprecedented in Earth's history. Unfortunately, the consequences of this disequilibrium on the functioning of the biosphere are poorly understood. How is the carbon balance of tropical forests affected by the increasing imbalance between P and C and N cycle? In this project, we use the process-based vegetation model ORCHIDEE to reproduce nutrients and carbon cycling data from several forest sites of French Guiana and the Democratic Republic of Congo. We then upscale our site-simulations of the N- and P-limitation to the entire Amazon and Congo basins, and test if nutrient availability can explain recent trends in the C cycle (e.g. the declining C sink of the Amazon).

D416 POSTER-0187: Improving the terrestrial carbon cycle simulated by BNU-ESM

Duoying Ji, Qian Zhang

The BNU-ESM (Beijing Normal University Earth System Model) uses the Common Land Model (CoLM) as its land component, which adopts a terrestrial carbon scheme based on the Lund–Potsdam–Jena (LPJ) dynamic global vegetation model to simulate vegetation distribution, species succession and related carbon pools change under climate change. The model can reproduce the main climatic variables controlling the spatial and temporal characteristics of the terrestrial carbon cycle. The major simulation biases include overestimated leaf area index and underestimated high-latitude soil organic carbon stocks as many other earth system models. In the new version of BNU-ESM, several improvements are implemented to reduce systematic biases in terrestrial carbon cycle. The dynamic global vegetation model is calibrated against observational datasets, which can simulate more reasonable spatial vegetation pattern, leaf area index and improved biomass allocation between leaf, stem and root carbon pools. The soil carbon scheme is improved with considering vertical distribution, and simulates better carbon accumulations and dynamics of organic matter over high-latitude regions. Especially, with introducing soil supercool water scheme, the model can simulate better permafrost extent compared to IPA permafrost map, which also helps to reduce soil carbon biases in cold regions.

D417 - POSTER-0377: Asymmetric Responses of Terrestrial Ecosystems to Climate Changes in China from 1982 to 2010

Junbang Wang1, Shaoqiang Wang1, Lei Zhou1, Honglin He1, Li Zhang1, Xiaoli Ren1, Hao Yan2, Fengxue Gu3, Guirui Yu1

1Synthesis Research Center of China's Ecosystem Research Network, Key Laboratory of Ecosystem Network Observation and Modeling, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, Beijing, China 2National Meteorological Center, China Meteorological Administration, Beijing, China 3Institute of Environment and Sustainable Development in Agriculture, Chinese Academy of Agricultural Sciences, Beijing, China

Chinese terrestrial ecosystems act as an important carbon sink offsetting 28–37 per cent of its cumulated fossil carbon emissions (Piao et al., 2009). Temperature and precipitation are two key driver and their changes in future may alter the carbon balance of terrestrial ecosystem. Whether the ecosystem act as a carbon source or sink in future climate projection, is greatly dependent on understanding on the responses of carbon absorption and release to climate change. The four validated ecosystem process model, BEPS, TEC, CEVSA and GLOPEM-CEVSA, were applied to analysis the responses of the simulated gross primary production (GPP) and ecosystem respiration (RE) to climate changes and its uncertainties were assessed by comparing with literatures and the multipscale synthesis and terrestrial model intercomparison project (MsTMIP). The results showed that GPP and RE will increase 3.0% and 4.0% respectively with a warming climate of 1 °C, which resulted a decreasing net ecosystem production (NEP) of 2.8%. Meanwhile, under precipitation increasing of 100 mm, the increasing of 6.8% and 1.9% in GPP and RE will result a more NEP of 9.2%. However, some differences were found among models, which could source to the model structure. But a further uncertainty assessment should trace the modules from photosynthesis, autotrophic respiration, and carbon allocation, to soil organic decomposition, even its theory and hypothesis. The studies have great implication to understand the carbon cycles mechanism of terrestrial ecosystems, parameterize biogeochemical cycle models, and predict the potential change of ecosystem in the future climate change.

D418 - POSTER-0379: Implementing surface boundary conditions in the Ent Terrestrial Biosphere Model: evaluation of land-surface albedo and sensitivity to forcing leaf area index

Carlo Montes, Nancy Kiang, Wenge Ni-Meister, Wenze Yang, Crystal Schaaf, Igor Aleinov, Qingsong Sun, Dominique Carrer

Land surface albedo is a major controlling factor in vegetation-atmosphere transfers, modifying the components of the energy budget, the ecosystem productivity and patterns of regional and global climate. General Circulation Models (GCMs) are coupled to Dynamic Global Vegetation Models (DGVMs) to solve vegetation albedo by using simple schemes prescribing albedo based on vegetation classification, and approximations of canopy radiation transport for multiple plant functional types (PFTs). In this work, we aim at evaluating the sensitivity of the NASA Ent Terrestrial Biosphere Model (TBM), a demographic DGVM coupled to the NASA Goddard Institute for Space Studies (GISS) GCM, in estimating VIS and NIR surface albedo by using variable forcing leaf area index (LAI). The Ent TBM utilizes a new Global Vegetation Structure Dataset (GVSD) to account for geographically varying vegetation tree heights and densities, as boundary conditions to the gap-probability based Analytical Clumped Two-Stream (ACTS) canopy radiative transfer scheme (Ni-Meister et al., 2010). High resolution (1 km x 1 km) land surface and vegetation characteristics for the Ent GVSD were obtained from earth observation platforms and algorithms, including the Moderate Resolution Imaging Spectroradiometer (MODIS) land cover and plant functional types (PFTs) (Friedl et al., 2010), MODIS-derived soil albedo (Carrer et al., 2014), and vegetation height from the Geoscience Laser Altimeter System (GLAS) on board ICESat (Ice, Cloud, and land Elevation Satellite) (Simard et al., 2011; Tang et al., 2014). Two LAI products are used as input to ACTS/Ent TBM: Beijing Normal University LAI (Yuan et al., 2011) and Global Data Sets of Vegetation (LAI3g) (Zhu et al. 2013). The sensitivity of 1 km resolution Ent TBM albedo simulations to LAI products is assessed, compared against the previous GISS GCM vegetation classification and prescribed Lambertian albedoes (Matthews, 1984), and against MODIS black-sky and white-sky albedo, and upscaled to GCM 2ºx2.5º coarser resolution.