B43D-1578
In Situ Measurements of Atmospheric O2 and CO2 in a Central Massachusetts Forest
Simultaneous and continuous measurements of O2 and CO2 made in the air around terrestrial ecosystems have the potential to improve our understanding of the biogeochemistry of the ecosystem, and may reduce uncertainties in estimates of terrestrial carbon uptake derived from atmospheric O2 measurements. Following the design of Stephens et al. [2007], we have constructed an instrument that performs continuous in situ measurements of atmospheric O2 and CO2 concentrations. We present design and performance data, along with recent results from a deployment at the Environmental Measurement Site at Harvard Forest in central Massachusetts. Preliminary results show that O2 and CO2 covary with well- defined stoichiometries that differ in summer and winter months. During the growing season, values are close to 1.0, consistent with photosynthesis and respiration of sugars. In the winter, values are elevated (around 1.2), possibly reflecting the signature of fossil fuel combustion occurring outside of the immediate study area.
B43D-1579
Surface CO2 Isotopologues: Influence of Stratosphere-Troposphere Exchange
Current estimates of gross carbon flux tend to ignore the downwelling flux of CO2 from the stratosphere, through the downward branch of Brewer-Dobson circulation at high latitudes. Recent observations showed that there is a phase shift between a time series for the concentration of the standard isotopologue C16O16O and that for C16O18O at the surface at Waliguan, China (36 17' N, 100 54' E, 3810 m). Using a 2-dimensional model of the troposphere and stratosphere, we show that this phase shift could be explained by the downwelling of CO2 from the stratosphere, where C16O18O is known to be enriched. Using O3 as a proxy of stratosphere-troposphere exchange, we find there is excellent correlation between O3 and C16O18O observed at the Waliguan site. The observed variability of C16O18O is consistent with the predictions of our model, thus providing the first evidence of stratospheric CO2 near the surface. Quantitative modeling may provide a powerful tool for constraining the sources and sinks of CO2 using the isotopically enriched CO2 from the stratosphere as a tracer.
B43D-1580
Investigation of the Spatial Heterogeneity and Fluxes of Carbon Dioxide Using a Lagrangian Approach From an Aircraft
Aircraft measurements of CO2, sensible heat, and 3-dimensional winds were conducted as part of the Mid- Continent Intensive campaign of the North American Carbon Program, in June of 2007. These "top-down" atmospheric budget observations were used to assess the horizontal and vertical spatial heterogeneity of CO2 concentrations throughout the boundary layer and the lower free troposphere and to compare with observed "bottom-up" budgets as measured by a network of tall towers surrounding a predominately agricultural landscape in Iowa (USA). Regional aircraft-based CO2 fluxes were estimated using a Lagrangian approach by measuring a net drawdown of mixed-layer averages of CO2 concentration from mid-morning to late afternoon due to primary productivity in the region. The results from over 53hrs of flights spanning 11 days in June 2007 are presented. A statistical analysis of the spatial variance of CO2 mixing ratios and the application to ecosystem exchange and transport models are discussed. http://ring2.psu.edu/
B43D-1581
?14C of Atmospheric CO2 over the Subtropical and Equatorial Pacific and at Point Barrow, Alaska
Δ14C is a unique tracer for studying the carbon cycle, especially for discriminating between fossil and biosphere carbon emissions. However observations of Δ14C variation in atmospheric CO2 are available for only a few locations. We have been measuring atmospheric 14CO2 in boundary layer air over the subtropical and equatorial Pacific and at stations in the US since 2002 to expand the 14CO2 database with high precision data that sample either at high temporal resolution (Point Barrow, Alaska) or high spatial resolution (cross-equatorial mid-Pacific). These data provide observational constraints for the roles of 14C isotope disequilibirum in the tropical terrestrial biosphere, the Southern ocean, and fossil fuel burning and enhance our understanding of the patterns of atmospheric 14CO2 distribution and its seasonal variation. Five transects of atmospheric 14CO2 were collected on shipboard over the Pacific Ocean between Los Angeles (34°N, 118°W) and Auckland, New Zealand (34°N, 177°W) from fall 2002 to summer 2005. Abundances of CO and CH4 in addition to CO2, and their stable isotopes were also measured for these samples. The high precision of our Δ14C analysis (~2‰ based on duplicate measurements) allows us to observe relatively small variations over the latitude span investigated. All five transects show that Δ14C in atmospheric CO2 were relatively uniform from the equatorial region to 30°S latitude. They also indicate a consistent decreasing trend in Δ14C (~7‰) northward of ~6°N to 30°N latitude, consistent with an increase in fossil fuel input in the northern hemisphere. From fall 2002 to summer 2005, Δ14C decreased by an average rate of 6‰/year, with a slightly higher rate of decrease over the southern ocean. Correlation between CO mixing ratio and Δ14C indicates short-term atmospheric circulation may significantly affect the 14CO2 distribution pattern and its latitudinal gradient. In addition, signals from seasonal variation of 14CO2 could be superimposed on the transects' latitudinal variation. We have also been measuring Δ14C in two air samples biweekly from the Point Barrow Observatory, Alaska (71°N, 157°W) since July 2003. In this period, Δ14C decreased by 5- 6‰/year, to ~52‰ in Feb-2007. We find a distinct seasonal cycles for 14C, with a broad minimum around April and a maximum in September with an amplitude of 5-7‰. This seasonal pattern is highly variable from year to year. Increasing 14C values may reflect injection of stratospheric air in April and May, and higher soil respiration with enriched 14CO2 between May to August; rapid declines may be due to reduction in soil respiration and changes in the poleward advection of fossil fuel burned in the winter months.
B43D-1582
Foraminiferal Calcium Carbonate Pump Response to Temporal Changes of Carbon dioxide System in the Sundarban Mangrove Environment, NE Coast of Bay of Bengal, India
Rising atmospheric CO2 to changes in seawater carbonate chemistry, which could affect marine biogenic calcification of the foraminifera. This study reports the production/ dissolution of foraminifera in relation to the air- water exchange of CO2 from mangrove water at the land-ocean boundary condition of Sundarban. Globigerina sp. and Ammonia beccarii were found to be most abundant amongst the foraminifera occurring in the 0-2 cm layer of the mangrove sediment coloumn. Other species were found to be Nominella stella, Milammina fusca, Protelphidium sp., Elphidium rugulosum, Tipotrocha sp., Discorbis sp., Rotalia sp., Tentularia sp. etc. Salinity of the water during the study period (Sept ´06 to Feb ¢07) varied between 15 and 28 psu and temperature, between 24.5 and 30°C. Air-Water flux of CO2 varied between – 4.8 and 101.9 mmol m-2h-1. Tidal water was supersaturated with respect to calcite (37 – 1243%). Foraminifera samples were classified into 7-9 size mode fractions. The observed radius (r) range was found to be between 7.47 and 52.45 mm. Number (N) / Volume (V) size distribution dN or dV/dlogr of the total foraminifera samples were observed either in trimodal or bimodal. Volume-size distribution with respect to depth showed prominent peak in the size range between 10-20 mm, 22- 42 mm and > 42 mm. However, the maximum volume of dV/dlogr (peak height) was found to be decreased from 1.1´ 109 mm3cm -2 at 0-2cm layer of the sediment column to 9.1´ 108 mm3cm-2 at 2-4cm and to 4.37´108 mm3 cm 2 at 4-6cm layer of the sediment column indicating dissolution. However, maximum values of dV/dlogr in the radius range 20-40mm at 0-2cm layer of the sediment column was found to be increased from 1.14 ´ 107 mm3cm-2 in September to 1.106x 10 9 mm3cm-2 in November indicating production while partial pressure of CO2 in the tidal water was found considerably decreased from 659 matm in September to 553 matm in November. Even though foraminifera showed calcium carbonate production in the surface sediment layer (0- 2cm), refered to as the calcium carbonate pump, water still showed emission flux to the atmosphere but at reduced rate.
B43D-1583
An Evaluation of Soil Carbon Layer Dynamic in the Context of Global Warming
Boreal soils contain approximately 40-45% of the world's terrestrial soil C and may play an important role in the feedback between the global carbon cycle and climate change. Although there are a number of predictions of future boreal soil C dynamics under various warming scenarios, there are still many uncertainties due to the complexity and uniqueness of boreal soils. In most modeling analyses of boreal C cycle responses to warming, the deep, organic C rich layers are given little attention or assumed to have little role in future C dynamics. The objective of the study was to improve our understanding of the sensitivity of boreal soils to climate warming. A soil C model with dynamic soil layers and multiple C pools structure (fine, coarse, humic) was used to simulate C cycling in the Northern Study Area Old Black Spruce site of the Boreal Ecosystem and Atmosphere Study. The C model with the parameters used in the study has been shown to provide the best simulation on both soil 14C profile and total C for the study site. The simulation was designed in a factorial combination of two factors: temperature increase (0°C, 2°C, 4°C, 6°C, 10°C) and warming period (2000-2200, 2000-2400, 2000-2600, 2000- 3000). The model predicted that warming would lead to significant losses of soil C in the future that these losses would scale linearly with the magnitude of warming. The simulation results also indicated a shift in the sources of CO2 from soil decomposition from equally balanced between surface and deep (e.g. 55 cm) C sources under current conditions to a dominance of deep C contributions to CO2 by 2200, even with low to moderate warming. This result is driven by centennial scale changes in soil thermal regimes that facilitate the decomposition of large deep C stocks, despite their relative recalcitrance. These simulations also indicate that organic rich, deep boreal soils have long equilibration times, exceeding 1000 years in some cases, in response to temperature perturbations. Collectively, these results suggest that the deep soil C stocks in boreal forests are likely to be critically important to the future carbon dynamics in these regions and future work is needed to examine the feedbacks between vegetation change, primary productivity and soil carbon cycling in these regions.
B43D-1584
Twentieth Century Changes of Forest Carbon Pools of a Northern Hardwood Forest with Emphasis on Soil Organic Carbon Pools
Changes in forest Soil Organic Carbon (SOC) pools are difficult to detect because 1) the long timescales involved, 2) relative differences are small compared to aboveground biomass C, 3) the number of samples may not capture spatial heterogeneity of soils in forest stands, 4) sampling methods may sacrifice accuracy for sample size and 5) uncertainty about land use changes. These challenges are addressed in this study by using detailed field data and by applying ecosystem models to simulate changes in biomass C and SOC of 22 even- aged northern hardwood stands stretched throughout the Green Mountains of Vermont. A century long "snapshot" of results reveal overall small (less than 10 percent) decreases in SOC throughout the Twentieth Century as a result of clearing the land for settlement and logging. These legacy effects, borne out only by simulations, indicate that managed forest systems may be steadily, albeit slowly, declining in SOC content. Net Ecosystem Production for even-aged stand in this region was still positive in the past century because SOC decreases were small compared to aboveground biomass increases. The potential for simulation models to estimate future responses of forest growth and changes in SOC pools was also explored and are presented. Overall, results indicate that simulation outputs are reliable and adequate at both local and regional scales given sufficient information about vegetation, soils and land-use history.
B43D-1585
North American Results From a New Global Atmospheric Inversion
The objective of this study is to characterize North American carbon flux at regional spatial resolution and monthly temporal resolution to inform our mechanistic understanding of the underlying carbon cycle dynamics. A top- down method is used: a global atmospheric Bayesian synthesis inversion in the TransCom style, with 10 regions in North America (47 land regions and 11 ocean regions globally). We address the problem of sparse measurements in global inversions by complementing the traditional, primarily marine boundary layer, global carbon dioxide measurement network with high-precision, well-calibrated carbon dioxide mixing ratio observations at North American flux towers. Mid-day observations (when the atmospheric boundary layer is likely to be well-mixed at these continental sites) and virtual tall tower estimates of mid-day mid-mixed layer mixing ratios are tested. The method uses the NASA PCTM tracer transport model with Goddard Modeling and Assimilation Office GEOS-4 meteorological fields appropriate to the inversion time period. Forward integrations with the PCTM model support solutions for monthly fluxes for 2001-2003 as well as a monthly mean, cyclostationary, solution covering the same time period. Background fluxes include seasonal fossil fuel emissions and air-sea flux, and interannually- varying seasonal terrestrial biosphere and biomass burning fluxes. Preliminary results from the inversion will be shown. http://cheas.psu.edu/data/people/mpbutler/AGU2007
B43D-1586
Contributions of Deciduous and Evergreen Trees to the Seasonal Dynamics of CO2 and Water Vapor Exchange Over Developed Land in the Midcontinental United States
Half the world's population currently lives in urbanized areas, a proportion expected to increase to 60% by 2030. The clearing of agricultural and natural ecosystems for urban and suburban development is consequently one of the fastest rates of land use change around the world. Although developed land areas represent major sources of CO2 and alter hydrology, they are also ecosystems with significant vegetation cover that takes up CO2 and humidifies the atmospheric boundary layer. Direct measurements are needed to develop a mechanistic understanding of how vegetation contributes to the land-based CO2 sinks and evapotranspiration in developed areas, especially residential developed land which accounts for most of the land-use change in the United States. We quantified whole-tree transpiration, and modeled canopy conductance and canopy photosynthesis using thermal dissipation sap flow and leaf-level gas exchange measurements on stands of deciduous and evergreen trees in a suburban residential neighborhood of Minneapolis-St. Paul, Minnesota. A suite of environmental variables was continuously monitored at each site. During the 2007 growing season, seasonal drought and synoptic and diurnal variations of vapor pressure deficit (VPD) strongly controlled transpiration and photosynthesis in both the deciduous and the evergreen stands. The results are analyzed against a phenology data set (including leaf area index (LAI), and plant and soil biophysical properties) measured at a larger number of stands within a suburban region during the 2006 growing season. These results will allow us to determine the relative contributions of different tree plant functional types to the seasonal dynamics of CO2 exchange and evapotranspiration in developed land, and to scale up the effect of these land cover types on regional carbon and water budgets. This study is a contribution to the North American Carbon Program (NACP) Mid-Continent Intensive Field Campaign.
B43D-1587
Examining the Influence of Teleconnection Patterns on CO2 Fluxes at an Old-Growth Forest Scaling from Stand to Region Using MODIS
In this study, year-to-year variability in three of the major Pacific teleconnection patterns were examined to determine if CO2 and H2O fluxes at an old-growth forest in the Pacific Northwest were affected by climatic changes associated with these patterns. The three cycles examined are the Pacific Decadal Oscillation, Pacific/North American Oscillation and El Niño-Southern Oscillation. We centered our study on the Wind River Canopy Crane, an AmeriFlux tower located in a 500 year old conifer forest in southern Washington State. CO2 and H2O fluxes have been measured continuously for six years using the eddy covariance method. The objectives of this study are to: 1. determine to what extent teleconnection patterns influence measured CO2 and H2O fluxes through mechanistic anomalies; 2. ascertain if climatic shifts affect annual vegetation canopy characteristics; and 3. make comparisons at the local and regional scales using MODIS. The ecosystem was a significant sink of carbon (-207 gC m-2 year-1) in 1999 but turned into a large carbon source (+ 100 gC m-2 year-1) in 2003. NEE significantly (above the 95th CI) lags the PNA, ENSO and PDO indicating that these patterns affect the forest carbon budget across overlapping time scales. To ascertain the influence of atmospheric patterns on fluxes, we categorized the flux measurement years based on in-phase climate events (1999 = La Niña/cool PDO, 2003 = El Niño/warm PDO, 2000-2002, 2004 = neutral ENSO years). The results of this analysis indicate that the Pacific Ocean/atmospheric oscillation anomalies explain much of variance in annual NEE (R2 = 0.78 between NEE and the PDO, R2 = 0.87 for the PNA, and R2 = 0.56 for ENSO). Teleconnection patterns are found to be associated mostly with air temperature, precipitation, and incoming light radiation (cloudy vs. sunny conditions). Important meteorological driving mechanisms of fluxes include: water- use efficiency (WUE), light-use efficiency (LUE) and canopy structure parameters (e.g., fPAR). Tower-based fPAR was strongly related to NEE (R2 = 0.78) and climatic patterns (R2 = 0.84 with ENSO and R2 = 0.76 with PDO). Variability in fluxes may be a result of changes in the canopy structural characteristics; for example higher, fPAR (e.g., 2003) correlated well with increased respiration fluxes. MODIS data (200 km X 200 km area) were obtained to determine if anomalies in vegetation indices and canopy structure could be linked to teleconnection patterns at the site level and across the region. The MODIS-derived Enhanced Vegetation Index (EVI) correlated well with yearly cumulative NEE at the tower and regional EVI anomalies were strongly negatively correlated with the annual PDO index (R2 = 0.9). MODIS-derived fPAR product correlated with yearly variability in the PDO (R2 = 0.34) at the site level. Therefore, there is reasonable expectation that structural changes, as a result of climate variability during strongly positive or negative teleconnection patterns, will be observed in other parts of the Pacific Northwest. MODIS data is useful for identifying the effects of teleconnections across a regional scale.
B43D-1588
Tree Age, Disturbance History, and Carbon Stocks and Fluxes in Subalpine Rocky Mountain Forests
Forest carbon stocks and fluxes in vary with forest age, and relationships with forest age are often used to estimate fluxes for regional or national carbon inventories. Two methods are used to estimate forest age: observed tree age or time since a known disturbance. To clarify the relationships between tree age, time since disturbance and forest carbon storage and cycling, we examined stands of known disturbance history in three landscapes of the southern Rocky Mountains. Our objectives were to assess the similarity between carbon stocks and fluxes for these three landscapes that differed in climate and disturbance history, characterize the relationship between observed tree age and time since disturbance and quantify the predictive capability of tree age or time since disturbance on carbon stocks and fluxes. Carbon pools and fluxes were remarkably similar across the three landscapes, despite differences in elevation, climate, species composition, disturbance history, and forest age. Tree age was a poor predictor of time since disturbance. Maximum tree age overestimated age since disturbance for young forests and overestimated it for older forests. Carbon pools and fluxes were related to both tree age and disturbance history, but the relationships differed between these two predictors and were generally less variable for pools than for fluxes. Using tree age in a relationship developed with time since disturbance or vice versa increases errors in estimates of carbon stocks or fluxes. Little change in most carbon stocks and fluxes occurs after a tree age of 60 years, simplifying landscape scale estimates. We conclude that subalpine forests in the Central Rocky Mountains can be treated as a single forest type for the purpose of assessment and modeling of carbon, and that the critical period for change in carbon is < 60-100 years.
B43D-1589
The Effect of Fire on Soil Respiration Rates in Siberia Scotch Pine Forest
Russian boreal forests contain approximately 25% of the world's terrestrial biomass storage, and have an annual wildfire activity averaging 10 to 15 million ha, that has been increasing in recent years. Wildfire activity, in response to changing climate has the potential to significantly affect the carbon storage capacity of Siberian forests. Experimental fires were conducted on a total of 13 plots (4 ha) at 3 Scotch pine sites in central Siberia from 2000 to 2003. Fire intensities on the plots ranged from low to medium with one high fire intensity plot. Soil respiration was measured immediately before and after burning, and has been measured annually on the burn plots and controls for the last five years through 2007. A systematic grid point sampling scheme was employed on each plot. The soil respiration rate was reduced by an average of 71% one day after burning for all plots measured. Regressions of soil respiration vs. fireline intensity, and other measures of fire severity had r2 values of .6 to .9. These relationships are tracked as the burns age and soil respiration recovers. On one of the sites, after 3 years soil respiration rates of the burn plots had recovered to approximately 60% of the adjacent control plot rates. Soil respiration rates in mid- summer are presented for Siberia Scotch pine forest for control and 1 to 5 year old burn plots of different fire severity classes. These results are part of the Russian FIREBEAR (Fire Effects in the Boreal Eurasia Region) Project, which quantifies the impacts of fire severity on ecosystem processes, emissions, and the carbon cycle. The annual measurements of soil respiration will continue, to get a better understanding of how soil respiration rates are changing and how they recover after fire in Siberia Scotch Pine forests.
B43D-1590
Two decades of forest disturbance and regrowth across the United States evaluated using the Landsat record for the North American Carbon Program
Forest disturbance and regrowth are assumed to be significant forces modulating North American carbon balance. Quantifying the carbon fluxes of forest changes requires the changes be assessed with appropriate spatial and temporal details. Landsat imagery accumulated since 1972 provides a unique data source for evaluating these processes over the last 30+ years. Through a NASA funded project -- "North American Forest Disturbance and Regrowth since 1972 (NAFD)", Landsat time series stacks (LTSS) have been assembled for 29 locations selected to represent United States dynamics. For each selected location, the LTSS is being used to map forest disturbance and regrowth with a nominal temporal interval of 2 years from 1972 to 2005. In this talk we will present the results we have derived for the TM/ETM+ era of the Landsat missions – from 1984 to 2005. Specifically, we will provide an assessment of the spatial and temporal characteristics of the changes mapped for each site. We will also discuss the regional variations in patterns derived from these sites. Results from this study will provide a basis for improved quantification of carbon fluxes arising from forest disturbance and regrowth across the United States. The derived forest change products will most likely also be highly valuable for forest resources management, ecosystem monitoring, and ecological applications.
B43D-1591
Accuracy and Precision in Measurements of Biomass Oxidative Ratio and Carbon Oxidation State
Ecosystem oxidative ratio (OR) is a critical parameter in the apportionment of anthropogenic CO2 between the terrestrial biosphere and ocean carbon reservoirs. OR is the ratio of O2 to CO2 in gas exchange fluxes between the terrestrial biosphere and atmosphere. Ecosystem OR is linearly related to biomass carbon oxidation state (Cox), a fundamental property of the earth system describing the bonding environment of carbon in molecules. Cox can range from -4 to +4 (CH4 to CO2). Variations in both Cox and OR are driven by photosynthesis, respiration, and decomposition. We are developing several techniques to accurately measure variations in ecosystem Cox and OR; these include elemental analysis, bomb calorimetry, and 13C nuclear magnetic resonance spectroscopy. A previous study, comparing the accuracy and precision of elemental analysis versus bomb calorimetry for pure chemicals, showed that elemental analysis-based measurements are more accurate, while calorimetry- based measurements yield more precise data. However, the limited biochemical range of natural samples makes it possible that calorimetry may ultimately prove most accurate, as well as most cost-effective. Here we examine more closely the accuracy of Cox and OR values generated by calorimetry on a large set of natural biomass samples collected from the Kellogg Biological Station-Long Term Ecological Research (KBS-LTER) site in Michigan.
B43D-1592
The Second "Ring of Towers": Over-sampling the Mid Continent Intensive region CO2 mixing ratio?
A central barrier preventing the scientific community from understanding the carbon balance of the continent is methodological; it is technically difficult to bridge the gap in spatial scales that exists between the detailed understanding of ecological processes that can be gathered via intensive local field study, and the overarching but mechanistically poor understanding of the global carbon cycle that is gained by analyzing the atmospheric CO2 budget. The NACP's Midcontinental Intensive (MCI) study seeks to bridge this gap by conducting a rigorous methodological test of our ability to measure the terrestrial carbon balance of the upper Midwest. A critical need in bridging this gap is increased data density. A primary goal of the project is to increase the regional atmospheric CO2 data density so that 1) atmospheric inversions can derive well-constrained regional ecosystem carbon flux estimates and 2) the trade off between data density and accuracy of the flux estimates can be determined quantitatively using field observations, thus providing guidance to future observational network designs. Our work adds a regional network of five communications-tower based atmospheric CO2 observations to the planned long-term atmospheric CO2 observing network (tall towers, flux towers and aircraft profiles) in the midcontinent intensive region. Measurements began in April–June 2007, If the measurements are shown to be spatially dense enough to over sample the CO2 mixing ratio, the experiment will provide an upper bounds on the density of measurements required to produce the most accurate flux possible with current atmospheric inversions. The five sites for "Ring 2" and deployment dates are Centerville, IA (Apr 07), Round Lake, MN (May 07), Kewanee, IL (Apr 07), Mead, NE (Apr 07), Galesville, WI (June 07). Two heights are sampled at each tower (30 m AGL and between 110 and 140 m AGL). More details are available at www.ring2.psu.edu. In addition, two systems in PSU's network of well-calibrated CO2 mixing ratio measurements deployed at Ameriflux towers are within the midcontinental region: Ozark, MO (30 m AGL) and Mead, NE (3-6 m AGL) (www.amerifluxco2.psu.edu). The instruments chosen for the Ring 2 deployment are Picarro Inc., Cavity Ring-Down Spectroscopy (CRDS) instruments. One advantage of the CRDS instruments is the reduced need for calibration compared to the systems used in PSU's Ameriflux CO2 network which are calibrated every four hours using four calibration tanks. Although the long-term stability is not exactly known, tests have shown accuracy to within 0.2 ppm on a monthly time scale without additional calibration. Preliminary results show spatial differences in daytime CO2 across the ring that are as large as 40 ppm, and highly variable in time. We will present observations and preliminary interpretation of these data. http://www.ring2.psu.edu
B43D-1593
Using narrowband vegetation indices to estimate ecosystem-atmosphere CO2 flux for a mixed hardwood forest
The productivity of terrestrial ecosystems is a fundamental component in the global carbon cycle. While satellite data are increasingly relied upon to derive information on earth surface characteristics that affect biosphere- atmosphere interactions, such as carbon cycling, there remain substantial uncertainties in the estimates of vegetation productivity and related processes derived from these sensors. In order to reduce these uncertainties data are needed to better understand the relationship between narrowband reflectance of shortwave radiation by vegetation canopies and carbon exchange between these canopies and the atmosphere. We mounted a narrowband spectroradiometer approximately 26 m above a mixed forest canopy to quantify midday canopy reflectance while simultaneously measuring CO2 and radiation fluxes at the University of Michigan Biological Station Ameriflux site throughout the 2006 and 2007 growing seasons. Narrowband spectral vegetation indices (SVIs) were used as proxies for canopy structure and physiological function and regressed against flux measurements to assess their efficacy for predicting CO2 fluxes. We found that certain narrowband SVIs were well correlated with factors such as canopy light absorption (r2 = 0.97, p<0.0001) and light use efficiency (r2 = 0.73, p<0.0001). Furthermore, when these SVIs were combined in a simple model to predict gross carbon uptake, they explained much of the variability in midday Net Ecosystem Exchange (r2 = 0.87, p<0.0001). These results show that optical signals from remote platforms may be useful for scaling carbon exchange of mixed temperate forests to scales beyond that of an individual tower flux footprint.
B43D-1594
Forest Structure Estimation and Pattern Exploration From Discrete Return Lidar in Subalpine Forests of the Central Rockies
Discrete return lidar has been used to accurately measure and characterize forest structure across a range of forest types, with canopy surface height and canopy profile indices used as explanatory variables in regression analysis. This study evaluates the ability of discrete lidar to estimate forest structure and forest biomass variables using both traditional lidar indices (i.e. mean height, max height, height percentiles, etc.) and statistically derived canonical correlation analysis (CCA) variables across three temperate subalpine forest sites in the Central Rockies. Modeling results with both lidar and CCA explanatory variables performed well with lidar models consistently having slightly higher explained variance, and a lower ratio of mean predicted value, relative to models derived with CCA variables. Adjusted R2 values for mean height, sum of leaf area and all carbon in live biomass were (0.93, 0.93), (0.74, 0.73) and (0.93 and 0.85) for the lidar and CCA explanatory regression models respectively. Investigation of forest complexity patterns using graphs of forest variable correlations with lidar canonicals one and two revealed distinct forest structure clusters within ordination space. Canonical one is highly correlated with forest height, biomass, and total leaf area, and canonical two is highly correlated with tree density. When canonicals one and two are considered in conjunction they represent a continuum of stand age and structure from young to mature forest. The lidar derived biomass estimates will be utilized in the US Forest Service Northern Global Change Research Program, where the extensive lidar derived biomass estimates will be compared with coincident intensive flux tower biomass estimates.
B43D-1595
Regional Carbon Fluxes and Atmospheric Carbon Dynamics in the Southern Great Plains during the 2007 CLASIC intensive
In June 2007, a regional campaign took place in the Southern Great Plains (SGP) to estimate land-atmosphere exchanges of CO2, water, and energy at 1 to 100 km scales. The primary goals of this campaign were to evaluate top-down and bottom-up estimates of regional fluxes and to understand the influence of moisture gradients, surface heterogeneity, and atmospheric transport patterns on these fluxes (and their estimation). The work was integrated with the Cloud and Land Surface Interaction Campaign (CLASIC), centered on the US DOE Atmospheric Radiation Measurement (ARM) Program SGP region. CO2 concentration data were collected from tower and airborne platforms. Eddy flux towers were deployed in the four major land cover types, distributed over the region's SE to NW precipitation gradient. In addition, CO2, water, and energy fluxes were observed with the Duke Helicopter Observation Platform (HOP) at various heights in the boundary layer, including in the surface layer (the few meters near the surface). One aircraft carried precise CO2, CO, and CH4 continuous measurement systems, and 14C, radon, and NOAA 12-flask (carbon cycle gases and isotopes) packages. Continuous CO2, CO, and radon concentrations, NOAA 2-flask package, and isotope diel flasks (14C, 13C, and 18O) were also collected from a centrally located 60 m tower. Flights were planned to constrain simple boundary layer budget models and to conduct Lagrangian air mass following experiments. We present these data in the context of characterizing surface carbon exchanges via bottom-up and top-down approaches. We also describe results from forward (using MM5-LSM) and inverse (using STILT) modeling to estimate regional surface carbon and energy fluxes. In addition to characterizing the influence of the land surface on the atmosphere, the aircraft data (in combination with observations of atmospheric dynamics) provides a very well characterized southern boundary condition to the NACP Mid-Continent Intensive. http://esd.lbl.gov/ARMCarbon/
B43D-1596
Annual CO2 flux Across Moisture and Vegetation Gradients in the Amazon.
The Amazon region is important to global climate both due to its large size and the large fluxes of energy, moisture and carbon exchanged there between the atmosphere and terrestrial biosphere. The spatial extent and large flux magnitude in the region firmly couple both the global circulation of the atmosphere and carbon flux to the region. Landsurface models have a poor record for simulating the annual cycle of exchange of carbon in the Amazon basin. In general, models have predicted uptake of carbon during the wet season and efflux during dry months, while observations show the opposite. Observational research has suggested a number of biogeophysical mechanisms to explain all or part of this annual cycle, including: 1) increased canopy phenological response to increased insolation during dry season 2) ability of deep roots to maintain transpiration as surface soil dries 3) Hydraulic redistribution of soil moisture by roots to bring deep water to surface. Using the Simple Biosphere model we were able to show that a control simulation has an erroneous response consistent with previous modeling efforts. We were also able to show that, when the above mechanisms were included in the model physics, a much more realistic simulation of the annual cycle of CO2 flux (NEE) at the Tapajos River km83 site was possible. We now conduct a regional analysis across 8 tower sites throughout Amazonia across multiple climactic regimes. Annual precipitation ranges from less than 1000 mm to more than 2000 mm, and length of dry season (defined as a month with <100 mm precipitation) ranges from none to 6 months or more. Vegetation types consist of tropical forest, pasture, and savanna. We find that by including the aforementioned mechanisms into the model we remove stress and obtain more reasonable annual cycles of NEE in the forest sites, but in the pasture and savanna sites the modified model removes vegetation stress in an unrealistic manner. Regional simulations require realistic input information on soil and rooting depth to perform realistically when confronted with data. When we include these data in the simulations the model carbon flux is comparable to observed across moisture and vegetation gradients in Amazonia.
B43D-1597
Identifying Temporal Patterns in Light use Efficiency for two Loblolly Pine Plantations in a Drained Lower Coastal Plain Region of North Carolina, U.S.A.
Light Use Efficiency (LUE) is a variable present in most ecosystem models driven by remote sensing. Among other factors, LUE varies with time. In this study we evaluate the temporal variation of LUE over a one year period for two loblolly pine plantations – one mid-rotation and one recently harvested. Specifically, we determine the most reasonable measurement time periods for LUE, and if these periods vary with stand age. The underlying hypothesis is that short term temporal changes in LUE cancel out over a certain amount of time, and therefore, to estimate forest productivity at landscape to regional scales using moderate resolution satellite data these intensive measurements are unnecessary. To test this hypothesis we use data collected in two micrometeorological tower sites that are a part of the Ameriflux network. They are located in the coastal plain region of North Carolina, U.S.A and are less than five kilometers apart. For this study eddy covariance measurements and photosynthetically active radiation (PAR) sensors are used to obtain gross primary production (GPP) and the fraction of incident photosynthetically active radiation absorbed by the canopy (ƒAPAR). LUE is calculated as GPP divided by fAPAR. The analysis of the data consists of first calculating daily LUE averages for the entire study period. Changes in both the trend and variance of LUE are being assessed using autoregressive conditional techniques for time series analysis.
B43D-1598
Low-cost sensor packages for measuring narrowband canopy reflectance: implications for modeling regional C exchange
Recent advances in understanding the relationships between spectral reflectance of vegetation canopies and the structural/physiological drivers of canopy-atmosphere net ecosystem carbon exchange highlight the potential for using narrowband spectral vegetation indices to spatially scale C fluxes beyond the area of a tower footprint. However, ground reference observations of narrowband spectral reflectance in support of satellite observations can be challenging to obtain because 1) automated sampling of both upwelling and downwelling radiation is required over extended time periods to characterize diurnal and seasonal variability, 2) the quality of hyperspectral spectroradiometer data can be sensitive to environmental factors such as temperature and humidity, and 3) hyperspectral spectroradiometers are quite expensive, therefore greatly limiting prospects for widespread sampling. We have therefore developed lightweight (<0.5 kg), relatively low cost (ca. $325) sensor packages capable of measuring upwelling and downwelling radiation in 10nm-wide wavebands centered at 532 nm, 568 nm, 680 nm, and 800 nm, as well as the full PAR range (400-700 nm). These measurements can be combined to calculate several spectral characteristics (e.g. the photochemical reflectance index, PRI; the normalized difference vegetation index, NDVI; green NDVI, and the fraction PAR absorbed by the canopy, fPAR) found useful in modeling canopy trace gas exchange. Here, we discuss data collected using these sensor packages at shrub-steppe (Idaho, USA), mixed hardwood forest (Michigan, USA), and tropical rainforest (French Guiana) ecosystems and their implication for improving models of regional-scale carbon dioxide fluxes.
B43D-1599
On Using CO2 Concentration Measurements at Mountain top and Valley Locations in Regional Flux Studies.
Data from the Regional Atmospheric Continuous CO2 Network in the Rocky Mountains (Rocky RACCOON) are used to investigate atmospheric controls on temporal and spatial variability of CO2 in mountainous terrain and the usefulness of mountain top and valley measurement for the estimation of regional CO2 fluxes. Rocky RACCOON consists of four sites installed in fall of 2005 and spring of 2006: Niwot Ridge, near Ward, Colorado; Storm Peak Laboratory near Steamboat Springs, Colorado; Fraser Experimental Forest, near Fraser Colorado; and Hidden Peak, near Snowbird, Utah. The network uses the NCAR-developed Autonomous Inexpensive Robust CO2 Analyzer. These units measure CO2 concentrations at three levels on a tower, producing individual measurements every 2.5 minutes precise to 0.1 ppm CO2 and closely tied to the WMO CO2 scale. Three of the sites are located on a mountain top while one site is located in a valley. Initial analyses show interesting relationships between CO2 concentration and atmospheric parameters, such as wind speed and direction, temperature, and incoming solar radiation. The nature of these relationships is further investigated with an atmospheric mesoscale model. Idealized and realistic simulations are able to capture the observed behavior of spatial and temporal CO2 variability and reveal the responsible physical processes. The implications of the results and the value of the measurements for providing information on local to regional scale respiration and photosynthesis rates in the Rockies are discussed.
B43D-1600
Remote measurement of photosynthetic efficiency using laser induced fluorescence transient (LIFT) technique.
An understanding of spatial and temporal diversity of photosynthetic processes, water and energy exchange of complex plant canopies is essential for carbon and climate models. Remote sensing from space or aircraft platforms provides the only practical way to characterize the vast extent of plant canopies around the globe, but the basis for relating physiological processes to remote sensing is still largely theoretical. Experiments that bridge this gap are needed. Chlorophyll fluorescence measurements have been widely applied to quantify photosynthetic efficiency and non- photochemical energy dissipation non-destructively in photosynthetically active organisms. The most commonly used Pulse Amplitude Modulated (PAM) technique provides a saturating light pulse and is not practical at the canopy scale. We report here on a recently developed technique, Laser Induced Fluorescence Transient (LIFT), capable of remote measurement of photosynthetic efficiency of selected leaves at a distance of up to 50 m and we present here continuous studies on plans growing under natural conditions during the beginning of the winter season and the onset of summer drought in this Mediterranean climate. i) Lichens showed a strong diurnal variation in photosynthetic efficiency which correlated with relative humidity; ii) Photosynthetic efficiency of annual grass decreased with progressing drought stress; iii) An oak canopy showed very little variation of quantum yield from leaf out in spring to summer; iv) The combined effect of low temperature and high light intensity during an early winter strongly reduced the photosynthetic efficiency of four different species in response to chilling stress. These measures with the LIFT correlated well with (more limited) sampling by PAM fluoromentry and gas exchange. The ability to make continuous, automatic and remote measurements of photosynthetic efficiency of leaves with the LIFT provides a new approach for studying the heterogeneity of photosynthetic efficiency within canopies and for integrating these effects from the leaf to the canopy scale.
B43D-1601
Carbon Distribution and Net Primary Production in a Forest-Peatland Landscape Mosaic
We characterized the distribution of carbon and annual NPP in a mixed forest and peatland landscape in the Marcell Experimental Forest in northern Minnesota, USA. We estimated vegetation biomass and production (aboveground and belowground) and the carbon content of detrital pools (forest floor, woody debris, and mineral soil or peat) in a 1-km2 area that encompassed multiple vegetation cover types, including forested uplands dominated by aspen, mixed-hardwood, or pine, and peatlands dominated by alder, conifers, or ericaceous shrubs and Sphagnum mosses (open peatlands). In aspen dominated areas, which account for >70% of our study area, total C storage was 164 ± 9 Mg C ha-1 while pine and hardwood dominated averaged 190 ± 16 and 153 ± 19 Mg C ha-1 respectively. Total ecosystem carbon content in peatland areas averaged 1380 ± 170 Mg C ha-1 and was highly dependent upon peat depth. Among upland cover types, NPP was greatest in pine-dominated areas (6.2 ± 0.6 Mg C ha-1) and similar in aspen- and hardwood- dominated areas (4.7 ± 0.2 and 4.5 ± 0.5 Mg C ha-1, respectively). We found considerable variability in NPP among peatland cover types; in coniferous peatlands, alder peatlands, and open peatlands, NPP was 6.0, 2.8 ± 0.4 and 1.4 ± 0.2 Mg C ha-1 respectively. Large differences in NPP among peatland cover types as well as smaller but significant differences between deciduous and coniferous upland cover types illustrate the importance of these cover types when scaling carbon cycling to landscape and regional levels.
B43D-1602
Ground-based Remote Sensing of Carbon Trace Gases: Validation of AERI CO Retrievals and Validation of Satellite CO Observations
Global to regional satellite measurements of CO and other trace gases in the boundary layer can provide essential constraints for air quality forecasting and climate change science. We present validation of a carbon monoxide (CO) retrieval algorithm developed for zenith-viewing Atmospheric Emitted Radiance Interferometers (AERI) using co-located in situ measurements at the United States Department of Energy (DOE) Southern Great Plains (SGP) Atmospheric Radiation Measurements (ARM) site near Lamont, Oklahoma. The AERI measures downwelling radiation emitted from the atmosphere every 7-8 minutes, providing near real-time temperature and moisture profiling of the boundary layer since January, 1997. AERI CO averaging kernels indicate approximately 70% of the retrieved information comes from the boundary layer, with rapidly decreasing sensitivity to the free troposphere. We report a comparison of AERI CO retrievals with a combination of ground-based continuous measurements of CO mixing ratios at 60 m above the surface and weekly aircraft flask profiles (from 300 to 5000 m above ground level) at the SGP site between March, 2006 and April, 2007. Results show that AERI retrievals are well correlated with measured CO. A subset of the ten-year record of AERI CO retrievals are then compared with retrievals of lower tropospheric CO obtained from satellite-borne instruments including the Measurement Of Pollution In The Troposphere (MOPITT) instrument onboard Terra, the Atmospheric InfraRed Sounder (AIRS) onboard Aqua, and the Tropospheric Emission Spectrometer (TES) onboard Aura. Results from this comparison show that the satellite measurements successfully retrieve the column CO over the SGP.
B43D-1603
Fluxes by eddy correlation over heterogeneous landscape: How shall we apply the Reynolds average?
Top-down estimates of carbon exchange across the earth's surface are implicitly an integral scheme, deriving bulk exchanges over large areas. Bottom-up estimates explicitly integrate the individual components of exchange to derive a bulk value. If these approaches are to be properly compared, their estimates should represent the same quantity. Over heterogeneous landscape, eddy-covariance flux computations from towers or aircraft intended for comparison with top-down approach face a question of the proper definition of the mean or base state, the departures from which yield the fluxes by Reynolds averaging. 1)≠Use a global base state derived over a representative sample of the surface, insensitive to land use. The departure quantities then fail to sum to zero over any subsample representing an individual surface type, violating Reynolds criteria. Yet fluxes derived from such subsamples can be directly composed into a bulk flux, globally satisfying Reynolds criteria. 2)≠Use a different base state for each surface type. satisfying Reynolds criteria individually. Then some of the flux may get missed if a surface's characteristics significantly bias its base state. Base state≠(2) is natural for tower samples. Base state≠(1) is natural for airborne samples over heterogeneous landscape, especially in patches smaller than an appropriate averaging length. It appears (1) incorporates a more realistic sample of the flux, though desirably there would be no practical difference between the two schemes. The schemes are related by the expression w̄*a*)C - w̄'ā')C = w̄'ã̄)C+&wtilde;̄ā')C+ &wtilde;̄ã̄)C Here w is vertical motion, and a is some scalar, such as CO2. The star denotes departure from the global base state≠(1), and the prime from the base state≠(2), defined only over surface class≠C. The overbar with round bracket denotes average over samples drawn from class≠C, determined by footprint model. Thus ā')C = 0 but ā*)C ≠ 0 in general. The tilde denotes the departure of base-state≠(2) from base-state≠(1). It represents surface≠C's characteristic bias. The equation is defined only over class≠C. A similar equation applies to each surface class. The first and second righthand terms express interaction of the departure quantities with surface≠C's characteristic bias. These terms are zero if the base states are simple means. The third term becomes important if class C has a significant bias both in vertical motion and in its characteristic values of a. A practical example from 2005 June 18 at 1015 LST in Illinois is illustrative. Turbulence measurements were made by aircraft at 20≠m above ground along a 50≠km track approximately evenly divided between corn and soybean. Corn (type≠C) was growing quickly, increasing the mixing ratio of moisture (r) and reducing that of CO2 (a), relative to soybean. Soybean characteristically heated the air and favored updrafts. These biases were evident in r̄*)C, ā*)C, θ̄*)C, and w̄*)C relative to their corresponding averages over soybean. In particular the bias in CO2 mixing ratio, negative over corn and positive over soybean, was about 20% of the standard deviation of a*. Nevertheless, neither surface type strongly favored vertical motion, giving the encouraging result that the two approaches do not differ by more than an insignificant few per cent. The theoretical analysis indicates care, however, where extensive areas of both bare soil and vegetated land may enhance the bias in vertical motion between different components of the landscape.
B43D-1604
Does Western US forest thickening increase carbon stored in aboveground biomass?
Forest thickening in Western United States forests, which is characterized by an increase in stem density, may have arisen from alterations in historic fire regimes brought about by fire suppression. This has led to a shift in forest structure and demography from a comparatively sparse forest composed of a few large trees to one composed of a higher number of small trees. This study compares plots from the contemporary Forest Inventory Analysis dataset with the historic Weislander Vegetation Type Mapping dataset in designated California wilderness areas to quantitatively ascertain the change in carbon stored in aboveground live biomass due to forest thickening and associated demographic shifts for California's forests over a 60 year period. There was no significant change in stand density when averaged over all forest types. However, live aboveground carbon pools decreased by nearly 50 Mg C/ha, which corresponds to an average loss of 0.8 Mg C/ha/year. Mid-elevation forests showed a likely thickening with a reduction in live aboveground carbon. This research helps refine carbon budgets for Western North America by improving our understanding of forest thickening as an atmospheric carbon sink.
B43D-1605
Carbon Consequences of Positive NDVI Anomalies in North America
We present a modeling approach to investigate the carbon consequences of ecosystem disturbance and land cover dynamics in regions of increasing NDVI in North America, which have altered terrestrial patterns of net ecosystem productivity from 1982-2005. Terrestrial carbon fixation and respiration is substantially affected by components of global change (e.g. land cover land use change, and changes in climate variables e.g. warming and drying). The Carnigie-Ames-Stanford approach (CASA) Biosphere model was run on a monthly time interval to simulate seasonal patterns in net plant carbon fixation, biomass and nutrient allocation, litterfall, soil nitrogen mineralization, and microbial CO2 production. CASA is a bucket type model, which allocates carbon between pools based on logarithmic scalars derived from in situ studies of terrestrial mechanistic processes. Carbon was reallocated between pools based on anthropogenic and abiotic disturbances identified with ancillary and remote sensing data (Landsat, Ikonos, aerial photography, agriculture production statistics, and fire and logging data). We developed modules for CASA and simulated expansion of irrigated agriculture, logging and subsequent recovery, warming in the northern latitudes, and fire with subsequent recovery. Simulations with modules initiated predicted altered carbon allocation in ecosystems with disturbance, and sequestration increased in some of our selected study sites during the 24 year (1982-2005) period while others experienced increased soil respiration. The regions investigated represent spatially complex transient pools of carbon in North America vegetation altered by abiotic and anthropogenic land cover conversions which contributed to the terrestrial Northern Hemisphere sink.
B43D-1606
Controls on DOC Export Through Time Along the Eastern Coast of the U.S.
The riverine transport of carbon is an important component of the net terrestrial and oceanic carbon cycles. However, the inclusion of riverine fluxes in local, regional and global carbon budgets is only performed at rough scales. As a preliminary step towards the development of a spatially explicit map of the lateral export of DOC from terrestrial landscapes to coastal margins, we investigated the controls on hydrology, DOC concentration and DOC flux along the eastern coast of the US. The effects of changes in temperature, precipitation, and land-cover on patterns of changing organic carbon concentrations across small and large watersheds within the US will be discussed.