Biogeosciences [B]

 CC:714B  Monday  0800h

CarboNA: Continental Carbon Cycle Studies in North America III

Presiding:  N Cavallaro, U.S. Department of Agriculture; A Muhlia, CIBNOR, S.C.; C Ste-Marie, Canadian Forest Service


Proposed Gulf of Mexico Intensive Study on Carbon Fluxes

* Coble, P G (, College of Marine Science, University of South Florida, 140 Seventh Ave.S., St. Petersburg, FL 33701, United States
Robbins, L (, Florida Integrated Science Center, St. Petersburg, U.S. Geological Survey, 400 Sixth St. S., St. Petersburg, FL 33701, United States
Lohrenz, S (, Department of Marine Science, The University of Southern Mississippi, Stennis Space Center, MS 39529, United States
Cai, W (, Department of Marine Sciences, The University of Georgia, Athens, GA 30602, United States

The Gulf of Mexico is an ideal site for the study of land-ocean carbon cycle coupling processes. A recent synthesis suggests that Gulf of Mexico air-sea CO2 flux may dominate the net flux of the entire North American margin because of the Gulf's large size and strong carbon signals. Northern Gulf waters appear to be a strong local CO2 sink due to high primary productivity stimulated by river input of anthropogenic nutrients from the North American continent. Nutrient discharge from the Mississippi River has been implicated in widespread hypoxia on the shelf. The surface drainage system of the Gulf covers more than 60% of the U.S. and more than 40% of Mexico; thus, large-scale changes in land-use and water-management practices in both countries, as well as changes in temperature and rainfall due to climate change, will profoundly affect Gulf carbon fluxes. Nevertheless, major sources of uncertainty in the North American carbon budget remain because of largely unsampled areas, undocumented key fluxes, such as air-sea exchange of carbon dioxide, associated carbon fluxes, and poorly characterized control mechanisms. An intensive study in which the Gulf is considered as a whole system, including watersheds, margins, open Gulf of Mexico, overlying atmosphere, and underlying sediments, will be discussed. The study is best addressed using a three-pronged approach that incorporates remote sensing observations, field observations and experiments, and physical and biogeochemical modeling. Societal issues related to carbon management and land-use/land-change must be an integral part of such a study. International cooperation with Mexico, Canada, and Cuba will be essential for the success of this study.


Inter-annual Variability in Net Ecosystem Exchange of Carbon Dioxide and Methane Emissions in a Temperate Freshwater Marsh

* Bonneville, M (, Department of Natural Resource Sciences, McGill University, 21111 Lakeshore Rd., Ste Anne de Bellevue, QC H9X 3V9, Canada
Strachan, I (, Department of Natural Resource Sciences, McGill University, 21111 Lakeshore Rd., Ste Anne de Bellevue, QC H9X 3V9, Canada

There exists very little information on greenhouse gas (GHG) exchange in marsh wetlands, especially in temperate climates. Measurements of carbon dioxide (CO2) and methane (CH4) fluxes were made from May 2005 to June 2008 in a temperate freshwater cattail marsh in Eastern Ontario, Canada. The net ecosystem exchange (NEE) of CO2 was measured continuously using the eddy covariance technique, and closed chambers were used to measure CH4 emissions from open water, soil, and vegetated portions of the marsh. Based on NEE, we found that the marsh accumulated 264 g C m-2 from May 2005 to April 2006 and 185 g C m-2 and 308 g C m-2 in 2006-2007 and 2007-2008, respectively. Lower spring temperature in 2005 seems to have delayed the initial growth of cattails and therefore led to a later switchover time from a net CO2 source to a net CO2 sink compared to spring 2006 and 2007. The lower cumulative NEE measured in 2006-2007 is mainly due to the cloudy conditions (i.e. low average incoming photosynthetically active radiation) that occurred through late summer and early fall 2006, which greatly decreased cattail photosynthesis and induced an earlier death of the pants, which in turn resulted in a lower average CO2 uptake compared to the other years. During the 2005, 2006 and 2007 growing seasons, the carbon uptake period was 109, 104, and 116 days in length, which is consistent with the inter-annual variability in NEE observed. The results suggest that the timing of the fall switchover from a net CO2 sink to a net CO2 source is probably the main factor influencing the annual CO2 accumulation. The average CH4 flux measured from open water was 658 mg CH4 m-2 d-1 in 2005, 381 mg CH4 m-2 d-1 in 2006, and 352 mg CH4 m-2 d-1 in 2008. The average CH4 flux from vegetation was 1001 mg CH4 m-2 d-1 in 2005, 1640 mg CH4 m-2 d-1 in 2006, and 1260 mg CH4 m-2 d-1 in 2008. The CH4 flux from soil was only measured in 2006 (255 mg CH4 m-2 d-1) and 2008 (224 mg CH4 m-2 d-1). It is known that the presence of the aerenchyma tissue that runs through the cattail's roots, stem and leaves facilitates the transport of gases such as CH4 from the production site (i.e. soil) to the atmosphere, which explains the higher fluxes measured from vegetated portions of the marsh. Knowing that CH4 is produced under anaerobic waterlogged conditions, the higher CH4 fluxes measured from plants in 2006 may be related to the higher rainfall observed during that year. This long-term record of CO2 and CH4 fluxes combined with continuous measurements of meteorological and ecosystem properties allows us to investigate the overall carbon budget of this marsh wetland as well as the major controls on the GHG fluxes for this type of ecosystem. This study provides valuable knowledge that can complement the existing information on other wetlands types and may be useful for predicting the impacts of climate change on CO2 and CH4 fluxes and for estimating national carbon stocks.


The Coupling of Methane and Carbon Dioxide Fluxes From Boreal Lakes and Reservoirs

* Prairie, Y T (, Department of BIological Sciences, UQAM, Box 8888, station Centre-Ville, Montreal, Qc H3C 3P8, Canada
del Giorgio, P A (, Department of BIological Sciences, UQAM, Box 8888, station Centre-Ville, Montreal, Qc H3C 3P8, Canada

The flooding of land for the construction of hydropower dams alters considerably the sources, amounts and dynamics of carbon degradation processes occurring in these aquatic systems. In natural boreal lakes of the Eastmain region, diffusive methane flux at the air-water interface follows a constant relationship with carbon dioxide flux, methane accounting for about 1.1 percent of the total carbon gas flux. In the early years after flooding, the fluxes of both CO2 and CH4 flux in the Eastmain-1A reservoir were greatly enhanced relative to natural lakes and the relative contribution of CH4 surprisingly decreased to about 0.5 percent of CO2 fluxes. In this paper, we explore whether the decreased relative importance of methane in conditions of higher overall organic matter decomposition rate is attributable to decreased methanogenesis or to increased water column methane oxidation. In both reservoirs and natural lakes, methane flux remain a highly predictable function of carbon dioxide flux, opening the possibility of easily inferring methane flux from CO2 surveys, and upscaling to the regional scale.


Evaluating Contributions of Wetland and Lake Emissions of Methane to Atmospheric Methane Concentrations with a Processed-Based Biogeochemistry Model and an Atmospheric Chemistry Transport Model and Satellite Retrieval Data in Northern High Latitudes

* Zhuang, Q (, Purdue University, Department of Earth & Atmospheric Sciences and Department of Agronomy, West Lafayette, IN ,
Tang, J (, Purdue University, Department of Earth & Atmospheric Sciences and Department of Agronomy, West Lafayette, IN ,
Lu, Y (, Purdue University, Department of Earth & Atmospheric Sciences and Department of Agronomy, West Lafayette, IN ,
Xu, K (, Purdue University, Department of Earth & Atmospheric Sciences and Department of Agronomy, West Lafayette, IN ,
Xiong, X (, National Environmental Satellite, Data, and Information Service (NESDIS), NOAA, Center for Satellite Applications and Research (STAR,
Melillo, J (, The Ecosystems Center, Marine Biological Laboratory, Woods Hole, MA,
Prinn, R (, Massachusetts Institute of Technology, Cambridge, MA,
McGuire, A (, University of Alaska Fairbanks, Fairbanks AK,

Northern high latitudes (north of 45oN) contain vast areas of wet tundra, wetlands, and water bodies (lakes and ponds), that are emitting a large amount of methane to the atmosphere each year. To date, the magnitudes and the inter-annual variations of these emissions are uncertain. The seasonal variations of these emissions due to changes of inundation of wetlands, the effects of spring thaw and winter freezing, and the effects of permafrost degradation in lands and lakes are also uncertain. In addition, how these emission dynamics affect the temporal and spatial distributions of atmospheric methane concentrations is not yet well understood. Here we use a process-based biogeochemistry model called TEM (the Terrestrial Ecosystem Model) to quantify these emissions with three different datasets for wetland and lake distributions. The effects of spring thaw and winter freezing on methane emissions are also incorporated into the TEM simulations. These estimated emissions together with wildfire emissions are then incorporated into a 3-D atmospheric chemistry transport model (GEOS-Chem) to simulate atmospheric methane concentration profiles. The satellite retrieval data of AIRS are then compared with the simulated methane concentration profiles for the region. We find that the current regional land and lake methane emissions range from 65 to 150 Tg per year. Seasonal changes of emissions of land and lakes due to spring thaw and winter freezing, together with fire emissions, play a significant role in determining the seasonal atmospheric methane concentration profiles simulated with GEOS-Chem. Comparison between multiple GEOS-Chem simulations driven with different seasonal dynamics of land, lake, and wildfire emissions and AIRS retrievals, suggests that accurately simulating the timing of these emissions as affected by spring thaw and winter freezing and wildfires, is critical for GEOS- Chem to capture the atmospheric concentration profiles over the region.


North American Carbon Project (NACP) Regional Model-Model and Model-Data Intercomparison Project

* Huntzinger, D N (, University of Michigan, 1351 Beal Avenue, Ann Arbor, MI 48109, United States
Post, W M (, Oak Ridge National Laboratory, Environmental Sciences Division, Oak Ridge, TN 37831, United States
Jacobson, A R (, Cooperative Institute for Research in Environmental Sciences, University of Colorado, UCB 216, Boulder, CO 80309, United States
Jacobson, A R (, NOAA Earth Science Research Laboratory, 325 Broadway, Boulder, CO 80305, United States
Cook, R B (, Oak Ridge National Laboratory, Environmental Sciences Division, Oak Ridge, TN 37831, United States

Available observations are localized and widely separated in both space and time, so we depend heavily on models to characterize, understand, and predict carbon fluxes at regional and global scales. The results from each model differ because they use different approaches (forward vs. inverse), modeling strategies (detailed process, statistical, observation based), process representation, boundary conditions, initial conditions, and driver data. To investigate these differences we conducted a model-model and model-data comparison using available forward ecosystem model and atmospheric inverse output, along with regional scale inventory data. Forward or "bottom-up" models typically estimate carbon fluxes through a set of physiological relationships, and are based on our current mechanistic understanding of how carbon is exchanged within ecosystems. Inverse or "top-down" analyses use measured atmospheric concentrations of CO2, coupled with an atmospheric transport model to infer surface flux distributions. Although bottom-up models do fairly well at reproducing measured fluxes (i.e., net ecosystem exchange) at a given location, they vary considerably in their estimates of carbon flux over regional or continental scales, suggesting difficulty in scaling mechanistic relationships to large areas and/or timescales. Conversely, top-down inverse models predict fluxes that are quantitatively consistent with atmospheric measurements, suggesting that they are capturing large scale variability in flux quite well, but offer limited insights into the processes controlling this variability and how fluxes vary at fine spatial scales. The analyses focused on identifying and quantifying spatial and temporal patterns of carbon fluxes among the models; quantifying across-model variability, as well as comparing simulated or estimated surface fluxes and biomass to observed values at regional to continental scales for the period 2000-2005. The analysis focused on the following three questions: 1. Do model results and observations show consistent spatial patterns in response to the 2002 drought? From measurements and model, can we infer what processes were affected by the 2002 drought? 2. What is the spatial pattern and magnitude of interannual variation in carbon sources and sinks? What are the components of carbon fluxes and pools that contribute to this variation? 3. What are the magnitudes and spatial distribution of carbon sources and sinks, and their uncertainties during the period 2000-2005? Examining and comparing results of inverse and forward model simulations with each other and with suitable benchmark spatial measurements help evaluate model strengths/weaknesses and utility, thereby providing multiple views of spatial and temporal patterns of fluxes, leading to better understandings of processes involved, and providing an improved basis for making projections.


The Uncertainty in 20th Century Carbon Budget due to Land Use Change Emissions

* Arora, V K (, Canadian Centre for Climate Modelling and Analysis, Environment Canada, University of Victoria, PO Box 3065, STN CSC, Victoria, BC V8W3V6, Canada
Boer, G J (, Canadian Centre for Climate Modelling and Analysis, Environment Canada, University of Victoria, PO Box 3065, STN CSC, Victoria, BC V8W3V6, Canada

The uncertainty in 20th century carbon budget due to land use change (LUC) emissions is assessed using the Canadian Centre for Climate Modelling and Analysis (CCCma) first generation Earth System Model (CanESM1). CanESM1 is based on CCCma third generation coupled general circulation model and includes terrestrial and oceanic carbon cycle components. LUC emissions are modelled interactively on the basis of specified changes in land cover that determine the amount of deforestation as well as the corresponding albedo changes at the land surface. Eight fully coupled climate-carbon cycle simulations are performed using different reconstructions of the 1850-2000 land cover that are based on historical data sets of increase in cropland and pasture area. The reconstructions of 1850-2000 land cover are based on two approaches: in the linear approach the changes in fractional coverage of natural plant functional types (PFTs) are in proportion to changes in cropland and/or pasture area and in the rule-based approach the natural PFTs are deforested in a specified order. These simulations allow to estimate implied LUC emissions, the contribution of increase in cropland versus pasture area on LUC emissions, the uncertainty associated with using different historical data sets of crop area as well as the manner in which the historical land cover is reconstructed. The amount of deforested biomass for the 1850-2000 period ranges from 63 Pg C for the case where only increase in cropland area is taken into account following the rule-based approach to 145 Pg C where increases in cropland and pasture area are both taken into account following the linear approach. In absence of historical LUC the CO2 concentration in the atmosphere is about 20 ppm below the observation-based value of ~370 ppm in the year 2000. Inclusion of increase in pasture area although increases the amount of deforested biomass it does not change the atmospheric CO2 substantially because pastures also sequester CO2 in soil carbon. Overall the atmospheric CO2 in most simulations with LUC varies between 370±4.5 ppm depending on the cropland/pasture data set used and the manner in which historical land cover is reconstructed.


North America Terrestrial Carbon Budget: A Model Analysis of the Combined Effects of CO2, Nitrogen, Climate, Land Use Changes and Management

* Jain, A (, University of Illinois, 105 South Gregory Street, Urbana, IL 61874, United States
Yang, X (, University of Illinois, 105 South Gregory Street, Urbana, IL 61874, United States

Several mechanisms have been proposed to explain present trends in the terrestrial uptake of CO2. These mechanisms include physiological responses of terrestrial ecosystems to increasing ambient CO2 concentrations, anthropogenic N deposition, and variations in productivity due to climate variability and changes in land-use and management. Each of these mechanisms may be playing a significant role in the global CO2 budget. In addition, changes in soil management can potentially increase the accumulation of soil organic carbon (SOC) in addition to anthropogenic disturbances, which include clearing of land for agriculture, conversion of forest to pasture, and harvest of forest products. We present the concurrent effects of all important ecosystem processes and anthropogenic disturbances and management practices on North America terrestrial carbon budget, for the historical period 1900-2000, using an Integrated Science Assessment Model (ISAM), a geographically explicit advanced terrestrial ecosystem model which simulates the carbon and nitrogen fluxes to and from different compartments of the terrestrial biosphere with 0.5-by-0.5 degree spatial resolution.