B33E-1656
FLUXNET: Data from a Global Network of Eddy-Covariance Flux Towers
FLUXNET is an internationally coordinated global network of long-term micrometeorological flux measurement sites that focus on measuring and interpreting the land-atmosphere exchanges of carbon dioxide, water vapor, and energy. The FLUXNET Data and Information System compiles flux data and related site characteristics data from regional flux networks from around the world for modeling the carbon cycle, analyzing the analysis of the controls on carbon, water, and energy fluxes, and validating remote sensing products. Flux data are being used to evaluate ecosystem model outputs and to validate remote sensing products, such as the photosynthesis product derived from the MODIS sensor on the Terra and Aqua satellites. Comparison of information from multiple types of local and regional studies is needed for understanding the dynamics of ecosystem-atmosphere carbon dioxide exchange and for extrapolating site studies to larger-scale products derived from remote sensing or global-scale modeling. We will demonstrate data products and tools for facilitating acquisition and intercomparison of the flux data, model output, and remote sensing data. http://www.fluxnet.ornl.gov/fluxnet/index.cfm
B33E-1657
Creating and Accessing the Global Fluxnet Data Set
The recently gathered FLUXNET synthesis dataset contains on the order of 900 site years from over 260 sites. The size of this dataset makes browsing of the data difficult for users without additional help. For instance, a search of the dataset for sites with particular meteorological or flux characteristics would require a download of the complete dataset and then running all of the data through a preliminary analysis. Instead we have developed a Scientific Data Server which enables browsing of the data on-line and then download of only the data needed for an analysis. The Scientific Data Server leverages modern database technology and stores the data in a database. This server allows individual researchers to concentrate on science rather than data management. We leverage database tools such as data cubes and web reports to enable Excel pivot table and browser access to the data. It is our belief that by using these tools a researcher can quickly and easily evaluate the quality and availability of the data and identify sites able to support a specific analysis. In addition, we have leveraged available collaboration technology to incorporate support for contact between site PIs and researchers hoping to use their data. In this talk we will give a brief introduction to the data server and its available features. http://www.fluxdata.org
B33E-1658
Mechanisms responsible for lags in the development of photosynthetic capacity of deciduous forests in the spring and their implications for the TG model of gross photosynthesis.
We have recently developed a new Temperature and Greenness (TG) model for estimation of GPP at local to global scales. This model is based entirely on the enhanced vegetation index (EVI) and land surface temperature (LST) from MODIS and consequently outputs are obtained on a per pixel basis. In most cases, the predictions of the TG model have been as good as or better than the more complex MOD17 product from MODIS. However, there are still some areas where the TG model could be improved. For the deciduous forest sites, there is a lag in the development of tower measured GPP relative to modeled GPP. In this study, we examined the causes of this lag so that we can understand how to modify the TG model to better predict it. Specifically, we wished to know to what extent this lag is due to a lag in development of maximum light use efficiency (LUE) as opposed to a lag in development of leaf area? We used data from a wide range of eddy covariance flux towers across North America in combination with MODIS data for the 3 km region around each flux tower. Results suggest that both of these factors are important. In deciduous forest sites, increases in EVI, indicating expansion of leaf area, lag the increase in LST in the spring. But LUE also lags the increase in both EVI and LST, indicating that not only do leaves expand slowly in the spring but that their photosynthetic capacity is initially low. In contrast, evergreen forest sites show LUE values that are more constant during the growing season and have a more linear response to LST when low temperatures limit photosynthetic performance. Potential modifications to the TG model to account for these lags will also be discussed.
B33E-1659
Water-use efficiency at different scales and its between-site variability
The eddy covariance technique is widely applied for estimating the carbon and water exchanges between terrestrial ecosystems and the atmosphere, and also gross primary productivity can be reliably derived from such measurements. With such data we study intrinsic water-use efficiency (iWUE) at ecosystem scale that is the ratio between GPP and stomatal conductance which can be approximated as the ratio between GPP times VPD and evapotranspiration. Effects of LAI on iWUE in deciduous forests and grasslands during the growing season are detected. This effect is explained by combined impacts of light absorption on photosynthesis and bare soil evaporation. Increasing iWUE under drought conditions found at leaf level is partly confirmed at ecosystem scale at daily resolution, but the effect is small. At longer time scales, however, other ecosystem processes are suggested to override the impact of stomatal conductance on iWUE. This hypothesis is based on the pronounced between-site variability of iWUE (replacement of time by space), and on specific relationships of mean annual iWUE to environmental conditions. Besides the impact of maximum LAI, mean annual iWUE relates to both plant- available water-holding capacity of the soil (WHC) and mean annual soil moisture. Carbon isotope ratios independently confirm these findings. The first potential explanation of increasing iWUE with WHC, high nitrogen availability, is rejected by the anti-correlation between mean annual iWUE and leaf-nitrogen content. Rather, memory effects of droughts on maximum carbon assimilation are suggested to explain this observation. Mean annual iWUE of herbaceous ecosystems is lower than that of forests, and deciduous broad-leaved forests mostly show higher mean annual iWUE values than evergreen needle-leaved forests. Presented relationships of mean annual iWUE at ecosystem scale to different ecosystem properties can be further used to extrapolate mean annual iWUE in space for deriving global maps as the basis for further diagnosis and evaluation of ecosystem models.
B33E-1660
Towards a Land Surface Model Evaluation Framework
As a result of the increased availability of eddy-covariance flux data globally, the area of Land Surface Model (LSM) evaluation has an opportunity to move beyond calibrate-and-compare-with-observation perspective that usually constitutes evaluation. This presentation will briefly outline several innovative model evaluation techniques that demonstrate the increased depth of understanding projects such as Fluxnet can provide to the LSM community. Issues addressed include: efficiency of utilisation of information available to LSMs in predicting fluxes; the meaning of calibration in the context of systematic model bias, and; the appropriateness of vegetation type divisions.
B33E-1661
Nocturnal Respiration: Can we use the Temporal Dynamics of Advection to Derive an Alternative to the u*-Threshold Filtering Technique?
Even moderately complex topography can lead to significant horizontal and vertical advection and a consequent underestimation of nocturnal CO2 fluxes derived from eddy covariance measurements on a single tower. The presence of advective transport is identified using the time series of CO2 concentrations. Given a continuous input of carbon (soil-, wood- and leaf respiration) into a control volume (CV) we expect concentrations in that control volume to increase through time if there is no flux through the upper lid and no advective transport. Time series measurements usually show a strong increase in carbon concentration once the atmosphere is stably stratified, but after this initially strong increase the temporal build-up of concentrations is slower and may even remain relatively constant. As this steady-state situation is not associated with fluxes through the top of the CV we can assume that advection is draining CO2 away. We have analysed data from 24 Fluxnet sites where turbulent flux (Fc) and change of storage (Sc) of CO2 are routinely measured. Analysis of Sc confirms the observation by Aubinet et al. [1] that a common feature in canopy flows is that a maximum rate of change of CO2 occurs in the early evening. This suggests that after this maximum, the advection terms in the mass balance become non negligible. Based on this observation we have developed an alternative to the u*-threshold filtering technique [2]. We have shown that applying this alternative approach leads to good agreement between eddy covariance derived estimates of respiration and independent chamber measurements on a study site in a medium dense Eucalyptus forest in south-east Australia. In this contribution we discuss the applicability of this approach for several sites ranging from the tropics to subpolar latitudes and located in terrain that varies from almost ideally flat to very complex. We compare results from three different estimates of respiration and discuss differences between them. [1] Aubinet, M., Berbigier, P., Bernhofer, Ch., Cescatti, A., Feigenwinter, C., Granier, A., Gruenwald, Th., Havrankova, K., Heinesch, B., Longdoz. B., Marcolla, B., Monagnani, L. and Sedlak, P., 2005, Comparing CO2 storage and advection conditions at night at different Carboeuroflux sites. Boundary-Layer Meteorol., 116, 63-94. [2] van Gorsel, E., Leuning, R., Cleugh, H. A., Keith, H. and Suni, T. 2007. Nocturnal carbon efflux: reconciliation of eddy covariance and chamber measurements using an alternative to the u*-threshold filtering technique. Tellus, 59B, 397-403.
B33E-1662
Inverse modeling of parameters in a process-based model Biome-BGC at Duke Pine Forest
One important source of uncertainty in model prediction of climate change on terrestrial ecosystems is the uncertainty in model parameters. To quantify the uncertainty in model parameters, a Bayesian inverse modeling approach was used to estimate model parameter and the associated uncertainty in a process-based ecosystem biogeochemical model (Biome-BGC). Data from eddy covariance measurements at the Duke pine forest site were used to constrain the model parameters. Markov Chain Monte Carlo (MCMC) simulations using the M-H algorithm were conducted as the re-sampling method. Results showed that maximum stomatal conductance, fraction of nitrogen in Rubsico, and specific leaf area had large influences on carbon and water fluxes. The parameters could be relatively well constrained by the flux data and showed strong interannual variations among years. To further reduce the variation in ecosystem respiration parameters, soil and plant respiration measurements may be needed. The improved model parameters and the uncertainty could be used to quantify and reduce the uncertainty in model estimation and prediction.
B33E-1663
Deriving temporal and spatial variation in ecosystem parameters from FLUXNET data
To understand the global variation in carbon and water balances and to predict the ecosystem responses to climate changes it is important to identify the processes driving the differences and thus make progress beyond the simple regressions end empirical relationships that have been found. This study presents a method using the FLUXNET data and a simple ecosystem model to obtain five model parameters. Two parameters (reference respiration and activation energy) are related to ecosystem respiration and three parameters (carboxylation capacity, water use efficiency and light use efficiency) to photosynthesis and transpiration. The model is constrained by both the observed carbon and water fluxes, and determines the parameter uncertainty from the uncertainty in the observations. The main question is how the parameters are varying in time and space and if this can be related to environmental variations. The parameters are derived for four European forests. Annual parameter values are significantly different between sites and years, with no overlap in parameter space, taking into account the uncertainty. These site year parameters are very useful for comparing sites. The carboxylation capacity and light use efficiency are lowest for dry Mediterranean sites, while here the water use efficiency is highest. The light use efficiency is higher for the more northern sites. Adaptation to the environment may explain most of this, because sites are generally water limited in the south and light limited in the north of Europe. Whereas the parameter values appear meaningful when the model is applied on an annual basis, the correlation between observations and simulations improves when the model is applied on smaller time scales, but noise increases. The change in performance raises the interesting question how the different time scales are related. Improvement of the model could, for instance, be including this relation between time scales constrained by the environment.
B33E-1664
Quality control and improvement of eddy flux data for the CarboEurope-IP network
A site evaluation based on footprint modelling and a comparison of eddy-covariance software are two major components of QA/QC activities within the CarboEurope-IP network. We applied a site evaluation approach combining Lagrangian Stochastic footprint modelling with a quality assessment approach for eddy-covariance data to 25 forested sites of the CarboEurope-IP network. The analysis addresses the spatial representativeness of the flux measurements, instrumental effects on data quality, spatial patterns in the data quality, and the performance of the coordinate rotation method. Our findings demonstrate that application of a footprint filter could strengthen the CarboEurope-IP flux database, since only one third of the sites is situated in truly homogeneous terrain. Almost half of the sites experience a significant reduction in eddy-covariance data quality under certain conditions, though these effects are mostly constricted to a small portion of the dataset. Reductions in data quality of the sensible heat flux are mostly induced by characteristics of the surrounding terrain, while the latent heat flux is subject to instrumentation-related problems. The Planar-Fit coordinate rotation proved to be a reliable tool for the majority of the sites using only a single set of rotation angles. Overall, we found a high average data quality for the CarboEurope-IP network, with good representativeness of the measurement data for the specified target land cover types. Furthermore, the results from seven commonly used software packages for four five-day datasets from different sites were compared to assess the uncertainty of carbon dioxide flux estimates due to differences in post-processing. Data preparation, coordinate rotation and the implementation of the correction for high frequency spectral losses were identified as crucial processing steps leading to significant discrepancies in the carbon dioxide flux results. The overall comparison indicated a good although not yet perfect agreement among softwares within 5 to 10% difference for the 30-minute carbon dioxide flux values.
B33E-1665
New Approach to Estimate Daytime Ecosystem Respiration From Conventional Eddy Covariance Data Using Conditional Sampling Methods
Daytime respiration from tall-forested ecosystems remains among the least understood components in the total carbon balance. These forests pose unique challenges to respiration measurements because of the large number of respiring organs, their complex vertical distribution and their high spatial variability in forest floor carbon dioxide efflux. The definition of the net vertical flux (FN) as covariance between fluctuations of vertical velocity and perturbations in carbon dioxide density using the Eddy covariance (EC) technique does not lend itself to a formal decomposition into its 'ecologically desirable' component fluxes photosynthesis (FA) and respiration (Re) at daytime. Here, a new approach based on conditional sampling methods, quadrant analysis, and Relaxed Eddy Accumulation formulation is explored on carbon dioxide, water vapor, and vertical velocity time series to arrive at independent estimates of daytime Re directly from conventional EC measurements. The conceptual framework is based on the assumption that organized updrafts carry an unambiguous imprint of different scalar sinks and sources within the canopy volume. The new method is tested against datasets from 4 coniferous and 1 deciduous sites in North America and Europe (4 AMERIFLUX, 1 FLUXNET) most of them providing multi-level EC measurements. Results from comparisons of i) daytime Re against Re = FN at night, ii) temperature sensitivity coefficients (Q10) and base respiration rates (R10) derived from soil carbon dioxide efflux chambers and the new method, and iii) intercepts of light-response curves (FN versus shortwave down-welling radiation) to bulk respiration at daytime showed the effectiveness of the proposed method at 3 out of the 5 sites. Limitations were posed by the dense canopy of the deciduous site, and possibly by a summer drought reducing the coherence of scalar exchange in the carbon dioxide and water vapor signals at one coniferous site. A predictive skill indicator based on simple metrics of the canopy (i.e. drag coefficient, leaf area density, and canopy height) was explored to identify success or failure. The new method has the potential to become a powerful tool for flux network data analysis and syntheses, as it can be applied to both existing and current high-frequency EC measurements to independently constrain ecosystem daytime respiration in addition to other methods.
B33E-1666
Comprehensive comparison of gap filling techniques for eddy covariance net carbon fluxes
Review of fifteen techniques for estimating missing values of net ecosystem CO2 exchange (NEE) in eddy covariance time series and evaluation of their performance for different artificial gap scenarios based on a set of ten benchmark datasets from six forested sites in Europe. The goal of gap filling is the reproduction of the NEE time series and hence this present work focuses on estimating missing NEE values, not on editing or the removal of suspect values in these time series due to systematic errors in the measurements (e.g. nighttime flux, advection). The gap filling was examined by generating fifty secondary datasets with artificial gaps (ranging in length from single half-hours to twelve consecutive days) for each benchmark dataset and evaluating the performance with a variety of statistical metrics. The performance of the gap filling varied among sites and depended on the level of aggregation (native half- hourly time step versus daily), long gaps were more difficult to fill than short gaps, and differences among the techniques were more pronounced during the day than at night. The non-linear regression techniques (NLRs), the look-up table (LUT), marginal distribution sampling (MDS), and the semi-parametric model (SPM) generally showed good overall performance. The artificial neural network based techniques (ANNs) were generally, if only slightly, superior to the other techniques. The simple interpolation technique of mean diurnal variation (MDV) showed a moderate but consistent performance. Several sophisticated techniques, the dual unscented Kalman filter (UKF), the multiple imputation method (MIM), the terrestrial biosphere model (BETHY), but also one of the ANNs and one of the NLRs showed high biases which resulted in a low reliability of the annual sums, indicating that additional development might be needed. An uncertainty analysis comparing the estimated random error in the ten benchmark datasets with the artificial gap residuals suggested that the techniques are already at or very close to the noise limit of the measurements. Based on the techniques and site data examined here, the effect of gap filling on the annual sums of NEE is modest, with most techniques falling within a range of ±25 g C m-2 y-1.
B33E-1667
Characteristics of Sensible Heat, Water Vapor, and CO2 Fluxes Over a Rice Paddy
An eddy-correlation system consisting of a sonic anemometer and an open-gas analyzer was used for understanding the characteristics of sensible heat, water vapor, and CO2 fluxes over a subtropical rice paddy in Taipei, Taiwan. The results showed that about 35-40 percent of net radiation was used for latent heat flux, 13 percent for sensible heat flux, and the rest (about 50 percent) was absorbed by the water and soil in the rice paddy. Based on the background measurements (where no rice was growing), it was found that CO2 emission from the soil surface was small, just about 0.074 micro mole per square meter per second. We also found that the relative turbulent transport efficiencies of heat to water and heat to carbon dioxide depended on Bowen ratio. However, in average, heat, water vapor, and carbon dioxide were transported with the same rate above this rice paddy.
B33E-1668
Scaling approach of terrestrial carbon cycle over Alaska's black spruce forests: a synthesis of field observation, remote sensing, and ecosystem modeling
Spatio-temporal scale up of the eddy covariance data is an important challenge especially in the northern high latitude ecosystems, since continuous ground observations are rarely conducted. In this study, we measured the carbon fluxes at a black spruce forest in interior Alaska, and then scale up the eddy covariance data to spatio- temporal variations in regional carbon budget by using satellite remote sensing data and a process based ecosystem model, Biome-BGC. At point scale, both satellite-based empirical model and Biome-BGC could reproduce seasonal and interannual variations in GPP/RE/NEE. The magnitude of GPP/RE is also consistent among the models. However, spatial patterns in GPP/RE are something different among the models; high productivity in low elevation area is estimated by the satellite-based model whereas insignificant relationship is simulated by Biome-BGC. Long- term satellite records, AVHRR and MODIS, show the gradual decline of NDVI in Alaska's black spruce forests between 1981 and 2006, resulting in a general trend of decreasing GPP/RE for Alaskafs black spruce forests. These trends are consistent with the Biome-BGC simulation. The trend of carbon budget is also consistent among the models, where the carbon budget of black spruce forests did not significantly change in the period. The simulated results suggest that the carbon fluxes in black spruce forests could be more sensitive to water availability than air temperature.
B33E-1669
Estimating Climate-induced Variability in Carbon and Water Fluxes in the Upper Oyster River Watershed, British Columbia
Forest carbon budgets are commonly estimated using empirical approaches. Such models are not responsive to inter-annual variability in climate, limiting their ability to predict potential carbon cycle responses to climate change. In this study, we have revised a radiation-use efficiency model, Physiological Principals Predicting Growth (3-PG), to be compatible with the carbon pool structure and disturbance scheme of the Carbon Budget Model of the Canadian Forest Sector (CBM-CFS3) in order to investigate climate-induced variability in forest carbon cycling in British Columbia, Canada. The model was evaluated by comparing simulations with independent estimates of total aboveground biomass (AB), evapotranspiration (ET) and gross primary production (GPP) in stands of Douglas-fir within a 25 square km area within the upper Oyster River watershed on the east coast of Vancouver Island. The model was able to reproduce the salient features of variability in AB, including regeneration and overmature decline, as shown in growth and yield curves. Simulated annual ET was mainly driven by net radiation, but was also positively correlated with summer air temperature. Simulated annual GPP was positively correlated with annual average air temperature and soil water content, stemming from strong relationships during spring and summer, respectively. Inter-annual variability was assessed based on the standard deviation of annual total ET and GPP and compared with eddy covariance (EC) measurements, spanning 1998-2006. Inter-annual variability in simulated ET and GPP was 15 mm and 149 g C m-2, which closely matched EC-based estimates of 13 mm and 115 g C m-2. These results suggest that the model is a promising tool for further analysis of the effects of climate on long-term carbon cycling within the region and comparison with empirical estimates derived from CBM-CFS3.
B33E-1670
Modeling Carbon Flux from Inter-Tidal Salt Marshes
Coastal salt marshes have the potential to accumulate carbon at high rates over longer periods of time due to continuous accretion and burial of sediments rich in organic matter, giving soils in coastal wetlands a distinct advantage over many other environments with respect to the sequestration of organic carbon. Even though ecosystem level fluxes of carbon and energy have been studies in great detail from terrestrial (eg. FLUXNET) and oceanic environments (JGOFS - Joint Global Ocean Flux Study), considerable gaps exists in our understanding of ecosystem fluxes of carbon and energy from inter-tidal salt marshes that form the interface between terrestrial and oceanic ecosystems. The present study is based on observations from a flux tower based on eddy covariance methodology, set up in the lagoonal salt marsh in the Eastern shore of Virginia (37.41°N 75.83°W). The prominent vegetation in these mud flats is the salt marsh cod grass (Spartina alterniflora). The results presented include the trends in assimilatory response of the systems to various environmental forcings. The short nature of the vegetation along with a tidal range which can submerge the vegetation during different times of the day provided an interesting scenario for understanding tidal forcings on carbon flux. The study also addresses the development of a biophysical model for simulating the carbon and energy dynamics of the system by incorporating the theories of turbulent transfer (LNF theory) and by solving the energy balance equations within the different layers of the plant canopy. Spartina alterniflora is considered to be the primary sink for atmospheric carbon, and its distribution was modeled as a C4 photosynthetic mechanism. The model incorporates effects of tidal activity which influences the number of layers available for the model to operate and also the source/sink distribution. The study also addresses the possibility of including the air sea CO2 fluxes as a part of source/sink distribution for carbon dioxide for such inter-tidal ecosystems.
B33E-1671
Estimating ET for a Mixed Oak Forest Using Two Methods: Eddy-Covariance and the Soil Water Budget
We examined the evapotranspiration (ET) of a mixed Oak forest stand located within a woodland preserve in NW Ohio using two independent methods. The depth to ground water (DGW) in 2005 to 2007 ranged from 1.35 m at peak recharge (early to mid-May), to below 3.0 m during the driest period near the end of the growing season. The DGW during peak recharge was more shallow in 2007 than the other two years, and while the slopes of the ground water decline during the growing season were similar in 2005 and 2006 (-9 mm and -6 mm per day, respectively), the rate of decline was very steep in the first half of the 2007 growing season (-14 mm per day). Daily ET estimated from the eddy-covariance measurements (EC-ET) averaged 1.6 mm per day during 2005 to 2007. Growing season averaged EC-ET values (May-Oct) ranged from 0 to 5.32 mm per day, being highest in 2007 (2.5 mm per day), and lowest in 2005 (2.2 mm per day). The average growing season temperatures and relative humidity were also highest in 2007 when the highest precipitation and most shallow DGW values were recorded. Finally, weekly EC-ET values were highly correlated with the ET values estimated by the variation in DGW over the same 7 day period (GW-ET). This correlation was highest for 2005 (r=-0.77, p<0.05) and lowest for 2006 (r=-0.55, p<0.05). The correlations increased slightly when the 7 day sums of through-fall precipitation were added to the ground water sums. Soil water storage is included in the final GW-ET estimate, and lateral flows will be estimated from the change in water-table recession.
B33E-1672
Determination of Representative Canopy Height of Forests from Aerodynamic Data
The canopy height is the essential length scale for micrometeorological studies, especially for intersite comparison studies, but there has been a lack of consensus of how to determine it. We proposed the `aerodynamic canopy heightf ha to provide a universal method of determining representative canopy height, which was originally introduced by Maki (1975) and determined by simple linear regression between zero- plane displacement d and roughness length z0 (Thom, 1971), without the need for stand inventory data. The applicability of ha was confirmed in five different forests, including a forest with a complex canopy structure. The linear relationship between d and z0 was explained by assuming that the logarithmic wind profile above the canopy and the exponential wind profile within the canopy are continuous at canopy height, indicating that ha is essentially the same as that previously defined by the inflection point of the vertical wind profile (Thomas and Foken, 2007).
B33E-1673
Evaluation of Two New Models of Net Radiometers and Comparison to a Model to Predict Net Radiation
Net radiation is a key component to the surface energy balance, but it is difficult and expensive to measure accurately. Two new net radiometer models (Hukseflux NR01 and Kipp & Zonen CNR2) have been released in the past year. We evaluated and compared these models to two Kipp and Zonen model CNR1 net radiometers, and to two less expensive, older model net radiometers (Kipp & Zonen NR-Lite and REBS Q*7.1). Additionally, we predicted net radiation from solar radiation, air temperature, and absolute humidity measurements using a commonly used model that calculates net longwave radiation using a Brunt (1932; 1952) approach for predicting net emissivity. The model uses the ratio of measured solar radiation to predicted clear-sky solar radiation as a surrogate for cloud cover. Net shortwave radiation is determined by direct measurement of solar radiation and the albedo of the surface. Hourly averages and daily totals (over the course of the study; 33 days) from three replicate sensors of the two new net radiometers compared quite well to the CNR1 radiometers. The difference was generally less than +/- 5 %. Three replicates of the two older model net radiometers did not agree as well with the newer models, with differences generally less than +/- 15 %. Our data matched what others (Cobos and Baker, 2003; Brotzge and Duchon, 2000) have shown for these older radiometers. The net radiation model yielded hourly average and daily total values that were 10-15 % higher than the CNR1 radiometers. Our findings indicate that accuracy increases with increasing cost. Prediction of net radiation from the model yielded adequate results for some applications, such as evapotranspiration predictions and irrigation scheduling, but the model has considerable error at night due to some simplifying assumptions. Accurate net radiation measurements depend on proper placement of the sensor, proper leveling, and routine maintenance to keep the sensing surfaces clean.
B33E-1674
Multi-year Measurements of Stomatal and Non-stomatal Fluxes of Ozone to a Northern Mixed Hardwood Forest
Measurements of ozone, sensible heat, and latent heat fluxes, as well as relative humidity, temperature, pressure, wind speed, leaf area index, ambient ozone, and plant physiological parameters were made at a northern mixed hardwood forest located at the University of Michigan Biological Station in northern Michigan from June 27 to September 28, 2002; August 7 to October 10, 2003; May 22 to October 16, 2004; and June 12 to September 1, 2005. An analog resistance model was used to calculate aerodynamic, boundary layer, and canopy surface conductances for water vapor and ozone, and ozone canopy conductance was partitioned into stomatal and non-stomatal components. Analysis was limited to the period when leaves were fully grown, but not yet senescing: June 15 to September 15, according to LAI measurements. For each year, data available within this time frame was binned by hour-of-day and averaged. Mean daytime (0800-0200h) ozone canopy conductance showed little year-to-year variability: 0.39 mol m-2 s-1 (2002), 0.41 mol m-2 s-1 (2003), 0.52 mol m-2 s-1 (2004), and 0.43 mol m-2 s-1 (2005). Stomatal conductance showed expected patterns of behavior with respect to photosynthetic photon flux density (PPFD) and vapor pressure deficit (VPD), and mean daytime stomatal ozone conductances were: 0.17 mol m-2 s-1 (2002), 0.33 mol m-2 s-1 (2003), 0.40 mol m-2 s-1 (2004), and 0.20 mol m-2 s-1 (2005). Ozone non-stomatal conductance increased monotonically with increasing PPFD, and increased with temperature before falling off again at high temperature. Mean daytime non-stomatal ozone conductances accounted for as much as 61% (2002), 31% (2003), 36% (2004), and 57% (2005) of canopy conductance. Total ozone flux varied diurnally, with downward flux reaching -100 μ mol m-2 h-1 at midday, and at or near zero at night. Estimated daily average stomatal ozone burden (flux) was 2.9 x105 nmol m-2 (2002), 5.6 x105 nmol m-2 (2003), 6.6 x105 nmol m-2 (2004), and 4.1 x105 nmol m-2 (2005), with the non-stomatal partition representing 59% (2002), 26% (2003), 35% (2004), and 46% (2005) of the total flux. Non-stomatal ozone conductance was neither constant nor low. While this study leaves unanswered questions as to the mechanisms of non-stomatal conductance (non-stomatal ozone conductance may be occurring on surfaces or in gas-phase), it is important in the formulation of mechanistic models of ozone fluxes to quantify their magnitude and identify patterns with respect to major environmental factors in order to determine impacts of tropospheric ozone in forest ecosystems.
B33E-1675
Relating Soil CO2 Efflux and Root Respiration to Climate and Canopy Gas Exchange in Mature Ponderosa Pine
Soil respiration (Fs) is the second largest flux of carbon in terrestrial ecosystems and will play a key role during global climate change, yet our understanding of factors that control Fs are notably weak. We examined a six year automated chamber based record of Fs and the underlying components at a seasonally drought stressed pine forest in relation to climate and canopy gas exchange. Inter-annual variability of Fs was large (CV=17%) ranging between 612 and 1039 gC m-2y-1. On average 76 % of the variation of daily mean Fs could be quantified using a simple empirical model with year specific basal respiration rate, that was a linear function of above ground annual net primary productivity (ANPP), modulated by a common response to soil temperature and moisture. Using the natural range of climate variability across the 6 yrs to determine the degree of control on Fs by temperature and soil moisture; seasonal total Fs was twice as sensitive to soil moisture variability during the summer months compared to temperature variability during the same period and almost insensitive to the natural range of variability in spring temperature. Soil autotrophic respiration showed a strong seasonal pattern that was tightly linearly correlated with tree transpiration measured using sapflow techniques (r2=0.87) and gross ecosystem productivity (r2=0.81) as determined using the eddy flux approach. Diel patterns of soil autotrophic and heterotrophic respiration were consistent with location of carbon source in the soil profile and a simple diffusion model. Collectively the results suggest future models of soil respiration should consider the inclusion of canopy processes.
B33E-1676
Cloud Effects on Photosynthetically Active Radiation, Tropical Forest Canopy PAR Absorption and Photosynthesis in West Kalimantan, Indonesia
This study combines field measurements of total and diffuse photosynthetically active radiation (PAR, 400-700 nm) with a 3-D model of forest canopy radiative transfer to investigate the effects of atmospheric scattering by clouds on the surface PAR regime, canopy PAR absorption and photosynthetic rates at an observation site in Putussibau, West Kalimantan Province, Indonesia (0.84N, 112.93E). A time-series of PAR measurements collected from December 2006 to July 2007 is characterized by large diffuse fractions and frequent occurrences of enhanced total and diffuse irradiance, occasionally exceeding the extraterrestrial flux, which result from reflectance and scattering by the broken cloud fields that typify this tropical site. Simulations of canopy-absorbed PAR and gross photosynthesis with the Forest Light Environmental Simulator (FLiES) demonstrate the potential role of clouds in controlling forest canopy light-use efficiency and rates of photosynthesis and vegetation- atmosphere carbon exchange in this moist tropical environment.
B33E-1677
The effects of high ozone episodes on photosynthetic activity of temperate wetland plants
Photosynthetic activity in certain wetland plant species is reduced when ozone (O3) exposure increases. Few studies have examined the response of whole ecosystems during episodic events. We investigated net ecosystem exchange (NEE) response in a temperate wetland to periods of high ozone levels. The autochamber system at Sallie's Fen, Barrington, New Hampshire provides long-term measurements of NEE of carbon dioxide (CO2). These data were paired with ozone data from the AIRMAP site at Thompson Farm, Durham, NH (10 km to the south) for analysis of trends. Two chambers with the highest percentage cover of Carex rostrata were chosen for preliminary analysis. Residual analysis was conducted from non-linear regression in which a rectangular hyperbola was used to fit photosynthetically active radiation (PAR), which is known to affect NEE. This indicated a trend in the residuals for PAR in the two autochambers during the month of September toward a reduction in productivity with increased ozone mixing ratios for the period of 2001-2005. Leaf level photosynthesis measurements were also conducted at this fen during the summer of 2007 for Carex rostrata, Chamaedaphne calyculata and Alnus incana ssp.rugosa using a Li-6400 portable photosynthesis system. Initial findings showed a decrease in the photosynthetic activity of C. rostrata when O3 levels increased. C. rostrata was expected to be the most sensitive of the three species due to high stomatal density and thus higher uptake levels.