Ameriflux Annual PI Meeting Abstracts

Links: Meeting Agenda | Abstracts | List of Participants | Meeting Packet | Breakout Posters

Dennis Baldocchi (, UC Berkeley

Exhortation for more synthesis activities (Plenary invited talk)

With Sebastian Wolf

We will discuss past, current and future synthesis activities based on the AmeriFlux dataset. The main goal of this talk is to stimulate new ideas and activities with the new and growing dataset.

Deborah Huntzinger (, Northern Arizona University

Using observational data to evaluate global terrestrial biospheric models: challenges and opportunities? (Plenary invited talk)

With Christopher Schwalm, School of Earth Sciences and Environmental Sustainability, Northern Arizona University; Anna Michalak, Department of Global Ecology, Carnegie Institution for Science, Stanford;; Joshua Fisher, NASA Jet Propulsion Laboratory; Benjamin Poulter, Department of Ecology, Montana State University; Yaxing Wei, Environmental Sciences Division, Oak Ridge National Laboratory; Robert Cook, Environmental Sciences Division, Oak Ridge National Laboratory; Kevin Schaefer, National Snow and Ice Data Center, Boulder, Colorado; Andrew Jacobson, NOAA Earth System Research Lab Global Monitoring Division and MsTMIP modeling teams.

Terrestrial biosphere models (TBMs) have become an integral tool for extrapolating local observations and understanding of land-atmosphere carbon exchange to larger regions. The North American Carbon Program (NACP) Multi-scale synthesis and Terrestrial Model Intercomparison Project (MsTMIP) is a formal model intercomparison effort focused on improving the diagnosis and attribution of carbon exchange at regional and global scales. MsTMIP builds upon current and past synthesis activities, and has a unique framework designed to isolate, interpret, and inform understanding of how model structural differences impact model estimates of carbon uptake and release. Existing estimates of land-atmosphere carbon exchange from TBMs vary widely and models disagree on the strength of the global net land sink. Uncertainties in, or variations among, TBM estimates are driven by a complex combination of assumptions, scientific hypotheses, and model choices. Yet there are no direct observations of land-atmosphere carbon flux at the spatial resolutions (e.g., 0.5° by 0.5°) and scales (e.g., continental to global) at which TBMs are generally applied. This greatly complicates the evaluation of different model approaches and the use of model-data comparisons to inform model development. This presentation provides a brief overview of the MsTMIP activity, as well as ways flux tower data has been used in the past to evaluate TBM performance, drawing from recent model-data comparisons such as the NACP interim synthesis activates. Drawing from these model-data synthesis activities, I discuss potential pathways (and challenges) for using AmeriFlux-based observations for evaluating global and continental scale models. The ultimate goal of this talk is to foster discussion on ways to improve and expand upon the use of site-level observations in TBM benchmarking.

Funding for this project was provided through NASA ROSES Grant # NNX10AG01A. Data management support for preparing, documenting, and distributing model driver and output data was performed by the Modeling and Synthesis Thematic Data Center at Oak Ridge National Laboratory (, with funding through NASA ROSES Grant # NNH10AN68I.

Adrien Finzi (, Boston University

Utility and vision for coupling belowground carbon dynamics to eddy-covariance measurements of ecosystem CO2 exchange with the atmosphere (Presentation)

With Marc-Andre Giasson, Allison L. Gill, Rose Z. Abramoff, Department of Biology, Boston University, Boston MA

Terrestrial ecosystems exchange large quantities of carbon with the atmosphere at daily to decadal time scales. A large proportion of this carbon is allocated belowground but there is presently little information on how partitioning varies across the growing season or among biomes. Using NEE and soil respiration data from a synthesis of Harvard Forest EMS and Hemlock tower data we show significant temporal patterning in above- vs. belowground respiration that is related to phenological events aboveground. We extend this analysis spatially and temporally through a meta-analysis to show that the mean offset [in days] between the peak in root and shoot production increases significantly from the tropics to the boreal forest. We also compiled a global-scale data base of eddy covariance data, total belowground C flux and soil resource availability [N, P, water] to show that there is >2-fold variation in belowground C allocation as a proportion of GPP and up to 11-fold variation in the carbon cost of soil resource acquisition. We conclude by suggesting a concerted effort towards belowground research at the AmeriFlux sites would greatly enhance our understanding of belowground C flux and their influence on the largest pool of carbon in the ecosystem, the soil.

This research was supported by Office of Science (BER), US Department of Energy, Terrestrial Ecosystem Science Program (SC0006916) and the National Science Foundation Long-Term Ecological Research Program (131706-5056666). Additional support was provided by the Office of Science, Graduate Fellowship Program to ALG.

Julie Shoemaker (, Harvard University

Combining soil scale and tower scale fluxes to understanding below-ground processes (Presentation)

Co-authors (pending)

Abstract text (pending)

Sebastian Wolf (, University of California Berkeley

With Dennis Baldocchi (UC Berkeley), Joshua B. Fisher (Jet Propulsion Laboratory, California Institute of Technology), Trevor F. Keenan (Macquarie University), Ankur Desai (University of Wisconsin – Madison), Russ Scott (USDA-ARS Southwest Watershed Research Center), Marcy Litvak (University of New Mexico), Andrew Richardson (Harvard University), Nathaniel Brunsell (University of Kansas), Bev Law (Oregon State University)

Synthesis of the 2012 U.S. drought: impact on ecosystem fluxes and implications for the future (Presentation)

Drought severely impacts biosphere-atmosphere carbon and water fluxes of terrestrial ecosystems by reducing productivity, carbon uptake and water transport to the atmosphere. The 2012 US drought was among the most intense and widespread drought events in the US since the “Dust Bowl” period in the 1930s. Drought conditions started developing during an exceptionally warm spring, intensified throughout the summer and were most severe in the Central US (Great Plains), with devastating effects on agricultural production. We synthesize the impact of the 2012 drought on ecosystem carbon and water fluxes across the Contiguous United States using eddy covariance data from 30 AmeriFlux sites and remote sensing data from MODIS. We found widespread reductions in gross primary productivity and evapotranspiration of more than 30% in the Great Plains and beyond during the summer months. Drought intensity and duration were directly linked to these changes in ecosystem fluxes. As drought frequencies and intensities are predicted to increase in the future, we discuss the implications of our results for drought susceptibilities of different land-use types.

We appreciate data contributed by AmeriFlux site PIs J. Baker, K. Bible, P. Blanken, G. Bohrer, D. Bowling, K.L. Clark, T. Griffis, D. Hollinger, S. Ma, A. Noormets, S.A. Papuga, F. Rahman, A.S. Suyker, M. Torn, S. Wharton. Support was provided by the AmeriFlux Management Team at the Lawrence Berkeley National Lab (LBNL), the Carbon Dioxide Information Analysis Center (CDIAC) at Oak Ridge National Lab (ORNL), Bai Yang (CDIAC/ORNL), Dan Ricciuto (ORNL), and the High Plains Regional Climate Center (HRPCC). Funding sources included the European Commission (Marie Curie Fellowship) and the U.S. Department of Energy (DOE).

Fred Huemmrich (, NASA

With T. Hilker, L. Corp, E. Middleton, P. Campbell, Y-B. Cheng, Q. Zhang, N. Coops, W. Kustas, A. Russ, T.A. Black, A. Barr

Using optical signals to determine carbon fluxes (Presentation)

Understanding the dynamics of ecosystem carbon cycling requires an accurate determination of the spatial and temporal distribution of photosynthetic CO2 uptake by vegetation. Optical sampling using spectral reflectance can provide information on the physiological status of vegetation, which can be related to these carbon fluxes. Near surface optical measurements can provide temporally frequent observations that capture a range of physiological states due to diurnal and seasonal cycles. This study examines the use of spectral reflectance to determine inputs to a light use efficiency model, driven only by optical inputs. Examples will be shown using reflectance data collected using tower-mounted automated spectrometers in conjunction with flux towers at the SK-Old Aspen tower and in a cornfield in Beltsville, MD.

Ankur Desai (, University of Wisconsin-Madison

With Matthias Mauder, KIT IMK-IFU, Stefan Metzger, NEON, Inc., Mike Dietze, Boston University

How do we make Ameriflux useful for ecosystem models? (Presentation)

The proliferation of eddy covariance flux towers for long term observation of ecosystem carbon and water cycling has been a boon to the carbon budget and ecosystem modeling communities. Numerous syntheses have documents latitudinal, climatic, and biotic controls of photosynthetic and respiratory processes across tower sites and how they compare to models. However, many of these efforts have been limited by the types of data analyzed and faced legitimate criticisms on the quality, availability, breadth, and reliability of flux data products for synthesis, upscaling, and model parameterization. As Ameriflux transitions to a new era of a distributed facility of core and affiliated sites, how can we make flux towers more useful for model parameterization and evaluation?

In particular, the community needs to move away from a primary focus on flux gap-filling, u* filtering, and flux GPP/RE partitioning, to a more comprehensive focus on providing the highest quality set of flux and meteorological observations with assessment of random and systematic uncertainty, representativeness, and provenance. Additionally, a method for systematic collection and machine-readable reporting of vegetation structure, physiology, soil biophysics, and component biogeochemical fluxes needs to be standardized soon. Recent advances in methods for deriving flux uncertainty (Salesky et al., 2012), quality control metrics (Mauder et al., 2013), footprint-weight spatial rectification (Metzger et al., 2013), meteorological gap-filling, and methods for assessing the value of flux and biological observations for ecosystem models (Dietze et al., 2014) can all be used to strengthen the observational database we provide to the community. Further, some of these new methods require analysis with high-frequency turbulent observations (e.g., 10 Hz) and call for systematic methods to archive and retrieve high-rate data from Ameriflux sites.

DOE LBNL Ameriflux Network Management Program Core Site Subcontract to ChEAS Cluster, NSF ABI #1062547 and #1062204, Karlsruhe Institute of Technology IMK-IFU MICMoR program, and NEON, Inc.

Jingfeng Xiao (, University of New Hampshire

With Kenneth J. Davis, Pennsylvania State University; Nathan M. Urban, Los Alamos National Laboratory; Keller Klaus, Pennsylvania State University; Scott V. Ollinger, NEON; Bobby Braswell, Applied GeoSolutions; Andrew D. Richardson, Harvard University; Donald C. Buso, Cary Institute of Ecosystem Studies; Gene E. Likens, Cary Institute of Ecosystem Studies

Assessing uncertainty of ecosystem models using AmeriFlux observations (Presentation)

In the realm of terrestrial carbon cycling, ecosystem models have been widely used to estimate carbon and water fluxes over various spatial and temporal scales. Uncertainty information of fluxes is essential for carbon and water cycle studies. Modeled fluxes with uncertainty bounds can facilitate the direct comparison between modeled and observed fluxes and the intercomparison of different approaches (e.g., modeling, inventory, atmospheric inversions). The availability of uncertainty information will also help interpret flux dynamics. Better understanding of the sources of uncertainty can also provide insight on future improvement of ecosystem approaches. Model-data fusion (or data assimilation) approaches have been increasingly used to estimate uncertain parameters of carbon cycle models. We used eddy flux observations and model-data fusion approaches to estimate the parameters of a simple Diagnostic Carbon Flux Model and examined the influence of parameter variability on regional flux estimates. Our results showed that parameter uncertainty could lead to substantial uncertainty in regional flux estimates. We also optimized the key parameters of a process-based ecosystem model – PnET-CN using multiple constraints from Bartlett Experimental Forest and Hubbard Brook. Our results show that the optimization based on multiple constraints could quantify the uncertainty of modeled fluxes and could also significantly improve the modelâ€TMs performance for simulating carbon and water fluxes. The parameter uncertainty can be propagated through model simulations to quantify the uncertainty of projections for the remainder of the 21st century.

This work is supported by NASA Terrestrial Ecology, the National Institute for Climatic Change Research (NICCR), Department of Energy (award 14U776), and NSF MacroSystems Biology (award 1065777).


With Paul Hanson, Jeff Riggs Oak Ridge National Laboratory; Stephen G. Pallardy and Kevin Hosman, University of Missouri

Tree mortality, ecosystem carbon uptake and water use under contrasting drought regimes in an Ozark forest (poster)

Droughts play a major role in forest ecosystem structures and functions. Climate change is expected to alter drought regimes characterized by their timing, intensity, duration, and frequency. However, our current understanding of ecological impacts of droughts is still limited due to lack of consistent long-term observations. MOFLUX, which is located along a critical climate gradient and in a biome transition zone, serves as an excellent site for studying forest ecosystem responses to climate variability and extreme events, particularly droughts. Large interannual variation in water availability is the most important climate feature at MOFLUX and controls its annual carbon budget. Although annual precipitation exceeds annual evaporation and transpiration demand, the precipitation does not come at the right time in most years and therefore drought is common. However, the timing, duration and intensity of drought vary from year to year, revealing complicated patterns of ecosystem responses to water availability. We found that integrated predawn leaf water potential is a strong predictor of annual carbon uptake. Also a threshold exists in the control of water availability on ecosystem carbon uptake. This threshold is defined when summer precipitation equals summer evapotranspiration. If summer precipitation is less than this threshold, carbon uptake is entirely determined by water availability while above this threshold, other factors play a role. Interannual variation in soil respiration remains difficult to explain and more observations and below-ground studies are going. It has been also observed that seasonal variation in carbon uptake is more closely related to variation in leaf biochemistry than to photosynthetically active radiation. These findings together with consistent, comprehensive datasets and diversity of climate regimes at MOFLUX provide opportunities to improve and test land surface models.

The support for this research came from the U.S. Department of Energy, Office of Science, Biological & Environmental Research Program (DE-AC05-00OR22725 for ORNL managed by UT-Battelle, LLC).

Stefan Metzger (, National Ecological Observatory Network, Fundamental Instrument Unit, Boulder, Colorado,

With Edward Ayres {National Ecological Observatory Network, Fundamental Instrument Unit, Boulder, Colorado, USA}; Peter Blanken {University of Colorado, Department of Geography, Boulder, Colorado, USA}; George Burba {LI-COR Biosciences, Environmental Division, Lincoln, Nebraska, USA}; Sean Burns {University of Colorado, Department of Geography, Boulder, Colorado, USA}; Ankur Desai {University of Wisconsin Madison, Department of Atmospheric and Oceanic Sciences, Madison, Wisconsin, USA}; Raymond Desjardins {Agriculture and Agri- Food Canada, Eastern Cereal and Oilseed Research Centre, Ottawa, Ontario, Canada}; Michael Dietze {Boston University, Department of Earth and Environment, Boston, Massachusetts, USA}; Andrew Fox {National Ecological Observatory Network, Data Products, Boulder, Colorado, USA}; Natascha Kljun {Swansea University, Department of Geography, Swansea, Wales, UK}; Henry W. Loescher {National Ecological Observatory Network, Data Products, Boulder, Colorado, USA}; Hongyan Luo {National Ecological Observatory Network, Data Products, Boulder, Colorado, USA}; Matthias Mauder {Karlsruhe Institute of Technology, Institute of Meteorology and Climate Research – Atmospheric Environmental Research, Garmisch-Partenkirchen, Germany}; David P. Moore {University of Arizona, Institute of the environment, Tucson, Arizona, USA}; J. William Munger {Harvard University, Center for the Environment, Cambridge, Massachusetts, USA}; Scott Ollinger {National Ecological Observatory Network, Data Products, Boulder, Colorado, USA}; Natchaya Pingintha-Durden {National Ecological Observatory Network, Data Products, Boulder, Colorado, USA}; Hans-Peter Schmid {Karlsruhe Institute of Technology, Institute of Meteorology and Climate Research – Atmospheric Environmental Research, Garmisch-Partenkirchen, Germany}; Penelope Serrano-Ortiz {Centro Andaluz del Medio Ambiente, Instituto Interuniversitario de InvestigaciÃ3n del Sistema Tierra, Granada, Spain}; Jeffrey R. Taylor {National Ecological Observatory Network, Data Products, Boulder, Colorado, USA}; Ke Xu {University of Wisconsin Madison, Department of Atmospheric and Oceanic Sciences, Madison, Wisconsin, USA}; Rommel Zulueta {National Ecological Observatory Network, Data Products, Boulder, Colorado, USA}

Gauging terrestrial environments: first lessons learned from NEON (poster)

One of the purposes of the National Ecological Observatory Network (NEON) is to provide free, high quality data products of the environment to the science community. These data products are designed to facilitate the discovery and understanding of the impacts of climate change, land-use change, and invasive species on ecology. To this end, NEON is currently designing and deploying human-based observations, aircraft remote sensing, and automated measurements in aquatic and terrestrial settings. In some cases, novel approaches are required for better facilitating unbiased measurements across a wide range of biomes and climates. Here, AmeriFlux, ICOS, NEON and other networks have been closely working together and have produced innovative solutions to challenging problems.

This paper provides an overview of NEON’s instrument-based terrestrial measurements and their algorithmic processing, with special focus on the results of collaborative efforts. Here, we show a few examples;

  • Inter-site consistency of tower- and soil-based measurements through scaling law relationships. For example, a geo-statistical sampling strategy is used to determine site-scale soil temperature, moisture etc. at pre-defined levels of accuracy and confidence.
  • Gas analyzer sampling system optimized for eddy-covariance applications. A combination of minimal-volume screened inlet, pleated mesh filter and short, heated intake tube provides up to 250% improved frequency response at 10 Hz for fast H2O and CO2 measurements.
  • Field validation hierarchy for automatically treating infrared gas analyzer drifts. Fast H2O and CO2 measurements are complementary filtered with simultaneous readings from a slow, high-accuracy analyzer and periodic validations with traceable gas standards.
  • Modular approach to eddy-covariance uncertainty quantification and QA/QC. Temporally and spatially explicit QA/QC approaches are combined to attribute quality properties and uncertainty in time and in space.
  • Rectifying the spatio-temporal representativeness of eddy-covariance measurements. A combination of frequency-domain processing, footprint modeling and machine learning is used to determine the exchange of momentum, heat, water vapor and CO2 over a target area.

While these approaches build on existing methodologies for measurement design and data processing, only cross-network collaborations have enabled their necessary advancement in a timely fashion. After rigorous testing, NEON is documenting the final algorithms and making them publicly available as open-source code.

The National Ecological Observatory Network is a project solely sponsored by the National Science Foundation and managed under cooperative agreement by NEON, Inc. This material is based upon work supported by the National Science Foundation under the grant DBI-0752017. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.

Kenneth Davis (, The Pennsylvania State University

With Chris Duffy, David Eissenstat, Yuning Shi, Susan Brantley, Yuting He, Jason Kaye, Margot Kaye, Henry Lin, Andrew Neal, Xuan Yu, Fuqing Zhang

Model-data synthesis of the carbon and water cycles at very high resolution in complex topography (poster)

The Susquehanna Shale Hills Critical Zone Observatory provides a test bed for studying the carbon and water cycles at very high resolution. The observatory, encompassing the Shaver’s Creek watershed, has focused primarily to date on the Shale Hills catchment, a 200×400 m2 first order watershed. The watershed hosts multi- state observations of the water cycle including evapotranspiration using eddy covariance, soil moisture, groundwater well and sapflux networks, and stream discharge. These data and the Penn State Integrated Hydrologic Model have been used to demonstrate a newly developed, ensemble Kalman-filter based model- data synthesis system. This system is able to provide high-fidelity, high-resolution reanalyses of the catchment water cycle. The observational and modeling systems at Shale Hills are being expanded to incorporate the carbon cycle. Observations include an eddy covariance flux tower, mapping of all trees in the catchment, litterfall and leaf area index measurements, an array of dendrometer bands, and soil carbon and soil carbon dioxide measurements, all of which, save the eddy covariance tower, map spatial variability within the watershed. The watershed clearly shows strong interactions between the carbon and water cycles, with marked differences in soil and vegetation properties on north vs. south facing slopes, and in the valley bottom vs. the slopes, swales and ridge tops. We are embarking on an investigation of the water-carbon cycle interactions in this watershed using observations and models capable of resolving watershed carbon and water cycling down to a spatial resolution of meters and temporal resolution of minutes. This high-resolution system will be applied both to climate impact studies and to evaluating approximations and parameterizations required for coarser-resolution, large-scale modeling of hydrology-ecosystem-atmosphere interactions.

This work has been supported by the National Science Foundation’s Critical Zone Observatory program, the National Oceanographic and Atmospheric Administration’s Hydrologic Sciences program, the Penn State Experimental Forest, and Penn State’s Earth and Environmental Sciences Institute.

Renato Prata de Moraes Frasson (, The Ohio State University, Civil, Environmental and Geodetic Engineering Dept

With Gil Bohrer and Peter S. Curtis

Predicting the impact of disturbances on the carbon cycle of a mixed-deciduous forest in the upper Midwest (poster)

Disturbances, either natural or anthropogenic, impact the carbon and water cycles. Therefore, understanding their immediate effect, as well as how fluxes evolve while forests recover from disturbances, is essential to carbon and water cycle modeling. Our study area is located in northern Michigan and encompasses the mixed- deciduous forest surrounding the University of Michigan Biological Station (UMBS). Two AmeriFlux affiliated towers provide the supporting data for our study. They are operated by the UMBS, and one overlooks an undisturbed footprint (US-UMB) while the second monitors the Forest Accelerated Succession ExperimenT (FASET) site (US-UMd), a 39 ha area where all aspen (Populus spp.) and birch (Betula papyrifera) trees were girdled.

We used the Ecosystem Demography model version 2 (ED2) to run three scenarios: a control (undisturbed) case and two disturbance cases (dist-1, dist-2) where all early successional trees were removed. Dist-1 featured a sudden removal of all early successional trees on 1 January 2010, while under dist-2, we prescribed a gradual removal of the trees, which happened monthly and was spread throughout two years. We parameterized ED2 using observations of monthly and yearly net ecosystem exchange (NEE), and latent and sensible heat fluxes from the undisturbed site (US-UMB). We forced the model using meteorological data recorded by the flux towers and evaluated the output of the three cases against NEE measured at US-UMB (undisturbed case) and against the US-UMd tower (cases dist-1 and dist-2) after the disturbance occurred.

All three scenarios showed a decline in the pine population. Moreover, the undisturbed scenario showed a decline in the mid and late successional deciduous populations, while the early successional deciduous trees kept growing, and NEE did not show any discernible change throughout the 50 years of simulation. Both the dist-1 and dist-2 scenarios showed steady growth of the late successional deciduous trees after the disturbance occurred. While dist-1 showed a slightly slower recovery of NEE, the model predicted that the net ecosystem exchange would return to pre-disturbance levels after approximately 15 years. However, after the transition to a late successional stage took place, ED2 predicted that NEE slowly converges to a neutral state. Our results indicate that although ED2 can predict the recovery of disturbed ecosystems, predicting their rate of recovery is still challenging.

Ellen Goodrich-Stuart (, Virginia Commonwealth University, Department of Biology

With Peter Curtis: Ohio State University, Dept of Evolution, Ecology and Organismal Biology; Robert T. Fahey: The Morton Arboretum, Forest Ecology; Christoph S. Vogel: University of Michigan Biological Station; Christopher M. Gough: Virginia Commonwealth University, Department of Biology

Net primary production resistance across a gradient of moderate disturbance in a temperate forest ecosystem (poster)

The global carbon (C) balance is vulnerable to disturbances that alter the balance between terrestrial C uptake and loss. Moderate disturbances that kill or defoliate only a subset of canopy trees such as insect defoliation, drought, and age-related senescence are increasing in extent and frequency; yet, little is known about the effect of moderate disturbance on forest production and the mechanisms sustaining or supporting the recovery of the C cycle across a range of moderate disturbance severities. We used a broad plot-level gradient of tree mortality within a large manipulation of forest disturbance to: 1) quantify how net primary production (NPP) responds to a range of moderate disturbance severities and; 2) identify the primary mechanisms supporting NPP resistance or resilience following moderate disturbance. We found that NPP was highly resistant to moderate disturbance, exhibiting stability following the senescence of 9 to ~ 60 % of the upper canopy tree basal area. As disturbance severity increased, proportional increases in canopy gap fraction and light availability to the subcanopy enhanced leaf-level C uptake and the growth of this formerly light-limited canopy stratum, helping to compensate upper canopy production losses. At higher levels of disturbance severity and leaf area loss, whole- ecosystem production efficiency increased maintaining NPP. These findings provide a mechanistic explanation for stable NPP across the disturbance gradient, in which the physiological and growth enhancement of undisturbed vegetation was proportional to the level of disturbance severity. Importantly, our results suggest a non-linear response to rising disturbance. We conclude that high functional resistance to moderate disturbance will maintain forest C uptake as disturbances increase in scale and frequency.

This research was supported by the Climate and Environmental Sciences Division, Office of Science, U.S. Department of Energy (DOE) [Award No. DE-SC0006708].

Sara Knox (, University of California, Berkeley

With Cove Sturtevant, Jaclyn Hatala, Laurie Koteen, Joe Verfaillie, and Dennis Baldocchi

Annual balances of CO2 and CH4 from drained agricultural peatlands and restored wetlands in the Sacramento-San Joaquin Delta, California (poster)

Worldwide, drainage of organic soils for agricultural and forestry purposes has resulted in significant soil subsidence, and some of the fastest rates and largest magnitudes of carbon loss attributable to land-use change. The Sacramento-San Joaquin Delta (referred to hereafter as the Delta) in California was drained over a century ago for agriculture and human settlement and has since experienced subsidence rates that are amongst the highest observed in the world. It is well recognized that drained agriculture in the Delta is unsustainable in the long-term, and thus to help reverse subsidence and capture carbon there is a growing interest in converting drained agricultural land-use types to flooded conditions (e.g. restored wetlands and rice agriculture). However, flooding may increase the emission of methane (CH4) as well as the loss of water via evaporation. I used multiple years of simultaneous eddy covariance measurements at conventional drained agricultural sites (a pasture and a corn field) and flooded land-use types (a rice paddy and two restored wetlands) to assess the impact of drained to flooded land-use change on CO2, CH4, and evaporation fluxes in the Delta.

I found that the drained sites were net greenhouse gas (GHG) sources, releasing between 134-350 g-C m-2 yr-1 as CO2 and up to 10 g-C m-2 yr-1 as CH4. Conversely, the restored wetlands were net sinks of atmospheric CO2, sequestering up to 450 g-C m-2 yr-1. However, they were also large sources of CH4, with emissions ranging from 40 to 57 g-C m-2 yr-1. In terms of the full annual GHG budget (assuming that 1 g-CH4 equals 25 g-CO2 with respect to the greenhouse effect over a time horizon of 100 years), the restored wetlands could be either GHG sources or sinks. Annual net CO2 exchange at the rice paddy ranged from -283 g-C m-2 to 95 g-C m-2 depending on management practices, and the site was always a moderate source of CH4. The flooded land-use types evaporated 45-200% more water than the pasture or corn sites. Therefore, from a subsidence perspective, restored wetlands and rice appear to provide a benefit for Delta sustainability as they are predominantly large carbon sinks. However, flooding also has secondary effects on the GHG budget through increased CH4 emissions and higher rates of evaporation.

Asko Noormets (, Department of Forestry and Environmental Resources, North Carolina State University

With Guofang Miao (Department of Forestry and Environmental Resources, North Carolina State University); Jean-Christophe Domec (Department of Forestry and Environmental Resources, North Carolina State University; Bordeaux Sciences Agro, University of Bordeaux); Carl C. Trettin (Center for Forested Wetland Research, USDA Forest Service); Ge Sun (Eastern Forest Environmental Threat Assessment Center, USDA Forest Service); Steve G. McNulty (Eastern Forest Environmental Threat Assessment Center, USDA Forest Service); John S. King (Department of Forestry and Environmental Resources, North Carolina State University)

Partitioning ecosystem respiration in a coastal plain forested wetland in the southeastern USA: Hydrologic effects and implications in climate change (poster)

Wetlands store and process a disproportionately large fraction of global carbon compared to their areal coverage. It is recognized that their historic role as carbon sink may be vulnerable to environmental change and land use pressure. Respiration, a key determinant affecting the role of an ecosystem as carbon source or sink, is less-well investigated in wetlands, and might be made more complex by frequent change in hydrologic regimes than in upland systems. To understand the hydrologic effects on respiration in wetlands and investigate the difference in the responses of respiratory components to change in environmental conditions, we established an eddy covariance flux tower in a coastal plain forested wetland in North Carolina, USA, and conducted measurements on ecosystem respiration (Re), belowground respiration (Rs) and decomposition of coarse wood debris (RCWD) over three years (2009-2011). Aboveground plant respiration (Ragp) was calculated as the residual between Re and Rs, RCWD. All the respiratory terms were separately quantified for flooded and non-flooded periods considering the different mechanisms. Belowground respiration responded rapidly to water level (WL) fluctuation whereas the Ragp did not exhibit significant difference between the two hydrologic regimes. Annual ecosystem respiration was partitioned to 51% from Rs, 12% from RCWD and 37% from Ragp in 2010, with the major contribution of 57% from Rs during non-flooded periods and 69% from Ragp during flooded periods. Overall, this forested wetland respired nearly 2000 g CO2-C m-2 y- 1 annually, comparable to many upland forests. The temperature sensitivity of respiratory components was similar to other ecosystems, but might be affected by WL fluctuation. Hydrologic regimes resulted in unique characteristics of respiration in this forested wetland, transiting between Rs:Re ratio similar to boreal and temperate upland forests under non-flooded conditions and one similar to some tropical rainforests during flooding. A linear increase in Rs:Re ratio with the WL drawdown was observed.

This work has been supported by DOE NICCR (award 08-SC-NICCR-1072), USDA Forest Service Eastern Forest Environmental Threat Assessment Center (award 08-JV-11330147-38). G.M. was partly supported by a graduate research assistantship from the USGS Southeast Climate Science Center (award G10AC00624). Partial support for the study was also provided by DOE-TES program (award 11-DE-SC-0006700) and Ameriflux Management Project (subcontract No. 7090112). The USFWS Alligator River National Wildlife Refuge provided helpful scientific discussions, the forested wetland research site, and in-kind support of field operations.

Timothy Morin (, The Ohio State University

With Gil Bohrer, Renato P.M. Frasson

Optimizing a gapfill model for an urban wetland’s methane fluxes (poster)

Wetlands are the largest source of methane (CH4) worldwide but offer a wide variety of ecosystem services and are commonly constructed in the United States to mitigate wetland loss, particularly in and near urban areas. CH4 emissions were measured at the Olentangy River Wetland Research Park (ORWRP) over three summers and two winters using an eddy flux covariance system. In this study, we used linear and neural network modeling with the Akaike Information Criteria to arrive at a general empirical model for methane emissions from the ORWRP. To account for the small-scale landscape heterogeneity, a typical characteristic of urban wetlands, we incorporated the patch-type composition within the flux footprint as part of the modeling process. Our methodology identified LE, the footprint, and soil temperature consistently as significant variables in the modeling of the observed CH4 fluxes. The predictive ability of these variables suggests that CH4 emissions are largely related to the volatization of water and also to the fraction of the flux signal originating from within the permanently flooded section of the wetland park. The validity of using flux variables in order to gap fill methane flux observations was also tested by evaluating the impact of different groups of explanatory variables on the uncertainty of the gap filled CH4 flux totals at the semi-hourly scale for different seasonal and diurnal combinations Our study concluded that fluxes which occurred within the first few additions from a stepwise linear model added enough explanatory power to the model to offset the added error from their own gap filling process.

Quality control inspection of observation and data processing from the site was conducted by Ameriflux in Aug-2011. The study was funded by NSF grant #CBET-1033451, US Geological Survey grant #G11AP20099 through the Ohio Water Resources Center projects 2011OH205B, 2012OH259B and Ohio Water Development Authority research grant #6560. Flux data processing was supported in part by Ameriflux Core Flux Site Agreement.

J William Munger (, Harvard University, School of Engineering and Applied Sciences; Harvard University, Dept. of Earth and Planetary Sciences

With Rosisin Commane; David Orwig (Harvard University, Harvard Forest); David Foster (Harvard University, Harvard Forest)

Recent observations from the Harvard Forest flux towers (poster)

Forests regrowing on abandoned agricultural landscape in the northeastern U.S. are significant carbon sinks. Understanding how this will respond to climate change, disturbance, and natural succession is an important question for future carbon management. Two long-term flux towers in combination with extensive plot-based sampling of forest biomass at the Harvard Forest are contributing to our understanding of forest carbon dynamics at multi-decadal time scales. The Harvard Forest Environmental Measurement Site (EMS) tower, established in 1990, represents deciduous-dominated mixed forest. The nearby Hemlock tower, established in 2000, represents hemlock dominated forest. Plot arrays for annual biomass measurements center on the towers and both are encompassed by a 35 ha vegetation dynamics plot that is surveyed completely on ~5 yr cycle.

Annual sums of carbon flux and biomass increments agree that both the deciduous and hemlock stands are accumulating carbon. The deciduous stand sustains a very high peak rate of photosynthesis for a short season spanning June-August, while the Hemlock stand has relatively constant rates of maximum daily uptake over a longer season starting in April and continuing to November. Uptake in the deciduous stand is considerably less in spring and fall months and quite variable from year to year. Some of the carbon uptake at EMS stand must be due to conifers in the subcanopy that can be active before the deciduous canopy leaves emerge. The hemlock site is poised to document changes in carbon, water and energy balances as a hemlock wooly adelgid infestation selectively kills hemlocks.

The EMS site supports new instrument development and testing such as a laser absorption spectrometer to measure carbonyl sulfide (OCS) fluxes. Because OCS is absorbed by foliage and reacts with carbonic anhydrase it is of interest as a potential tracer for photosynthetic uptake of CO2. Some key findings are that forests have previously unrecognized sources of OCS that are especially noticeable in hot, dry conditions. Secondly, the relationship between OCS and CO2 uptake is somewhat non-linear. In high light conditions when CO2 uptake is rapid and limited by mass transfer through the stomata it scales consistently with OCS. However, at low light levels and outside the peak growing season, CO2 uptake is limited by light or availability of Rubisco rather than stomatal conductance, and the relationship to OCS exchange is shifted. OCS may still be a useful proxy for photosynthetic CO2 exchange, but will require more detailed analysis than a simple linear scaling.

The Harvard Forest flux towers are supported by the Office of Science (BER), U.S. Department of Energy and are a component of the Harvard Forest LTER supported by the National Science Foundation

Jessica Osuna (, LLNL

With Mihail Bora, Tiziana Bond, Sonia Wharton

Progress toward developing a novel mini tunable diode laser for measuring CO2 fluxes in situ (poster)

We introduce a novel mini tunable diode laser (TDL) that is being developed at Lawrence Livermore National Laboratory (LLNL). This sensor has the potential to measure a multitude of trace gases and their isotopologues by choosing a laser tuned to the appropriate wavelengths. Additionally, multiple sample cells could be deployed as an array for improved spatial sampling of fluxes and isotopologues using a single laser source. This decreases the cost and maintenance normally associated with increasing spatial sampling at a site. The sensor is compact and requires low amounts of power through the use of data acquisition units (DAQs) meaning it is ideal for remote in situ deployment or airborne sampling.

We present preliminary tests of the LLNL mini-TDL using a vertical cavity surface-emitting laser (VCSEL) tuned to a narrow spectral range targeting CO2 absorption lines at 2012 nm. The sample cell is a multi-pass Fraunhofer White cell measuring 5cm x 4cm x 1cm with a total path length of 2.5m. The VCSEL was calibrated in the lab using a Bristol wavelength meter to match CO2 absorption peaks against the HITRAN 2012 database. We characterize sensor performance in the lab at a range of [CO2] from 0 ppm to 1000 ppm. Additionally, we show sensor stability under a range of temperature and humidity in an environmental chamber.

This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. LLNL-ABS-648409

Jessica Osuna (, LLNL

With R.D. Pyles (UC-Davis), M. Falk (UC-Davis), S. Ma (UC-Berkeley), D. Baldocchi (UC-Berkeley), K. Bible (U-Washington), S. Biraud (LBNL), S. Wharton (LLNL)

Evaluating the ACASA model as a tool for flux interpretation at Four AmeriFlux Sites (poster)

Many eddy-covariance flux sites are equipped with one measurement height above the canopy. Due to power requirements, instrument cost, and the maintenance involved it is often unrealistic to expand measurements across the full profile of the canopy. However, an understanding of the turbulence dynamics throughout the canopy could be valuable in interpreting flux data at measurements height. We apply a higher-order closure turbulence and flux model to four distinct flux sites in order to model fluxes and profiles of turbulence in and above canopies. We used the UC-Davis Advanced Canopy-Atmosphere-Soil Algorithm (ACASA) to study a range of ecosystems including a tall old-growth forest with a complex canopy (Wind River Ameriflux), a two-layered oak savanna (Tonzi Ameriflux), a grassland (Diablo Ameriflux), and a wheat field (Southern Great Plains Ameriflux). We modeled each ecosystem for at least the duration of the growing season. At each site, the model was reinitialized twice per month using data of soil moisture and temperature. We assess the conditions under which applying a complex model like ACASA is beneficial in interpreting within-canopy dynamics and flux signal quality. We describe ACASA performance at the four sites as well as the changes to the model and the tuning necessary to optimize performance for each of the unique sites. Novel improvements to the model include the incorporation of seasonal dynamics in leaf area index and photosynthetic capacity, adjustment of plant water use during seasonal drought, and adjustments for a clumped canopy

This work performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344. LLNL-ABS-648474

Dario Papale (, ICOS Ecosystem Thematic Center & DIBAF University of Tuscia . Viterbo (Italy)

Connecting ecosystem carbon observatories in Europe and USA – the COOPEUS project and the ICOS infrastructure (poster)

The COOPEUS project (, funded under the 7th Framework Programme for Research and Innovation of the EU, has the aim to bring together scientists and users being involved in Europe’s major environmental related research infrastructure projects with their US counterparts, starting from the one funded by NSF.

The intention is that by interlinking these activities new synergies are generated that will stimulate the creation of a truly global integration of existing infrastructures. The keys of this integration process will be the efficient access to and the open sharing of data and information produced by the environmental research infrastructures and the harmonization and standardization of methods, protocols and data products.

In Europe the Integrated Carbon Observation System (ICOS at infrastructure is going to monitor atmospheric, ecosystem and ocean carbon and GHGs concentrations and fluxes across the continent, under the coordination of Thematic Centers that will ensure standard methodologies and processing. The Ecosystem Thematic Center, coordinated by Italy with a contribution from Belgium and France, is developing standard protocols to acquire and process data from more than 60 sites in Europe all equipped with eddy covariance systems to measure atmosphere-ecosystem exchanges of CO2, CH4, N2O, water and energy.

In this initial phase of the infrastructure where the methods and products are defined, it is crucial to interact, discuss and harmonize the activities between ICOS and similar initiatives in USA. This is ongoing with both the National Ecological Observatory Network (NEON) and the AmeriFlux networks with exchange of information, common initiatives, protocols comparison and tools development. Increasing the level of harmonization and comparability will simplify global use of our data and will be an example also to the others emerging networks worldwide.

In this poster the ICOS ecosystem infrastructure and the ongoing coordination activities and exchanges with the NEON and Ameriflux initiatives will be presented with a special emphasis on the next future COOPEUS actions, with the aim increase the participation to the discussion level in the communities.

Daniel Ricciuto (, ORNL

With Jiafu Mao; Xiaoying Shi; Peter Thornton; NACP interim synthesis team

Evaluation of the Community Land Model with AmeriFlux data (poster)

Earth System modeling has advanced rapidly over the past several years both in mechanistic complexity and spatial resolution. Most land-surface components of these models now include detailed carbon cycle models, nutrient cycling and below-ground biogeochemistry. These advances have allowed such models to be used not only as tools to understand large-scale spatial patterns, but also to simulate and integrate site-level carbon cycle observations and experiments. The North American Carbon Program (NACP) site interim synthesis provided a consistent framework and driver data for models of varying complexity, including several Earth system models, to compare model output against eddy covariance tower observations. The Community Earth System Model (CESM) will continue to play a crucial role in model-data integration. Here we evaluate multiple versions of the Community Land Model (CLM4.0 and CLM4.5), the land component of CESM, with datasets from 14 different flux tower sites. The single-point implementation of CLM (PTCLM) is used for this analysis. Initial focus is on evaluation of gap-filed net ecosystem exchange (NEE) and gross primary productivity (GPP). Additionally, we evaluate the impact of using site-level vs. reanalysis driver meteorology and site-specific vs. global parameterization on these simulations. Although these simulations consider past land use change, CO2 fertilization and nitrogen deposition, the model consistently underestimates the observed carbon sink. We find that the more recent, mechanistically more complex version of CLM4.5 performs better at most sites. For both NEE and GPP, we find that the differences in driver datasets and model parameters cause at least as much variability in model predictions as the model structure.

This research is sponsored by the U.S. Department of Energy, Office of Science, Biological and Environmental Research (BER) programs, and performed at Oak Ridge National Laboratory (ORNL). ORNL is managed by UT- Battelle, LLC, for the U.S. Department of Energy under contract DE-AC05-00OR22725. Please contact for further information.

Crystal Schaaf (, University of Massachusetts Boston

With Ian Paynter (UMB), Edward Saenz (UMB), Francesco Peri (UMB), Xiaoyuan Yang (UMB), Yan Liu (UMB), Angela Erb (UMB), Zhan Li (Boston University), Alan Strahler (BU), Zhuosen Wang (NASA/GSFC), Bruce Cook (NASA/ GSFC), Keith Krause (NEON), Nathan Leisso (NEON)

Canopy Structure from Airborne and Terrestrial Lidar at Harvard Forest (poster)

Airborne and Terrestrial Scanning Lidar data are available from several field campaigns over Harvard Forest allowing comparison of canopy height models (CHM) and forest structure characteristics. In June 2012, the National Aeronautics and Space Administration (NASA) also flew Goddard’s LiDAR, Hyperspectral, and Thermal (G-LiHT) instrument package over Harvard Forest. Using the discrete 1550nm Lidar (Riegl), a 1m resolution CHM was derived. In August 2012, the National Ecological Observatory Network (NEON) flew their Airborne Observation Platform (AOP) with a full waveform 1064nm Lidar (Optech). Discrete returns from this instrument have recently been used to develop both 1m and 0.5m resolution CHM. In July 2007 and 2009, the full waveform 1064nm Echidna Validation Instrument (EVI) terrestrial Lidar was deployed in Harvard Forest and the resulting point clouds were merged to produce a 0.25 hectare sized 3D canopy reconstruction of a hardwood site in the vicinity of both the Ameriflux EMS tower and the new NEON tower. During the 2012 NEON campaign and again during June of 2013, the next generation Dual Wavelength Echidna Lidar (DWEL) was deployed at the same location. Additional discrete terrestrial Lidar scans of the hardwood site were taken in September and December of 2013 with the 905nm ultraportable rapid-scanning Canopy Biomass Lidar (CBL). These various measures of forest structure will be compared with the extensive field measures of standard forestry parameters (height, diameter breast height, species, crown characteristics, leaf area index) that were also collected during the 2007, 2009, 2012, and 2013 campaigns.

Cynthia M. Scheuermann (, Virginia Commonwealth University, Department of Biology

With Christopher M. Gough, Peter S. Curtis (Ohio State University, Department of Evolution, Ecology, and Organismal Biology), Brady S. Hardiman (Boston University, Department of Earth and Environment)

Carbon storage over stand development in North America’s temperate deciduous forests (poster)

The broad emergence of century-old forests in the US upper Midwest and Northeast has large potential implications for carbon (C) storage, as the long-assumed future decline of production in aging stands is expected to reduce continental C sink strength. As mixed temperate forests in the region broadly transition from early to middle stages of succession, short-lived canopy trees are dying and giving way to more structurally complex forests. At the same time, disturbances in the region are shifting from severe events to more moderate disturbances resulting in only partial canopy defoliation.

We review new evidence that temperate deciduous forests, against prior expectations, are likely to maintain their capacity to store C over the next several decades as diffuse mortality of canopy trees increases with advancing age. Forest production data from long-term observational studies and experimental chronosequences in the region do not support a decline in C storage during middle succession. Instead, sustained forest C storage in intermediate-aged forests corresponds with the accrual of structural complexity over time from small-scale, non- stand replacing disturbances such as age-related mortality. Increasing complexity with age gives rise to more efficient use of growth-limiting light and nutrient resources, which offsets the lost contribution of senescent canopy trees to forest production.

These findings show that older, structurally complex forests that emerge following diffuse mortality store C at rates comparable to or greater than their younger counterparts. We conclude that regional land-use decisions permitting age-related senescence in maturing forests or, alternatively, management outcomes emulating the structural features of older forests will support goals to maintain the region’s C sink strength.

Joel Biederman (Joel.Biederman@ARS.USDA.GOV), USDA-ARS Southwest Watershed Research Center

With Russell Scott1, Michael Goulden6, Marcy Litvak3, David Bowling2, Thomas Kolb5, Peter Blanken7, Enrico Yepez8, Enrique Vivoni10, Julio Rodriguez9, Christopher Watts9, Jaime Garatuza8, Walter Oeschel4, Sabina Dore5, Thomas Bell4, and Sean Burns7. 1USDA-ARS Southwest Watershed Research Center, 2University of Utah, 3University of New Mexico, 4San Diego State University, 5Northern Arizona University, 6University of California, Irvine, 7University of Colorado, Boulder, 8Instituto Tecnologico de Sonora, 9Universidad de Sonora, 10Arizona State University

Maturing flux datasets reveal ecosystem carbon uptake sensitivity to temporal climate variability across a summer-rainfall gradient in the Southwest (poster)

Quantification of global carbon budgets has suffered from scant observations in arid and semi-arid regions, which comprise around 40% of the terrestrial land surface. Dryland regions are often characterized by large temporal variability in precipitation and other drivers of carbon exchange, but common proxies of carbon exchange may fail to capture the quickly changing dynamics of these water-limited ecosystems. While remote sensing proxies (e.g. NDVI, fPAR) can characterize the spatial variability of productivity across climate gradients, they may be less sensitive to temporal dynamics. Likewise, traditional field measurements (e.g., aboveground primary production) do not capture belowground changes, and such periodic metrics may be poor in drylands, where productivity occurs erratically in response to precipitation pulses.

We use maturing, multiyear eddy covariance datasets to quantify temporal sensitivity of annual gross ecosystem production (GEP) to annual precipitation (P) across 15 sites receiving significant summer P in the southwest United States and northwest Mexico. Mean annual P varies from 150-900 mm with site-level annual variation of 25 to 50%, while corresponding vegetation ranges from sparse Sonoran desert scrub to mixed-conifer forest with persistent seasonal snowpack. GEP was significantly more sensitive to annual P than was suggested by a remotely sensed carbon uptake estimate (MODIS-17). Initial analysis suggests that annual GEP sensitivity to P may peak at mid-productivity sites such as a mesquite savanna and pinyon-juniper woodland, where perhaps the more diverse plant communities allow ecosystems to more efficiently utilize P pulses. In ongoing work, we use the measured evapotranspiration to ask how hydrologic partitioning regulates the temporal sensitivity of carbon uptake to precipitation. This work illustrates that maturing EC datasets are critical to understanding temporal variability of terrestrial carbon balance.

Jim Tang (, MBL Ecosystems Center

Physiological controls of stem respiration and soil respiration in a temperate forest (Poster)

Stem respiration from forest ecosystem is an important component of total ecosystem respiration and the forest carbon cycle. Our knowledge in understanding the variation in stem respiration and its governing drivers is limited, partially because empirical measurement of stem respiration is scarce. It has been reported that soil respiration is partially controlled by photosynthesis, but how stem respiration and soil respiration are controlled differently by photosynthesis over the diel scale is unknown.

The objectives of this research are to reveal the diel and seasonal pattern of stem respiration and soil respiration, the connection between stem and soil respiration, and how photosynthesis controls respiration through transport of photosynthate via phloem. The results will significantly improve our ability to model respiration and incorporate the process into earth system modeling.

We developed a novel system to automatically measure stem respiration at a half-hour frequency and to explore the diel pattern and its correlation with soil respiration and root respiration. We hypothesize that the peak value of stem respiration during a day reaches earlier than root respiration, resulting from the transit transport of newly assimilated photosynthate.

We found that the magnitude of stem-area-scaled stem respiration was at the same order as ground-based soil respiration. The diel pattern of stem respiration was primarily driven by temperature variation. But the peak stem respiration during the course of a day was controlled by tree photosynthesis. The peak value of stem respiration during a day reached earlier than root respiration. The CO2 source of stem respiration was primarily from locally produced stem metabolism, not from xylem water transported from roots.

This project is supported by DOE TES (DE-SC0006951) and NSF (AGS-1005663).

Bai Yang (, Oak Ridge National Laboratory

With Tom Boden, Misha Krassovski

AmeriFlux Data Management at CDIAC, ORNL (Poster)

The Carbon Dioxide Information Analysis Center (CDIAC) at the Oak Ridge National Laboratory serves as the long-term data repository for the AmeriFlux network. Datasets currently available include hourly or half-hourly meteorological and flux observations, biological measurement records, and synthesis data products.

In this presentation, we provide an update of this network database including a comprehensive review and evaluation of the biological data from about 70 sites and data support to two synthesis studies—2012 drought synthesis and FACE synthesis. Issues related to data quality and solutions in compiling datasets for these synthesis studies will be discussed. We will also discuss the development of a new product for flux uncertainty estimates, and re-formatting of Level-2 standard files. Work plans will be presented such as developing and producing other high-level products.

Margaret Torn (, LBNL

The AmeriFlux Management Project: Overview (Poster)

With Deb Agarwal and the Data Team; Sebastien Biraud and the Tech QA/QC team, Marilyn Saarni and the Outreach Team, and Dennis Baldocchi

AmeriFlux is a network of more than 100 sites using Eddy Covariance towers to measure ecosystem CO2, water, and energy fluxes across the Americas ( . The DOE AmeriFlux Management Project serves a broad community of flux sites and data users. This poster will present some new resources for the Network and highlights of the AmeriFlux Management Project.

Margaret Torn (, LBNL

AmeriFlux Data Collection, Processing, and Data User Support (Poster)

With Gilberto Pastorello LBNL, Deb Agarwal LBNL, Dario Papale (U Tuscia), Cristina Poindexter LBNL, Boris Faybishenko LBNL, Tom Boden ORNL, Rachel Hollowgrass LBNL

The AmeriFlux Data Management team is providing new and expanded services. These services include: archiving of high frequency data from sites; high frequency data processing; enhanced quality assessment capabilities; advanced data processing capabilities; improved biological, ancillary, disturbance, and metadata collection methods; and expanded user services available via the AmeriFlux web site. This new range of services and products from the Data Management team are the result of a close collaboration between the AmeriFlux Management Project, ICOS, and CDIAC personnel. This poster will provide an overview of the components of the system and a preview of upcoming functionality.

The Funding for the AmeriFlux Management Project was provided by the U.S. Department of Energy’s Office of Science under Contract No. DE-AC02-05CH11231.

Jonathan P Dandois (, UMBC

High spatial resolution three-dimensional mapping of vegetation spectral dynamics using computer vision and hobbyist unmanned aerial vehicles (Poster)

With Erle C. Ellis, UMBC Department of Geography and Environmental Systems.

Sébastien Biraud (, LBNL

AmeriFlux Management Program QA/QC Technical Team (Poster)

With Stephen Chan1, Chad Hanson2, David Billesbach2, and Margaret Torn. 1: Lawrence Berkeley National Laboratory; 2: Oregon State University; 3: University of Nebraska, Lincoln

The goal of AmeriFlux is to develop a network of long-term CO2 flux sites for quantifying and understanding the role of the terrestrial biosphere in global climate change. The network currently includes more than 100 sites started by many scientists and supported by multiple agencies (DOE, NSF, USDA, NFS). The AmeriFlux Management Program (AMP) Technical QA/QC at LBNL strengthens the entire Ameriflux Network and Core Sites by: (1) standardizing operational practices, calibration, and maintenance routines; (2) setting clear data quality goals, and (3) helping resolve instrument failure promptly. To ensure inter-comparability in the network, we conduct site comparisons with portable eddy covariance (PEC) systems, provide calibration gas standards and lab-quality sensors to check instrument performance, and identify uncertainties associated with data processing using data diagnostics and gold-standard files. During the past year, the LBNL AmeriFlux QA/QC lab transitioned from the AmeriFlux QA/QC lab at Oregon State University (OSU). In particular we built two new PEC systems. Our team is experienced in testing new instruments and working with manufacturers, and is building relationships with both vendors and investigators. We have completed eight site visits in 2013, and are planning eight site visits in 2014.

Christopher Gough (, Virginia Commonwealth University

Carbon and water cycling following low-severity disturbance in an Upper Great Lakes forest: Empirical and modeling results from an AmeriFlux core site (p​​oster)​​

With Project Director/PI: Dr. Peter S. Curtis, The Ohio State University. Co-investigators: Dr. Gil Bohrer, The Ohio State University; Dr. Knute Nadelhoffer, The University of Michigan; Dr. Christopher M. Gough, Virginia Commonwealth University. Senior Personnel: Dr. Christoph Vogel, The University of Michigan.

Disturbance modifies forest physical structure and composition, which may in turn alter biogeochemical processes central to carbon (C) and water cycling. In the Upper Great Lakes region, disturbances are transitioning away from severe events that cause forest stand replacement, and toward lower-intensity disturbances that result in partial canopy defoliation or loss of selected species. These low-severity disturbances include partial harvests, wind, pathogenic insects, diseases, and age-related senescence. The response and mechanisms supporting the recovery of C and water cycles following low-intensity disturbance are poorly characterized, with most prior empirical and modeling studies examining the biogeochemical implications of severe stand- replacing disturbance.

At the University of Michigan Biological Station Ameriflux core site, we are using long-term records of C and water cycling from unmanipulated control (US-UMB) and experimentally disturbed (US-UMd) forests to quantify disturbance-related changes in biogeochemical cycling and identify the underlying mechanisms supporting their resilience to disturbance. The Forest Accelerated Succession Experiment (FASET), in which >6,700 canopy dominant Populus (aspen) and Betula (birch) trees were stem girdled within a 39 ha area, employs C and water cycling measurements within paired US-UMB and US-UMd meteorological flux tower footprints.

The C cycle at our site has exhibited striking resilience to low-intensity disturbance, with sustained net ecosystem production (NEP) following the senescence of a third of canopy trees. Disturbance-related changes in canopy physical structure are associated with improved resource-use efficiency, providing a mechanism for sustained NEP as trees die. We found that the water cycle responded to low-severity disturbance through modified transpiration. Transpiration per tree and per sap-wood area increased following disturbance, with responses varying among hydrological functional types. Maples (a diffuse porous species) exhibited greater water stress and a decrease in transpiration because hydraulic stress increased following disturbance.

We find that ecosystem models poorly simulate biogeochemical responses to low-severity disturbance. In partnership with the DOE Pacific Northwest National Laboratory, we examined whether big-leaf (Biome-BGC) and gap (ED and Zelig) models accurately simulate NEP observed in our experimentally disturbed US-UMd site. Both ED and Zelig gap models substantially over-estimated the response of NEP to low-severity disturbance, predicting large declines in C uptake. The big-leaf Biome-BGC model more accurately simulated NEP following disturbance, but its mechanistic basis for resilience was not in agreement with observation. These results have implications for biogeochemical modeling at the global scale; both Biome-BGC and ED are part of the development stream of NCAR’s Community Land Model and DOE’s forthcoming ACME.

DOE TES Award No. DE-SC0006708

Kim Novick (, Indiana University​

The role of isohydric and anisohydric species in determining ecosystem-scale response to severe drought (p​oster)

With Roman, D.T., Brzostek, E.R., Dragnoi, D., Rahman, F., and Phillips, R.

There is a critical need to develop mechanistic frameworks for the species-specific contributions to stand-scale forest drought response in heterogeneous landscapes. Here, we develop such a framework that assumes plants fall along a continuum of isohydric to anisohydric regulation of leaf water potential. We use the framework to interpret a three-year record of weekly measurements of leaf-level gas exchange and leaf water potential, and continuous measurements of the net ecosystem exchange of CO2 (NEE) and gross ecosystem productivity (GEP) in a deciduous hardwood forest.A severe drought reduced the absolute magnitude of NEE and GEP by 55% and 40%, respectively, during the peak of the drought period, but species-specific drought responses varied. Oak species were characterized by anisohydric regulation of leaf water potential that promoted static gas exchange throughout the study period. In contrast, leaf water potential of the other dominant species was more isohydric, which limited gas exchange during the drought. Thus, the stand-scale drought sensitivity of GEP and NEE depends on the relative fraction of isohydric versus anisohydric species in a stand, and ongoing shifts in species composition in eastern U.S. forests may exacerbate the deleterious effects of water limitation on ecosystem-scale carbon cycling.

Roser Matamala, Argonne National Laboratory (

Differences in Biomass Distribution and Production Budgets between Cropland and Prairie Grassland ​(poster)

With Zhaosheng Fan and David Cook, Argonne National Laboratory, Argonne, IL, USA.

We study the mechanisms of ecosystem C uptake and retention at two sites with common soil types and climate but differing in management practices. Carbon uptake is measured by the eddy covariance technique. Aboveground and belowground plant biomass, soil microbial biomass, and soil organic carbon (SOC) dynamics are measured via nearby chronosequence on the same soils.

At our study sites (a corn/soybean rotation and a 25-year-old restored prairie near the agricultural site) gross primary production is very similar, averaging about 2500 g C m-2 y-1. Nevertheless, net ecosystem production (NEP) differs greatly due to the different management practices. The prairie site is a strong and stable carbon sink with an average NEP of -307 g C m-2 y-1 over 9 years. This NEP sustains net increases in biomass and production of belowground ecosystem components, including SOC stocks and the standing crop of roots and microbial biomass. For the prairie, we have established that, on a yearly basis, about 10-14% of NEP is allocated to SOC accrual, 7% of NEP supports net root production, and 2% of NEP supports net growth increments of microbial biomass. Thus, 23% of the year’s NEP sustains net gains of root and microbial biomass as well as SOC accrual rates averaging 43 g C m-2 y-1. The remaining 77% of NEP is allocated to production of aboveground litter (which is burned as a management practice to maintain the prairie ecosystem and does not contribute significantly to SOC accrual rates). The cultivated site is rotated annually between corn and soybean, which is a common regional practice. During years when the land is cultivated with corn, NEP averages -270 g C m-2 y-1. Thus, a proportion of NEP might potentially sustain accrual of SOC and soil microbial production. However, when the field is cultivated with soybean, the land becomes a source of CO2 (average NEP is 181 g C m-2 y-1), mostly because the short period of C uptake associated with soybean phenology results in large ecosystem carbon losses. Overall, the positive effects to the production budget that could occur under corn are cancelled by large losses during soybean years, resulting in a lack of energy to sustain SOC accrual under corn/soybean rotations in this area.

Overall, we found that phenological effects on the period of C uptake can be a key factor contributing to sustained ecosystem carbon gains as well as SOC accrual.