B23A-0921 1340h
Remote Measurements of Atmospheric Carbon Column Densities: Error Analysis and Early Results
A global network for tropospheric carbon column density observations is currently being developed. Fourier Transform Spectrometers (FTS) operating in near-infrared solar absorption will measure column densities of CO$_{2}$, CH$_{4}$, and CO simultaneously. Such observations provide atmospheric sampling complementary to in-situ measurements, and will greatly reduce uncertainties in the inference of surface-atmosphere carbon flux. Here we present a preliminary analysis of errors expected in the absorption spectra, and how they affect the column densities. Algorithms using both fixed and variable altitude profile shapes are considered ("profile scaling" and "profile retrieval", respectively). The types of error studied include both fixed and variable forward model errors such as molecular line broadening and the atmospheric temperature profile, as well as noise and smoothing error. We also present preliminary results of CO$_{2}$, CH$_{4}$, and CO column measurements made at Lauder, New Zealand, and compare their variability to that predicted by the error analysis.
B23A-0922 1340h
Ultra-Precise Ground-based Column CO2 Measurements Using a Fabry-Perot Interferometer
High precision measurements of CO2 on a worldwide scale are needed to reduce uncertainties in the global carbon budget. It has been calculated that total column measurements with a precision better than 0.3% are required for this purpose. No measurement from space has ever achieved this level of precision. We present an application of a Fabry-Perot interferometer that detects spectral absorption of sunlight by atmospheric CO2 at 1571 nm. This instrument could be used to provide ground truth information for the OCO instrument, once it is implemented. A space borne version of this instrument could potentially provide more comprehensive long-term validation. Simultaneous measurements of O2 at 762 nm provide surface pressure to permit corrections for atmospheric pressure (or terrain from an airborne or space borne instrument). For measurements of each species in this differential absorption technique, transmittance fringes from a Fabry-Perot interferometer are aligned with absorption lines to maximize sensitivity to changes in absorption. These Fabry-Perot channels are subsequently compared with reference channels monitoring changes in solar flux in the wavelength region of interest. Ground based measurements presented here meet the target precision of \<0.3% with a current detection limit of better than 1 ppm in less than ten seconds averaging.
B23A-0923 1340h
A CO$_2$ Differential Absorption LIDAR System
We are developing a CO$_2$ Differential Absorption LIDAR (DIAL) profiler at the NASA Goddard Space Flight Center (GSFC). This instrument is designed to provide continuous, range resolved, measurements of trace gas concentrations through the depth of the planetary boundary layer (PBL), providing the frequency and vertical resolution needed in order to capture diurnal variations in CO$_2$ and determine the relationship between the CO$_2$ diurnal cycles and the turbulent mixing occurring in the PBL. Three dimensional CO$_2$ distributions from a high resolution chemical transport model along with a line-by-line model are used to generate expected signal returns which are then used to quantify the measurement requirements of the DIAL technique. We present the results of this modeling study including error sources and the accuracy, precision, and sampling frequency necessary to resolve CO$_2$ variability.
B23A-0924 1340h
Novel Instrumentation for Atmospheric Measurements of Carbon Dioxide and Stable Isotopes of Carbon Dioxide
We present the development and testing of new compact, rugged and inexpensive instruments for measurements of carbon dioxide and stable isotopes of carbon dioxide in the atmosphere. These instruments, based on a new technology called Off-Axis Integrated Cavity Output Spectroscopy (Off-Axis ICOS), report measurements with high sensitivity, accuracy, and specificity in real time (no cross interferences). These instruments combine inexpensive, robust, and reliable room-temperature diode lasers operating in the near-infrared spectral region with Off-Axis ICOS, a patented and proven technology that provides extremely long optical paths (several thousand meters typical) to yield an instrument platform with important measurement capabilities. The first instrument reports measurements of carbon dioxide in air with a total uncertainty of less than 1 part in 3000. The second instrument reports measurements of the ratio of the isotopic abundances of 13CO2 and 12CO2 in ambient air with a precision of less than 0.2 del per mil. We will discuss the measurement strategies in detail and present recent results demonstrating real-time measurements without the need for any user intervention. Ongoing efforts to demonstrate the instrument's capabilities to record measurements with high accuracy and precision without calibration over extended periods as well as testing of the instrument at field sites in the AmeriFlux and FLUXNET networks, and NOAA/CMDL will also be discussed. By significantly increasing the accuracy and precision of carbon dioxide measurements in the field, these new instruments will significantly enhance studies of global warming and facilitate controlled multi-year studies and comparisons between geographically distant field sites. These studies will enable detailed monitoring of mass and energy exchange which is needed to determine the chemical composition of the atmosphere and the productivity of the biosphere. The instruments will thus help quantify in detail the global carbon cycle on local and large spatial scales and enable atmospheric chemists to generate reliable models of climate change.
http://www.LGRinc.com
B23A-0925 1340h
A tall-tower study of carbon exchange from developed land use in the U.S. Upper Midwest
A central goal of the North American Carbon Program is to reconcile differences in regional carbon exchange predictions from atmospheric and ground-based approaches. Within any large region there are embedded areas where land cover and land use have been strongly modified by development. However, studies involving continuous measurement of CO$_{2}$ exchange in developed land have begun only recently. These areas represent a significant percentage of the Upper Midwest region. U.S. Census housing density data show that while high-density urban areas occupy 3% of the land area, low-density residential land uses (generally suburbs) occupy $<$11% of the land area. We established a new tall-tower site to study CO$_{2}$, water vapor, and energy exchange from developed land use in the in the Minneapolis-Saint Paul metropolitan area. The objectives of the study are: 1. To quantify the components of the net ecosystem exchange of CO$_{2}$ (NEE) of developed land use and their variation over diurnal, synoptic, seasonal, and interannual time scales. 2. To determine the specific ecosystem controls on the net exchanges of CO$_{2}$, water vapor, and energy by analyzing the relationships between the fluxes and a suite of biophysical properties measured within the footprint of the flux tower. 3. To evaluate the relationship of the MODIS vegetation index, land surface temperature, and snow cover products to temporal dynamics of net ecosystem exchange of CO$_{2}$ and water vapor at a range scales from the flux tower to the metropolitan area. The design, implementation, and first results of the site will be presented. The long-term goal of this research is to understand how developed land use affects regional- to continental-scale carbon dynamics, how these effects may change with urban growth and development, and ultimately, how they could be managed to mitigate carbon sources and maximize sinks.
B23A-0926 1340h
Boundary-layer measurements of CO$_{2}$ concentration, carbon and oxygen isotopes of atmospheric CO$_{2}$ over montane forest regions in Colorado, USA
Air samples were collected with 100ml flasks in the atmospheric and canopy boundary layers using aircraft and ground-based samplers as an integral part of the Airborne Carbon in the Mountains Experiment (ACME) in May and July 2004. A total number of 524 flasks were collected during the two study periods for the analyses of concentration, carbon (\delta$^{13}$C) and oxygen (\delta$^{18}$O) isotopes of atmospheric CO$_{2}$. Air samples were collected in the early morning (beginning ~ 7:00 am) using a C-130 airplane to characterize regional-scale isotopic signals of respiration (\delta$^{13}$C$_{R}$) in the residual boundary layer. As the ground heated after sunrise and the convective boundary layer (CBL) increased, flasks were collected again to estimate isotope ratios of net CO$_{2}$ exchange between the atmosphere and the biosphere. In addition, vertical profiles of atmospheric CO$_{2}$, \delta$^{13}$C and \delta$^{18}$O in the CBL were measured with the aircraft. Two automated sampling systems conducted the ground-based sampling. The first sampled at 15-min intervals the night before a flight (9:00 pm to 12:30 am). These measurements were conducted within forest canopies and were used to capture \delta$^{13}$C ratios of nighttime respiration at the ecosystem scale. The second was programmed to sample every 45 minutes and started approximately two hours before the airborne flights (4:30 am to 3:00 pm). These measurements overlapped aircraft observations, providing information about daytime CO$_{2}$ exchange in the atmosphere-forest interface. Preliminary results showed that a CO$_{2}$ gradient of 27.9 ppm with a corresponding \delta$^{13}$C difference of 1.16 \permil was captured in the residual boundary layer on a July morning. Using a two-source mixing line approach, the regional-scale \delta$^{13}$CR was estimated to be -24.5 \pm 0.4 (S.E.) \permil, which was significantly (p < 0.05) more positive compared to measured nighttime /delta$^{13}$C$_{R}$ values within the forest canopy (-25.7 \pm 0.1 \permil). This difference partially reflects the disparity in sampling footprints represented by the two different methods for measuring CO$_{2}$. Similar results were found on several days, which could perhaps contribute to a better understanding of the CO$_{2}$ flux exchanged between montane forests and the atmosphere.
B23A-0927 1340h
Uncertainty Analysis of CO$_{2}$ Standard Transfer via NDIR Analyzers in the NOAA Climate Monitoring and Diagnostics Laboratory from 1979 to 2004
At present, state of art, non-dispersive infrared (NDIR) analyzers offer the most precise method of CO$_{2}$ detection. However, this technique requires very accurately calibrated standard gases. On the basis of the WMO (World Meteorological Organization) CO$_{2}$ Expert Meeting in 1975, the CO$_{2}$ standards are classified as WMO primary, secondary, tertiary, and working standards according to intended use. The primary standards are used to assign precise CO2 concentration values to the secondary standards which are maintained by each country as national standards. From 1979 through 1996, The CMDL secondary standards have been calibrated by the Scripps Institute of Oceanography (SIO) against the WMO primary standards approximately every 3 years. From mid-1996 to 2001, the assigned CO$_{2}$ values of the primaries have been jointly based on the SIO and CMDL manometric measurements, and completely on the CMDL manometric measurements alone from 2001 to the present. The secondary standards are transferred via NDIR analyzers to all other reference gas tanks, which are used routinely for measuring atmospheric CO$_{2}$ concentrations in the CMDL flask sampling network and at the CMDL observatories. In this presentation, the uncertainties for each reference tank calibrations by the NDIR against the secondary standards from 1979 to present are analyzed. The calibration uncertainties are about 0.03 mol/mol from September 1979 to March 1986, 0.007 mol/mol from April 1986 to August 1999, and 0.01 mol/mol from September 1999 to present. On the basis of these analyses, the total measuring uncertainties versus the WMO primary standards in the CMDL flask network and the observatories are estimated. These estimates can be considered as the upper limit of the precisions for determining atmospheric CO$_{2}$ concentrations during the period.
B23A-0928 1340h
Detecting patterns of forest canopy carbon uptake across multiple scales using flux measurements, high resolution remote sensing and ecosystem modeling.
We present detailed results from a study designed to asses the relative influence of forest canopy structural and chemical parameters on carbon uptake. Current remote sensing techniques for detecting productivity and gross photosynthesis rely predominantly on sensitivity to structural variables such as LAI and fPAR. While these parameters are certainly important, recent research has suggested that at certain scales, and particularly in dense canopies where the fPAR signal is saturated, variability in leaf nitrogen chemistry can explain observed differences in carbon uptake. Using high resolution hyperspectral and multi-angular optical remote sensing in combination with a biogeochemical vegetation model we explore our ability to scale surface measurements of gross carbon exchange measured at eddy co-variance towers to the resolution of broad scale sensors (MODIS and MISR). Both spatial and temporal dimensions of these data are presented. Remote sensing data and model results are presented from a series of sites (including Maine, Massachusetts, North Carolina and Florida) which span a climatic and vegetation gradient across the forests of the eastern United States.
B23A-0929 1340h
Quantifying Terrestrail Carbon Trends in the Conterminous U.S. - Overall Approach and Results From the Appalachian Forests
Estimating dynamic terrestrial ecosystem carbon (C) sources and sinks over large areas is crucial for C management, but it is complicated due to the variations of climate, soil, vegetation, and disturbances. The scaling of C sources and sinks from field to regional level has been challenging. As part of the U.S. Carbon Trends Project, we simulated the forest ecosystem C sequestration of the Appalachians region for the period of 1972 to 2000 using the General Ensemble biogeochemical Modeling System (GEMS). Land cover change was detected using sequential Landsat imagery within seventy-two sample blocks across the region. GEMS used the 60-meter resolution land cover change maps to capture stand-replacing events and used forest inventory data to estimate non-stand-replacing events that was not captured in those maps. GEMS also used Monte Carlo approaches to deal with spatial scaling issues such as initialization of forest age and soil properties. Ensemble simulations were performed to incorporate the uncertainties of input data. Simulated results show that from 1972 to 2000 the net primary productivity (NPP), net ecosystem productivity (NEP) and net biome productivity (NBP) ranged from 481 to 731(average 595), 97 to 329 (average 173), and 50 to 308 (average 136) g C m$^{-2}$ yr$^{-1}$, respectively. The inter-annual variability was mostly driven by climate. The inter-ecoregion variability showed the impacts from soil and land cover change. Model test revealed that without dynamic land cover change the annual C sink strength for this region would be over-estimated about 20 to 70 percent of the normal condition C sink. This over-estimated C sink was close to the amount of forest harvesting C in the normal condition. Detailed C budgets for the year 2000 were also analyzed. Within a total 148,000 km$^{2}$ forested area, average forest ecosystem C density was 161 Mg C ha$^{-1}$ , of which 81 Mg C ha$^{-1}$ was in biomass and 80 Mg C ha$^{-1}$ was in litter and soil. The total C stock of the Appalachian forests was estimated to be 2,375 Tg C including 1,191 Tg C in living biomass and 1,184 Tg C in litter and soil. The total net C sink of the forest ecosystem in 2000 was 13 Tg C y$^{-1}$.
B23A-0930 1340h
CASA-CQUEST: Decision Support Tools and Data Analysis for Ecosystem Carbon Management in the United States
Ecosystem modeling and satellite remote sensing are being used in combination to assess activities such as land use change and afforestation on carbon pools and fluxes across the United States. The main objectives of this research and application are to: 1) evaluate major forest and agricultural sinks of atmospheric carbon dioxide using NASA EOS satellite data and ecosystem modeling, 2) support U. S. Government interagency programs for voluntary greenhouse gas emissions reductions, and 3) develop an internet-based decision support tools for users nationwide. A new combination of ecosystem carbon modeling with CASA (Carnegie-Ames-Stanford Approach) and high-resolution land cover mapping of the country by the MODIS satellite sensor was used to estimate potential carbon sequestration rates in croplands and rangelands resulting from potential future afforestation activities. The high level of spatial detail produced from this remote sensing analysis permits state-level (and possibly finer) assessments of the areas best suited to crop and rangeland afforestation efforts, with results relevant to the North American Carbon Program.
http://geo.arc.nasa.gov/website/cquestwebsite/index.html
B23A-0931 1340h
Selective Cutting Impact on Carbon Storage in Fremont-Winema National Forest, Oregon
Management personnel of the Fremont-Winema National Forest in southern Oregon were interested in investigating how selective cutting or fuel load reduction treatments affect forest carbon sinks and as an ancillary product, fire risk. This study was constructed with the objective of providing this information to the forest administrators, as well as to satisfy a directive to study carbon management, a component of the 2004 NASA's Application Division Program Plan. During the summer of 2004, a request for decision support tools by the forest management was addressed by a NASA sponsored student-led, student-run internship group called DEVELOP. This full-time10-week program was designed to be an introduction to work done by earth scientists, professional business / client relationships and the facilities available at NASA Ames. Four college and graduate students from varying educational backgrounds designed the study and implementation plan. The team collected data for five consecutive days in Oregon throughout the Fremont-Winema forest and the surrounding terrain, consisting of soil sampling for underground carbon dynamics, fire model and vegetation map validation. The goal of the carbon management component of the project was to model current carbon levels, then to gauge the effect of fuel load reduction treatments. To study carbon dynamics, MODIS derived fraction photosynthetically active radiation (FPAR) maps, regional climate data, and Landsat 5 generated dominant vegetation species and land cover maps were used in conjunction with the NASA - Carnegie-Ames-Stanford-Approach (CASA) model. To address fire risk the dominant vegetation species map was used to estimate fuel load based on species biomass in conjunction with a mosaic of digital elevation models (DEMs) as components to the creation of an Anderson-inspired fuel map, a rate of spread in meters/minute map and a flame length map using ArcMap 9 and FlamMap. Fire risk results are to be viewed qualitatively as maps output spatial distribution of data rather then quantitative assessment of risk. For the first time ever, the resource managers at the Fremont-Winema forest will be taking into consideration the value of carbon as a resource in their decision making process for the 2005 Fremont-Winema forest management plan.
http://develop.larc.nasa.gov
B23A-0932 1340h
Extending the record of photosynthetic activity over the eastern United States into the pre-satellite period using surface diurnal temperature range
In this study, we demonstrate that mid-latitude surface measurements of diurnal temperature range (DTR) can be used to reconstruct decadal variability of spring, summer and fall terrestrial photosynthetic activity during and prior to the period with satellite retrievals of land surface greenness. The two relative maxima present in the seasonal evolution of DTR can determine the beginning and the end of the growing season of plants, and summertime average DTR can be used as a proxy of summertime terrestrial photosynthesis variability. In a case study over the eastern United States (1966-1997), we found from the DTR reconstructions significant decadal variability in photosynthetic activity. Satellite-observed positive trends in spring and summer photosynthetic activity are consistent with natural decadal variability and thus are not indicative of a secular, human induced trend. Comparison with temperature and precipitation data shows that summertime photosynthesis activity is primarily controlled by moisture availability while the timing of spring onset depends on both temperature and moisture conditions.
B23A-0933 1340h
Evidence of Vigorously Growing Old Trees in Eastern U.S. Forests
Many ecological, forestry and carbon sequestration models operate under the assumption that growth declines as trees age. Tree-ring studies at latitudinal and altitudinal treeline locations suggest that this may not always be true, especially over the last 150 years. Increment cores from < 1200 southern temperate trees were used to test the age-related decline hypothesis in the eastern U.S. Trees in this database range in age from 70 to 463 years, are comprised of four species ({\it Querucs alba, Q. prinus, Liriodendron tulipifera, Chameacyparis thyoides}) and include the oldest individuals documented by dendrochronology for {\it Q. alba} (463 years), {\it Q. prinus} (426 years), and {\it L. tulipifera} (335 years). {\it Quercus} trees were combined and grouped into six periods (1851-1900, 1801-1850,.pre-1651) to avoid a potential bias in growth trend by younger trees. Because age structure of {\it L. tulipifera} and {\it C. thyoides} forests are generally much younger, unique age classes were created for each species. The oldest trees for all species had periods of statistically significant, above average ring-width during the 20th century. The current trend of increased growth started in the mid- to late-1800s for {\it Quercus} and {\it L. tulipifera} while it started in the 1920s for {\it C. thyoides}. Using allometric equations to convert all chronologies to carbon increment reveals strong, positive trends in growth over the last century. These results show that even the oldest trees are taking up more carbon today than in the past. Because {\it L. tulipifera} and {\it C. thyoides} are less shade tolerant and have considerably different life-history traits than {\it Quercus}, it seems less likely that stand dynamics is the most important factor of these trends. Though old-growth forests are rare in temperate zones, our data can serve as a model for the large number of forests 100-180+ years old. It may not necessarily be true that forests in the eastern US will slow down in growth at ages 200 year and beyond. If ecosystem productivity declines with increasing tree age, changes in stand structure, environmental growth conditions or tree size may be the primary causes. Regardless, our results support treeline location studies to show that tree-scale productivity does not necessarily decline with age.
B23A-0934 1340h
Estimation of Regional Net CO2 Exchange over the Southern Great Plains
Estimating spatially distributed ecosystem CO2 exchange is an important component of the North American Carbon Program. We describe here a methodology to estimate Net Ecosystem Exchange (NEE) over the Southern Great Plains, using: (1) data from the Department Of Energy's Atmospheric Radiation Measurement (ARM) sites in Oklahoma and Kansas; (2) meteorological forcing data from the Mesonet facilities; (3) soil and vegetation types from 1 km resolution USGS databases; (4) vegetation status (e.g., LAI) from 1 km satellite measurements of surface reflectance (MODIS); (5) a tested land-surface model; and (6) a coupled land-surface and meteorological model (MM5/ISOLSM). This framework allows us to simulate regional surface fluxes in addition to ABL and free troposphere concentrations of CO2 at a continental scale with fine-scale nested grids centered on the ARM central facility. We use the offline land-surface and coupled models to estimate regional NEE, and compare predictions to measurements from the 9 Extended Facility sites with eddy correlation measurements. Site level comparisons to portable ECOR measurements in several crop types are also presented. Our approach also allows us to extend bottom-up estimates to periods and areas where meteorological forcing data are unavailable.
http://esd.lbl.gov/ARMCarbon
B23A-0935 1340h
The Impact of Coastal Meteorology on CO$_{2}$ Net Ecosystem Exchange Estimates: Implications for global inversion studies
A number of recent atmospheric inversion studies have concluded that there is a large northern hemisphere terrestrial sink for atmospheric CO$_{2}$. Many of the flask observations used in these analyses are collected at coastal stations or at remote offshore islands. Even though these sites are critical for our interpretation of continental-scale patterns of CO$_{2}$ fluxes, most of the work to quantify diurnal and synoptic variability in atmospheric CO$_{2}$ has been done at mid-continent locations and therefore the impacts of air-sea breezes, local topography, and transport of CO$_{2}$ from adjacent terrestrial ecosystems near the coastal domain are not well understood. Flasks are typically opened in the middle of the day, under windy conditions when sea-breeze circulations are common, and in locations with substantial topographic discontinuities, such as coastal bluffs and coastal mountain ranges. However, global transport models cannot resolve air-sea breezes, coastal discontinuities, and recirculation from nearby ecosystems. To determine the extent to which these factors impact the global inversions we apply a well-tested modeling framework that includes coupled meteorological (MM5), land-surface (ISOLSM), and tracer models. The MM5-ISOLSM model provides consistent predictions of net ecosystem CO$_{2}$, latent energy, and sensible energy exchanges at very fine resolutions. We apply the modeling framework to investigate atmospheric CO$_{2}$ sampling at Trinidad Head, CA over four months spanning the seasons in 2002. Our results demonstrate that sea and land breezes significantly impact atmospheric CO$_{2}$ concentrations, including those sampled in the middle of the day with an onshore breeze. Further, transport of CO$_{2}$ originating from ecosystem respiration north of Trinidad Head substantially impacts measured atmospheric CO$_{2}$. Variations from background associated with these `non-background' sources resulted in perturbations above background of 1.2, 0.5, 5.6, and 0.6 ppm during March, June, September, and December 2001, respectively, under strong onshore wind conditions. We characterize conditions (e.g., seasons, synoptic meteorology, topographical) where variations from expected background concentrations can be expected. This work will (1) help assess the impact of flask sampling criteria (such as minimum wind speed) on measured CO$_{2}$ concentrations at coastal sites; (2) identify biases induced by these criteria on long-term averages; (3) quantify the role of coastal mountain ranges and smaller-scale topographic heterogeneities on surface CO$_{2}$ concentrations; and (4) identify conditions where CO$_{2}$ originating from nearby terrestrial ecosystems can impact coastal sampling.
B23A-0936 1340h
Evaluation of Simulated Atmospheric [CO2] Using Analyzed Climate, Transport and Satellite Vegetation
We have simulated an hourly global atmospheric [CO2] field for the year 2000 with dual goals of gaining insight into the underlying mechanisms, as well as generating a global [CO2] field with associated uncertainties in order to provide a realistic lateral boundary condition field for regional model simulations and diurnally-varying priors to improve the performance of inversion studies. For our simulations, we are running Colorado State University's Simple Biosphere Model (SiB), versions 2.5 and 3.0, and Goddard Space Flight Center's Parameterized Chemical Transport Model (PCTM) in a step-wise coupled fashion, both driven by assimilated meteorological fields from the NASA Goddard - EOS Data Assimilation System (GEOS - 4), for the year 2000. Comparing the resulting [CO2] and CO2 flux field outputs with observations taken from flasks, continuous analyzers and aircraft campaigns (e.g., COBRA), we are diagnosing model strengths and weaknesses on various spatial and temporal scales. In addition, we are evaluating planetary boundary layer mixing, as this critical component of atmospheric transport and CO2 measurement is likely to be an important consideration in understanding the models' performance. By carefully considering these strengths and weaknesses together with driver data accuracy and "background flux" limitations (such as static fossil fuel emissions field for 1990), we are gaining insight into the underlying mechanisms as well as generating a global [CO2] field with associated uncertainties for use in regional model simulations and inversion studies. Note that by using surface meteorology from a self-consistent source (GEOS - 4) to drive biosphere CO2 fluxes, winds, planetary boundary layer turbulence and convective transport, we are allowing the models to "act in concert", as both CO2 flux and transport are influenced by identical forcings.
B23A-0937 1340h
Interannual Variability in Carbon Sources and Sinks over North America: How Important Compared to Other Regions?
We study the variability of carbon sources and sinks over N. America in relation to other major land regions, using both forward modeling and atmospheric inversion results. Simulation for the 20th century shows a dominant ENSO mode over global land regions, and another mode related to global temperature trends. Large sub-continental scale variations in carbon sources and sinks over North America and Eurasia are comparable to those in the tropics, and the total interannual variability over North America is about 1 Pg y$^{-1}$. Such spatiotemporal variability has implication for flux measurement network distribution. We highlight the key differences in ENSO related climate anomalies and plant/soil physiology in determining the distintly different contributions between N. America/Eurasia and the tropics. Fire, largely driven by drought, also contributes significantly to the total flux at a rate of about 1 Pg y$^{-1}$ globally, and 0.5 Pg y$^{-1}$ for North America and Eurasia. The robust variability in tropical fluxes agree well with atmospheric inverse modeling results. Even over North America and Eurasia, where ENSO teleconnection is less robust, the fluxes show general agreement with inversion results, an encouraging sign for fruitful carbon data assimilation.
B23A-0938 1340h
Influence of Organic Agriculture on the Net Greenhouse Effect in the Red River Valley, Minnesota
Fluxes for the suite of biologically-produced greenhouse gases (CH$_{4}$, N$_{2}$O and CO$_{2}$) are strongly influenced by agriculture, yet the influence of organic agriculture on all three gases, which comprise the net greenhouse effect (GHE), is not clear in the context of large-scale agricultural production. Greenhouse gas mitigation potential will depend upon the net balance for all three gases [GHE balance (CO$_{2}$ equiv.)= CO$_{2}$ $_{flux}$+ 23CH$_{4}$$_{flux}$ + 296N$_{2}$O$_{flux}$]. On-farm, field-scale experiments were performed to test the hypothesis that the net GHE at the soil-atmosphere interface is reduced under organic wheat production, compared with conventional, and that effects vary inter-seasonally. Trace gas fluxes were measured at the soil-atmosphere interface for organic and conventional wheat farms in the Red River Valley, Minnesota, one of the most productive agricultural regions in the US. We utilized 40-60 ha field pairs planted with hard red spring wheat (Triticum aestivum L.). Treatment pairs were located 6km apart and consisted of fields continuously cropped for wheat/soybean/sugar beet production for over 20 yr. Ten random, permanent points were generated for each 8.1 ha sub-plot nested inside each field. Each field pair was similar with respect to crop, climate, cultivation history, tillage, rotation, soil texture, pH, macronutrients, bulk density, and water holding capacity. Differences between treatments for the last five years were soil amendments (compost or urea) and herbicide/fungicide application versus mechanical weed control. We collected gas fluxes at each of the 41 points from April (wheat emergence) until the end of July (maturity) to determine the hourly and seasonally integrated net GHE for each management practice, given similar soil/plant/climatic conditions. Moreover, we analyzed inter-seasonal variability to determine the relationship between wheat phenology and flux under field conditions for soil temperature and moisture (water-filled pore space). The net GHE for organic fields was less spatially and temporally variable than conventional, with average daily flux between 0.48 and 1.44 g CO$_{2}$ equiv. m$^{-2}$ d$^{-1}$. Average daily flux in conventional fields ranged between 0.48 and 3.12 g CO$_{2}$ equiv. m$^{-2}$ d$^{-1}$, with highest values in April and May. While soil moisture in organic fields was significantly greater than conventional, it did not interact with treatment to affect trace gas flux. Instead, the effect of organic on N$_{2}$O, CO$_{2}$ and the net GHE was strongly influenced by crop stage, an agronomically meaningful proxy integrating time and plant growth conditions. Most CH$_{4}$ flux observations were 0. Integrated fluxes for each of the 40 sites over the growing season was averaged by field pair and treatment. Although the magnitude of the treatment effect for average seasonal integrated flux varied between field pairs for CO$_{2}$ and N$_{2}$O fluxes, the overall influence of treatment on the net GHE was similar. Overall, soils under organically produced wheat emitted 200 kg CO$_{2}$ equiv. ha$^{-1}$ per season less than conventionally produced wheat. We observed 1) the net GHE for similar field sites in the Red River Valley was reduced under organic versus conventional agriculture, 2) N$_{2}$O flux in organic fields was significantly lower than conventional fields for both field pair sites, and 3) the effect of treatment on CO$_{2}$ flux was site specific.
B23A-0939 1340h
Soil Carbon Changes in Transitional Grain Crop Production Systems in South Dakota
Corn-C (Zea Mays L.), soybean-S (Glycine max L.) and spring wheat-W (Triticum aestivum L.) crops were seeded as a component of either a C-S, S-W, or C-S-W crop rotation on silt-loam textured soils ranging from 3.0-5.0% organic matter. Conservation tillage(chisel plow-field cultivator) was applied to half of the plots. The other plots were direct seeded as a no-till (zero-tillage) treatment. Grain yield and surface crop residues were weighed from each treatment plot. Crop residue (stover and straw) was removed from half of the plots. After four years, soil samples were removed at various increments of depth and soil organic carbon (C) and nitrogen (N) was measured. The ranking of crop residue weights occurred by the order corn$<$$<$soybean$<$wheat. Surface residue accumulation was also greatest with residue treatments that were returned to the plots, those rotations in which maize was a component, and those without tillage. Mean soil organic carbon levels in the 0-7.5cm depth decreased from 3.41% to 3.19% (- 0.22%) with conventional tillage (chisel plow/field cultivator) as compared to a decrease from 3.19% to 3.05% (-0.14%) in plots without tillage over a four year period. Organic carbon in the 0-7.5cm depth decreased from 3.21% to 3.01% (- 0.20%) after residue removed as compared to a decrease from 3.39% to 3.23% (-0.17%) in plots without tillage applied after four years. The soil C:N ratio (0-7.5cm) decreased from 10.63 to 10.37 (-0.26 (unitless)) in the tilled plots over a four-year period. Soil C:N ratio at the 0-7.5cm depth decreased from 10.72 to 10.04 (-0.68) in the no-till plots over a four year period. Differences in the soil C:N ratio comparing residue removed and residue returned were similar (-0.51 vs. -0.43 respectively). These soils are highly buffered for organic carbon changes. Many cropping cycles are required to determine how soil carbon storage is significantly impacted by production systems.
B23A-0940 1340h
Soil Carbon Turnover and the Net Ecosystem Carbon Balance of a Northern Hardwood Forest, Michigan, USA
Soils are a major reservoir of stored carbon (C) in forested ecosystems, containing up to 70% of total ecosystem C. Heterotrophic activity largely dictates the rate of soil C turnover and directly impacts ecosystem C balance. Reliable estimates of net ecosystem productivity (NEP) from ecophysiological and biometric data as well as the refinement of process-based models predicting belowground changes in C storage depend on accurate quantification and partitioning of autotrophic and heterotrophic soil C fluxes. We used field and laboratory measurements of root, microbial and soil respiration in a northern hardwood forest to (1) quantify the annual soil C efflux attributed to heterotrophs and autotrophs from 1999 to 2003; (2) identify the extent to which microclimatic drivers impact interannual variability in microbial activity of the mineral soil and O-horizon; and (3) evaluate the sensitivity of estimated annual NEP to heterotrophic respiration. The study was conducted in an 85-year-old aspen-dominated mixed deciduous forest at the University of Michigan Biological Station Ameriflux site (UMBS$\sim$Flux) in N. lower Michigan, USA. Soil respiration was monitored from 1999 to 2003. Laboratory incubations of roots, mineral soil and the O-horizon at different temperatures were used to examine the relationship between microclimate and autotrophic and heterotrophic respiration. Empirical models relating root and microbial respiration to temperature were used in combination with soil respiration models and site soil temperature, moisture and root biomass data to estimate the contribution of autotrophic and heterotrophic respiration to total soil C efflux. Heterotrophic soil respiration estimates were combined with other C flux data to calculate annual NEP from 1999 to 2003. Microbially-mediated C turnover was responsible for $\sim$half of the total annual soil C efflux. Heterotrophic respiration varied by more than 1 Mg C ha$^{-1}$ yr$^{-1}$ among years primarily due to interannual variability in soil temperature rather than in the quantity of soil C inputs. Mean annual soil temperature explained over half of the interannual variability in heterotrophic respiration while fine root and litter inputs varied by no more than 6% among years and were not correlated with annual heterotrophic respiration. Heterotrophic respiration in 1999 was an annual high of 6.07 Mg C ha$^{-1}$ yr$^{-1}$ and contributed to a net ecosystem C loss of 0.25 Mg C ha$^{-1}$ yr$^{-1}$. In contrast, the ecosystem was a sink of 1.65 Mg C ha$^{-1}$ yr$^{-1}$ in 2001 when heterotrophic soil respiration was 5.02 Mg C ha$^{-1}$ yr$^{-1}$.
B23A-0941 1340h
Selecting Locations for Medium Intensity Sampling of Soil Organic Carbon in Alaska
A warming climate may lead to the release of the greenhouse gases carbon dioxide and methane from Arctic soils, resulting in a positive feedback. Quantifying soil carbon stocks is necessary for calibrating models of carbon flux. The total stocks of soil carbon in high latitudes are poorly quantified because there are few soil samples. As a guide to additional sampling, such as the new 3rd Tier Medium-Intensity Sampling planned in the North American Carbon Program, we have investigated methods of using existing soil databases to map the intensities and calculate the total quantities of soil carbon in Alaska. We have related the laboratory data on soil physical and chemical properties for specific locations (the pedon data) to the soil components of the digital maps in the State Soil Geographic (STATSGO) database. Both data sources are from the U.S. Department of Agriculture's Natural Resources Conservation Service. We evaluated 523 pedons for Alaska, selected those for which soil organic carbon could be calculated, and investigated several methods of matching them to the digital soil maps. The soil taxonomic classification (usually at the suborder, great group, or subgroup level), slope, and location were used in multiple combinations for linking the detailed profile descriptions to the general soil maps. We illustrate the uncertainty of estimations of carbon stocks by using alternative methods of performing the data selection and matching. Each matching method allows a many-to-many relation between the pedon point observations and the components of the soil maps. Once extrapolated to the map, a large measure of carbon stock attributed to a given sample point indicates an area in which additional sampling would efficiently reduce the uncertainty of the soil organic carbon estimates.
B23A-0942 1340h
Partitioned Soil Profile CO$_{2}$ Production in High and Low Resolution
Forest soil carbon exchanges are sensitive to changes in the physical environment associated with major disturbances such as forest management or climate change, which alter the distribution of roots and organic matter, heat and moisture through depth in soils. The individual responses of autotrophic and heterotrophic respiration ultimately determine the magnitude of feedback effects, shifts in temporal patterns, and relative importance of root and microbial respiration in forest soils. This paper reports preliminary findings of a project whose aim is to assess the physical controls on partitioned CO$_{2}$ production. Using trench plots and subsurface CO$_{2}$ measurements, autotropic and heterotrophic soil profile CO$_{2}$ production is being monitored at several forest sites, across a range of temporal resolutions from seasonal to 5 minute. A multi-layer diffusion model is being used to calculate CO$_{2}$ production values through depth, using soil diffusivity values obtained from in-situ diffusivity measurements. Emphasis will be given here to the spatial and temporal patterns of partitioned respiration, especially in the high-resolution record.
B23A-0943 1340h
Seasonal Changes of Soil Respiration Sources in a Boreal Forest
We studied how seasonal changes in soil temperature and moisture during the growing season (May-September) control soil respiration and potential sources (roots and microbial) at different forest recovery stages after fire. We investigated six black spruce stands in the BOREAS Northern Study Area in Canada (55N, 98W) on clay soils with underlying permafrost covering 0-150 yrs since fire. We measured the rate of CO$_{2}$ respired at the soil surface and CO$_{2}$ concentrations at various soil depths and its isotopic composition with dynamic chambers and soil probes. We sampled monthly, in 2004 we started automated continuous measurements. We determined the $^{14}$C signature of root-respired CO$_{2}$ in field incubations and that of microbial-respired CO$_{2}$ in laboratory incubations. The $\Delta$$^{14}$C signature of CO$_{2}$ was determined with accelerator mass spectrometry. In both years CO$_{2}$ fluxes from the soil surface increased from 6-21 mg C m$^{-2}$hr$^{-1}$ in spring to 35-135 mg in summer, and decreased to 25-65 mg in autumn. Fluxes were highest 40 yrs since fire. In early recovery stages burning intensity is a confounding factor to stand age, affecting the available C source and soil temperature and moisture. The $^{14}$C signature of the soil respired CO$_{2}$ followed the same trend with time since fire in both years. In early spring, fluxes were dominated by microbial respiration, the fraction of root respiration increased throughout the growing season. In autumn of the warm year 2003, soils of early recovery stages became a source of old mineral soil carbon to the atmosphere, while root respiration dominated in older stands. The incubations suggest that microorganisms within certain depth of the organic layer respire CO$_{2}$ with a constant $\Delta$$^{14}$C signature. However, increases in CO$_{2}$ fluxes at each depth with seasonal increases in temperature and moisture change the $\Delta$$^{14}$C of the CO$_{2}$ measured at the soil surface. Microbial respiration and its temperature and moisture response were highest in fresh litter. In young recovery stages, the dominant source of respiration was none of the mass dominating fractions (needles, moss, wood, roots, amorphous org. matter), and could potentially be dissolved organic C.
B23A-0944 1340h
Soil Carbon Stabilization Along Climate and Stand Productivity Gradients in Black Spruce Forests of Interior Alaska
The amount of soil organic carbon (SOC) in stable, slow turnover pools is likely to change in response to climate warming because processes mediating soil C balance (net primary production and decomposition) vary with environmental conditions. This is important to consider in boreal forests, which comprise one of the world's largest stocks of SOC. We investigated changes in soil C stabilization along four replicate gradients of black spruce productivity and soil temperature in interior Alaska to develop empirical relationships between SOC and stand and physiographic features. Total SOC harbored in mineral soil horizons decreased by 4.4 g C m$^{-2}$ for every degree-day increase in heat sum within the organic soil across all sites. Furthermore, the proportion of light fraction (density $<$1.6 g cm$^{-3}$) soil organic matter decreased significantly with increased stand productivity and soil temperature. Mean residence times of SOC (as determined by $\Delta$$^{14}$C) in dense fraction ($<$1.6 g cm$^{-3}$) mineral soil ranged from 282-672 years. The oldest SOC occurred in the coolest sites, which also harbored the most C. These results suggest that temperature sensitivities of organic matter within discrete soil pools, and not just total soil C stocks, need to be examined in order to project the effects of changing climate and primary production on soil C balance.
B23A-0945 1340h
The Source of Carbon for Root Respiration
In the Enriched Background Isotope Study (EBIS) that took advantage of a whole-ecosystem radiocarbon label that occurred in the temperate forest near Oak Ridge, Tennessee, we measured the radiocarbon signature of total soil respiration, heterotrophic respiration and root respiration, at different times during the last 3 growing seasons (2002-2004). By applying a mass balance approach, the relative and absolute contributions of heterotrophic and root respiration to total soil respiration were estimated. In contrast to heterotrophic respiration, root respiration seemed to be less affected by changes in soil moisture and temperature but rather showed a link to photosynthetic activity with a very similar pattern during the growing season as that of leaf area index. The radiocarbon signature of root respiration was very dynamic with low values in spring compared to the summer. The sources of variation can include changes in the local atmospheric signature and/or changes in the source of C being respired. Two different sites with different values and patterns of local atmospheric radiocarbon signature showed the same pattern in radiocarbon signatures of root respiration indicating that the source of variation was phenological. Low values during the spring could indicate the use of stored carbohydrates switching to more recent photosynthetic products as the summer progresses. As a first attempt to elucidate the source of C respired by roots, we will compare the radiocarbon content of starch, cellulose and soluble sugars in roots to that of bulk root material and root respired CO$_{2}$. These radiocarbon signatures can help us identify the pool of C that is most likely being respired by roots during the growing season. A better understanding of the source of C for root respiration has implications for understanding the role of root respiration in C cycling in temperate forests, specifically the timescale over which carbon is fixed through photosynthesis and returned to the atmosphere by root respiration.
B23A-0946 1340h
Whole Ecosystem Low-level 14C Pulse Labeling and CO2 Flux Measurements in a Boreal Forest
We developed a large volume, low level, 14C pulse-chase, field labeling method to determine the timing and contribution of recent photosynthetic products to total ecosystem respiration in a poorly drained black spruce forest stand in Manitoba, Canada. The site is part of a chronosequence of black spruce stands located in the BOREAS Northern Study Area (55N, 98W), and time since fire is 40 years. The radiocarbon addition was designed to produce a 14C signature of ~1500 times Modern for CO2 at ambient levels inside the ~37,000 L volume light chamber. At this level of labeling, the radioactivity in our 14C source (acidified sodium bicarbonate solution with specific activity of ~30 nCi/g) and in the chamber were well below levels that are regulated. We labeled two chambers in August 2004. The vegetation inside the first (37,000 L) chamber included black spruce trees (ranging from seedlings to 4 m tall) with feather moss and shrub understory. A second 14CO2 label was applied in a smaller chamber (500 L) containing only feather mosses. Both chambers were constructed from polyethylene plastic that allowed for 70 percent transmission of PAR. For seven days following the label, we measured the quantity and 14C content of soil respiration with small (10 L) dark chambers, above-ground respiration with branch bags, and total ecosystem respiration with a dark chamber. Live root and moss 14C content were measured by field incubations. Additionally, soil gas 14C content at two depths within the moss/organic layer was measured. Radiocarbon measurements are made using Accelerator Mass Spectrometry, which allows us to easily distinguish the presence of the label in small amounts (mg) of material. We will report the radiocarbon (delta 14C) signature of individual respiration sources. Preliminary results show that we can use these isotopic signatures to follow the labeled contribution of respiration from individual sources (moss, root/root exudates, and needle) to total ecosystem respiration. We plan on using this information to separate moss respiration from soil respiration, and furthermore, to partition fast from slow cycling soil respiration sources in this ecosystem. We will attempt to scale small chamber and branch bag respiration fluxes and 14C content to large (37,000 L) chamber measurements. Additionally, the large chamber CO2 flux measurements will be compared with eddy covariance measurements taken concurrently at the same site.
B23A-0947 1340h
Sensitivity analysis and quantification of uncertainty for isotopic mixing relationships in carbon cycle research
Quantifying and understanding the uncertainty in isotopic mixing relationships is critical to isotopic applications in carbon cycle studies at all spatial and temporal scales. Studies associated with the North American Carbon Program will depend on stable isotope approaches and quantification of isotopic uncertainty. An important application of isotopic mixing relationships is determination of the isotopic content of large-scale respiration ($\delta^{13}$C$_{R}$) via an inverse relationship (a Keeling plot) between atmospheric CO$_{2}$ concentrations ([CO$_{2}$]) and carbon isotope ratios of CO$_{2}$ ($\delta^{13}$C). Alternatively, a linear relationship between [CO$_{2}$] and the product of [CO$_{2}$] and $\delta^{13}$C (a Miller/Tans plot) can also be applied. We used an extensive dataset from the Niwot Ridge Ameriflux Site of [CO$_{2}$] and $\delta^{13}$C in forest air to examine contrasting approaches to determine $\delta^{13}$C$_{R}$ and its uncertainty. These included Keeling plots, Miller/Tans plots, Model I, and Model II regressions Our analysis confirms previous observations that increasing the range of measurements ([CO$_{2}$] range) reduces the uncertainty associated with $\delta^{13}$C$_{R}$. For carbon isotope studies, uncertainty in the isotopic measurements has a greater effect on the uncertainty of $\delta^{13}$C$_{R}$ than the uncertainty in [CO$_{2}$]. Reducing the uncertainty of isotopic measurements reduces the uncertainty of $\delta^{13}$C$_{R}$ even when the [CO$_{2}$] range of samples is small ($<$ 20 ppm). As a result, improvement in isotope (rather than CO$_{2}$) measuring capability is needed to substantially reduce uncertainty in $\delta^{13}$C$_{R}$. We also find for carbon isotope studies no inherent advantage to using either a Keeling or a Miller/Tans approach to determine $\delta^{13}$C$_{R}$.
B23A-0948 1340h
Atmospheric $^{14}$CO$_{2}$ Over the mid Pacific Ocean and at Point Barrow, Alaska, USA From 2002 to 2004
$^{14}$CO$_{2}$ is a useful tracer for studying the carbon cycle, in terms of determining residence times and fluxes between different carbon reservoirs, and understanding the various underlying processes. Knowledge of the regional and global distribution of atmospheric $^{14}$CO$_{2}$ is essential for many of these applications. We have recently begun measuring atmospheric $^{14}$CO$_{2}$ in the mid-Pacific and at stations in the US to add to the sparse available data for atmospheric $^{14}$C distribution. Our samples represent air collected over the time span of a few minutes from flasks. All of our samples were analyzed for $\Delta$$^{14}$C at the Keck AMS facility at UC Irvine. Atmospheric CO$_{2}$ samples were collected on a shipboard transect over the Pacific Ocean between Los Angeles (34$\deg$N, 118$\deg$W) and Auckland, New Zealand (34$\deg$N, 177$\deg$W) from fall 2002 to spring 2004. These samples were also measured for C-trace gas abundance (CO, CH$_{4}$ in addition to CO$_{2}$, and their stable isotopes). The 2002 transect (collected from Sept. 23 to Oct. 4, 2002) terminated in Manzanillo, Mexico (16$\deg$N, 109$\deg$W) instead of from Los Angeles. The 2002 transect showed that $\Delta$$^{14}$C in atmospheric CO$_{2}$ in this latitude range was relatively uniform spatially during the collection period. The average $\Delta$$^{14}$C value of all 24 samples was 79.9$\pm$2.3$\permil$ (1$\sigma$). The spread of the data was comparable to our analytical error estimated by standard measurements. There was a slight decreasing trend in $\Delta$$^{14}$C of air CO$_{2}$ northward of 6$\deg$N, consistent with an increase in fossil fuel inputs to air in the northern hemisphere. The 2003 transect (from July 2003 to Aug. 2003) was similar to that of 2002 transect, in terms of the latitudinal distribution. It gave an average $\Delta$$^{14}$C value of 75.4$\pm$2.7$\permil$ (1$\sigma$), indicating a decrease of approximately 4.5$\permil$ per year. Samples from a coastal mid-latitude sites (Monta\~{n}a de Oro State Park, CA (35$\deg$N, 121$\deg$W) show a similar average $\Delta$$^{14}$C value of 76.1$\pm$6.5$\permil$ from January 10, 2003 to April 27, 2003, while data from a mid-continental site at Niwot Ridge, CO (41$\deg$N, 105$\deg$W) were significantly higher during the same period (83.7$\pm$1.7$\permil$). The time series of atmospheric $^{14}$CO$_{2}$ at a coastal site at Point Barrow, Alaska (71$\deg$N, 157$\deg$W) from July 12, 2003 to August 18, 2004 shows a general decreasing trend with time. The average $\Delta$$^{14}$C of this time series was 66.6$\permil$ with a range of about 11$\permil$, showing a hint of a seasonal cycle and lower $^{14}$C values at high northern latitudes. Low $\Delta$$^{14}$C values in Pt. Barrow air correlate with wind direction, indicating that part of the temporal variation may be caused by the advection of low $^{14}$C air from lower latitudes. Our results confirm large-scale patterns in atmospheric $^{14}$C predicted using carbon cycle models coupled with models of atmospheric transport. We plan to continue measuring radiocarbon in CO$_{2}$ in the mid-Pacific and at the surface US stations in different seasons for the next several years for a fuller picture of seasonal and latitudinal variation in atmospheric $\Delta$$^{14}$C.
B23A-0949 1340h
Mapping Regional Patterns of Fossils Fuel CO$_{2}$ in the Planetary Boundary Layer Across North America Using Radiocarbon in Annual Plants
Radiocarbon levels in annual plants provide a means to map out regional and continental-scale patterns of fossil fuel emissions and biosphere-atmosphere exchange. The imprint of the local atmosphere is recorded within the leaves of these annual plants and represents a time-integral of atmospheric levels over a period of several months, complementing both flask and aircraft sampling techniques. Working with colleagues, we collected corn (Zea mays) from approximately 70 sites across North America. We designed a sampling protocol that captured regional and continental scale patterns of fossil fuel CO$_{2}$ levels; we specifically avoided areas directly influenced by point sources such as major roads or cities. We then analyzed the leaves in the W.M. Keck Carbon Cycle Accelerator Mass Spectrometer at the University of California, Irvine. In areas where the corn plants were exposed to sustained and elevated levels of CO$_{2}$ from fossil fuel emissions, such as the Ohio Valley, the $^{14}$C/$^{12}$C ratio of the corn leaves was reduced. We found that there was a drop of approximately 15 per mil between the western U.S. (Alberta, Idaho, Colorado and New Mexico) and the Northeastern U.S. (Ohio, Maryland and Pennsylvania). This corresponds to approximately 5 ppm increase in fossil fuel CO$_{2}$ levels, as air moves from west to east across the continent. These data provide a means to test our understanding of the coupling of biosphere atmosphere exchange, planetary boundary layer mixing, atmospheric transport, and fossil fuel emissions in mesoscale and global models that are used to estimate the spatial distribution of carbon sources and sinks.
B23A-0950 1340h
Fossil-Fuel-Derived Carbon Dioxide Emissions at Monthly Resolution for the Countries of the North American Carbon Program
Examination of national statistical databases has allowed for the widely-used, CDIAC-housed data set on annual, fossil-fuel-derived carbon dioxide emissions to be subdivided into monthly time intervals. This analysis focused on establishing reliable statistics that represent the solid, liquid, and gaseous fuels consumed in each country at monthly time scales. An intermediate product of this analysis was the fraction of the annual total consumption occurring in each month for each fuel. This monthly fraction was then multiplied by the annual carbon dioxide emission value to obtain a monthly emission estimate. This has the benefit of yielding monthly and annual emissions time series that are mutually-consistent. This presentation will give monthly emissions for multiple years for the United States, Canada, and Mexico.
B23A-0951 1340h
Assimilating carbon flux data to a soil carbon model to estimate pool sizes and their associated turnover time
Soil heterotrophic respiration depends not only on temperature but also on the dynamic properties of soil organic carbon (SOC). In modeling study, it is necessary to separate the SOC into multiple pools according to their turnover time. The pool sizes and their turnover time can only be determined indirectly. In this study, we show how the assimilation of carbon flux data into a soil carbon model can be used to obtain the needed information. An adjoint system is applied to minimize the difference between the flux data and the model simulation. It also provides the sensitivity of the model to each control variable.
B23A-0952 1340h
An Invasive Grass Species Alters Carbon Cycling in Hawaiian Dry Forest
At lower elevations on the leeward side of the island of Hawaii, remnant native forests are heavily invaded by an introduced African bunchgrass, {\it Pennisetum setaceum} (fountain grass). Our research is designed to determine the consequences of this invasion for carbon (C) cycling in Hawaiian dry forests. We examined above- and belowground C pools and fluxes in 400 m$^{2}$ replicated forest plots ({\it n} = 4) with fountain grass (grass plots) and in areas where fountain grass had been removed for $\sim$3 years (removal plots). C pools were estimated with direct sampling and allometric equations developed {\it in situ} for the dominant tree species. Aboveground net primary productivity (ANPP) was estimated as aboveground biomass increment plus litterfall minus loss from mortality (trees) and with clip plots (grass and herbaceous species); total belowground carbon allocation (TBCA) was estimated using a conservation of mass, C balance approach. Our results indicate that the invasion of a non-native grass in this ecosystem has considerable impacts on both C pools and fluxes. Aboveground, tree biomass did not differ between treatments ({\it P} = 0.57) but the presence of fountain grass led to a 7.5-fold increase in understory biomass in grass plots compared to removal plots ({\it P} $<$ 0.01). Tree ANPP was significantly higher in removal plots for both foliage (0.10 and 0.06 kg C m$^{-2}$ yr$^{-1}$ for removal and grass plots, respectively; {\it P} = 0.02) and wood (0.13 and 0.05 kg C m$^{-2}$ yr$^{-1}$ for removal and grass plots, respectively; {\it P} $<$ 0.01). However, grass ANPP was $\sim$35% greater than tree foliage productivity in grass plots. Despite this added foliar productivity, total ANPP (Tree + Grass ANPP) was significantly higher in removal plots ({\it P} = 0.04). Belowground, grass plots exhibited higher rates of soil-surface CO$_{2}$ efflux (1.09 and 1.38 kg C m$^{-2}$ yr$^{-1}$ for removal and grass plots, respectively; {\it P} = 0.03 ). Likewise, TBCA was significantly higher in grass plots (1.21 kg C m$^{-2}$ yr$^{-1}$) than in removal plots (0.97 kg C m$^{-2}$ yr$^{-1}$; {\it P} = 0.04). Tropical dry forests globally, and Hawaiian dry forests in particular, are among the most threatened terrestrial ecosystems. Our results indicate that the presence of an invasive, non-native grass species changes both ecosystem structure and function in these forests. These changes in above- and belowground C pools and fluxes are particularly important in light of the ubiquitous presence of invasive species in most terrestrial ecosystems and the need for a better understanding of the role that they will play in global C cycling and climate change.
B23A-0953 1340h
Growth Season Dynamics Of CO$_{2}$ Exchange In A Subarctic Mire: A Comparison Of Automated Chamber Measurements During Three Years
Peatlands are well-known to be a long-term sink for atmospheric carbon dioxide (CO$_{2}$). However the carbon balance and, hence, CO$_{2}$ flux can be significantly changed and peatlands may even become a significant atmospheric carbon source in a changing climate. Here we present results of CO2 flux measurements obtained by an automatic chamber method in a subarctic mire (Stordalen, $68\deg$22'N, $19\deg$03'E) during 3 seasons, 2002 to 2004. The study years had quite different climate (temperature, precipitation), causing different seasonal CO$_{2}$ flux patterns. In this presentation a detailed analysis of the causes for interannual differences in the carbon balance will be presented. Three different ecotypes (dry ombrotrophic, mesotrophic and wet minerotrophic) are studied and significant differences between their functional responses to different climate conditions were found. All three were atmospheric sinks in terms of accumulated CO$_{2}$ fluxes during the growing season (90 days). The dry ombrotrophic system accumulated 20-30 g C/m$\^{2}$, the mesotrophic between 37 and 43 g C/m$\^{2}$ and the wet minerotrophic system between 70 and 115 g C/m$\^{2}$. The interannual variability was mainly controlled by variations in snow-melt and precipitation patterns and the subsequent effects these have on the soil moisture regime. The CO$_{2}$ flux measurements presented provide a useful compliment to landscape scale micrometeorological (eddy correlation) measurements of CO$_{2}$ exchange over the mire conducted at the same site. Extrapolating the automated chamber fluxes to the mire as a whole gives mean growing season uptake rates that compare well with the corresponding numbers obtained with the eddy correlation method.
B23A-0954 1340h
Comparison of the two biosphere models to estimate global terrestrial carbon fluxes
For understanding and predicting global warming, it is important to accurately estimate the global terrestrial carbon fluxes, especially Net Primary Production (NPP) and Net Ecosystem Production (NEP). However, as actual measurements of these fluxes in global scale are difficult to realize, the only possible way is to use biosphere models. Recently, many biosphere models have been proposed, but the confidence of carbon fluxes derived from these models are not clear. This study aims at comparing spatial and temporal patterns of NPP and NEP between two terrestrial biosphere models (Sasai et al. model and Sim-CYCLE) and three re-analysis climate datasets (NCEP/NCAR, NCEP/DOE, and ERA40). The two models differ in their concept; Sim-CYCLE is more prognostic and driven by sole climate data, while Sasai model is more diagnostic and driven by both climate and satellite data (fAPAR and LAI by NOAA/AVHRR). Both models incorporate a carbon cycle (photosynthesis, respiration, litter fall, and soil decomposition) and hydrology (evaportranspiration and runoff) scheme. Simulations were performed globally from 1982 to 1999, using the three datasets for air temperature, downward short and long wave radiation, precipitation, and humidity. By comparing the six results (two models by three datasets) of NPP and NEP, we expect to address the current uncertainty in our model simulations. First, comparison of climatic variables revealed that the three datasets were mostly consistent in large-scale patterns and interannual variability, but inconsistencies were found for several features (e.g., incoming solar radiation; NCEP/NCAR < NCEP/DOE < ERA40). Second, interannual changes in the estimated NPP and NEP were approximately comparable between models and input data (i.e., within _}1GtC/yr). Here, we focused on the nature of differences and found that the differences in model structure and forcing data affected the estimation of NPP and NEP, since photosynthesis and respiration are sensitive to climatic conditions. Therefore, when we derive some conclusions from simulations, we should pay attention to specific characteristics in models and datasets. Further model validations with observational data, such as flux measurements, are also required. Finally, intercomparison studies between models and datasets, as presented here, should carry implications for global carbon-cycle researches.
B23A-0955 1340h
The Importance of Soil Mineralogy to Plant Nutrient Availability in the Northeastern U.S.
In the northeastern U.S., acid deposition poses a threat to nutrient availability, via leaching of base cations (Ca, Mg, K, Na). While silicate mineral weathering may be the source of most base cations released from soils over millennia, more easily weathered minerals may provide nutrients necessary to meet short term demand resulting from acid deposition and forest harvesting. Through sequential leaching of soils and their parent materials, we can determine whether easily soluble trace minerals are potentially available to plants. Previous studies have shown that apatite, a calcium phosphate mineral present in trace amounts, may provide a significant source of Ca to vegetation at the Hubbard Brook Experimental Forest, New Hampshire. In this study, we explore the regional availability of apatite in soils across the northeastern United States. Soils derived from granitoid and sedimentary rocks were collected from 20 sites across the northeast U.S. and sequentially leached to determine relative availability of base cations. A leach using 1M nitric acid extracted Ca and P from soils developed on crystalline parent materials (0.02 to 0.04 mmol Ca/g soil, 0.01 to 0.03 mmol P/g soil). The Ca:P ratio is 5:3, the stoichiometric ratio of apatite. The presence of apatite in these soils was verified by SEM analysis. The lack of K, Na and Si in this leach suggests that silicate mineral dissolution is not the source of Ca. The Ca and P concentrations indicate that amount of apatite varies in granitoid soil parent materials across the northeastern U.S. Sedimentary rock-derived soils did not contain appreciable amounts of apatite. With the exception of carbonate-derived soils, only small concentrations of Ca ($<$0.006 mmol/g soil) were leached from sedimentary rocks in 1M nitric acid.
B23A-0956 1340h
The Yearly Carbon Balance of a Low-Arctic Wetland.
Changes in the high latitude carbon pools have been an important topic in the climatic change debate over the last decades, due to the large amounts of carbon stored in especially wetland ecosystems of the region. The predicted temperature rise in most parts of the high latitudes is likely to have effect on both ecosystems and the atmospheric concentration of especially CO2 and methane. Over the last four years continuous measurements of the CO2 exchange have been carried out over the Stordalen mire in the Sub-Arctic region of Sweden, where the permafrost in wetlands is on a rapid retreat. Our measurements show a small annual uptake of CO2 (10-30 g C m-2 y-1) and a corresponding emission of CH4 in the order of 5 g C m-2 y-1, which makes this mire site a small net sink of carbon but a likely source of greenhouse gas to the atmosphere. Here we describe the inter-annual variations and the functional relationship between the exchange of the two gases and prevailing climatic conditions of the area.
B23A-0957 1340h
Permafrost Melting and the Age of Carbon Respired From Arctic Tundra
Up to 450 Pg of soil carbon (C) has accumulated in high latitude ecosystems after the retreat of the last major ice sheets. This soil C has until now been largely protected from decomposition by cold temperature, waterlogging, and permafrost. Recent studies suggest that, due to climate warming, these ecosystems may no longer be accumulating C, and in some cases may be losing stored C to the atmosphere. We hypothesize that sustained transfers of C to the atmosphere that could cause a significant positive feedback to climate change must come from old C, which forms the bulk of the soil pool. We used radiocarbon measurements of carbon dioxide to detect the age of C respired from tussock tundra near Denali National Park, Alaska. At this alpine tundra site, permafrost has been observed to warm and melt over the past several decades, causing the ground surface to subside as ice volume in the soil decreased. We established three sites within this area that differed in vegetation and surface topography. These sites represent differences in time since the onset of permafrost melting and/or in the magnitude of ecosystem change. We made radiocarbon measurements of ecosystem respiration and of incubations of above and belowground plant biomass to determine the age of C respired from these sites. Ecosystem respiration radiocarbon values ranged from +58$\permil$ to +114$\permil$ and there was as much variation within a site as there was between sites reflecting the large variability in surface topography at all sites. Respiration from belowground plant biomass had an average value of +83$\permil$ and was significantly higher than the value of +67$\permil$ respired from aboveground plant biomass. Respiration from plant biomass in general was higher than the current atmospheric radiocarbon value of +60$\permil$, suggesting that plant respiration was derived in part from a C storage pool within stems and rhizomes that was more than a year old. Most ecosystem respiration radiocarbon values were higher than values for plant respiration, indicating that the majority of decomposition of soil organic matter was derived from C fixed over the past several decades. In contrast, some individual plots had radiocarbon values below that of plant respiration, suggesting a larger contribution to respiration from organic matter that was more than 50 years old, likely as a result of deeper soil thaw.