B31A-0278
Characteristics of Land-Atmosphere Water Fluxes for Diverse Plant Functional Types in the FLUXNET Network
We used the FLUXNET database to quantify the relative contribution of biological and physical factors that control seasonal evaporation for a range of plant functional types. With meteorological and evaporative flux measurements, we performed a sensitivity analysis of Penman-Monteith equation with respect to surface resistance, aerodynamic resistance, available energy, and vapor pressure deficit in each PFT on an annual scale. Each PFT showed distinct seasonal sensitivity of ET on each variable. The sensitivity of ET on length of growing season varies depending on each PFT. The mechanisms that control different sensitivity of ET on length of growing season are discussed.
B31A-0279
Characteristics of Heat, Water Vapor, and CO2 Fluxes above a Mountainous Cypress Forest in Taiwan
Results from 2-year measurements of sensible heat, water vapor, and CO2 fluxes above a homogeneous mountainous Cypress forest are presented. The site is located at 1650 m a.s.l. in northeastern Taiwan, close to a nature preserve. Vertical turbulent scalar fluxes are obtained by eddy- covariance method. The data were collected during the period from April 2005 to May 2007. Two gap filling strategies, including linear and nonlinear regressions between net radiation and scalar fluxes, were discussed and applied for filling the missing and rejected data. We examined the seasonal patterns of net radiation, sensible heat, latent heat, and CO2 fluxes. The results showed that all three fluxes change seasonally, in consist with net radiation. Bowen ration variation with season and relative transport efficiencies between heat and water vapor and carbon dioxide under different atmospheric stabilities were also discussed.
B31A-0280
Water-use Comparison of the Invasive Tree Species, Melaleuca Quinquenervia, and two Native Tree Species,Taxodium Distichum and Pinus Elliottii, in Southwest Florida
Melaleuca (Melaleuca quinquenervia), an invasive tree species in southern Florida, is generally thought to have higher transpiration rates than the native vegetation, however little empirical data is available to support this claim. In this study, thermal dissipation probes were used to measure transpiration rates of the three species growing in a hydric ecotone in southwest Florida. Transpiration rates of melaleuca, slash pine (Pinus elliottii), and bald cypress (Taxodium distichum) were compared to assess seasonal variability between the wet and dry seasons. Individually trees of both bald cypress and slash pine showed significantly higher water fluxes than melaleuca (p<0.05). However, when individual tree fluxes were scaled to the ecosystem-level, melaleuca contributed 21% of the water flux and bald cypress contributed 72% during the wet season. Melaleuca's increased contribution at the landscape-level results from higher tree densities at our study site. Following leaf senescence in the early dry season, bald cypress continues to be a significant water user at the landscape level. With higher atmospheric demands for water, bald cypress was the least conservative of the three species with respect to water use, whereas on days with low atmospheric demands for water the three species function similarly. These results do not support the hypothesis that melaleuca uses more water than the native Florida tree species, however, they suggest the density of melaleuca at the landscape-scale is important in our understanding of its role in the hydrologic cycle.
B31A-0281
Analysis of Periods With Strong and Coherent CO2 Advection Over a Forested Hill
Horizontal and vertical advection of CO2 were measured at the Wetzstein site, a spruce ecosystem located on a hill near the village of Lehesten, Germany. These fluxes were measured during the CARBOEUROPE-IP advection experiment (ADVEX), carried out between April and June of 2006, when four additional towers were installed around the existing one, forming a polygon-like set-up [Feigenwinter et al. (2008)]. Measurements of wind speed, temperature, water vapour, and CO2 concentration were performed at 4 levels on all five towers. Horizontal gradients of CO2 were measured at 1.5 m in two perpendicular transects inside the area of the cube formed by the 4 towers. An additional tower was installed at one of the slopes of the hill. This tower was located at approximately 900 m from the hillcrest on the southwest sector, as this is the prevailing wind direction at the site. Horizontal and vertical advection terms were found to be related to cross-ridge flows, under specific micrometeorological conditions. Additionally, modelling results that support the existence of advection at low hills covered with a canopy were verified. According to these models, the flow passing over a low hill influences the wind velocity and scalar concentration fields above and below the vegetation, inducing advective transport of quantities such as CO2 [Finnigan and Belcher (2004); Katul et al. (2006)]. The first evidence was the reverse flow at the downwind side of the hill, obtained from the above and below-canopy measurements of wind direction at the tower located on the slope. The second experimental evidence came from the analysis of the horizontal gradient of CO2 inside the square formed by the ADVEX towers. According to this result, the CO2 accumulates close to the downwind side of the ridge, as confirmed by the modelling of the transport of scalars across a low hill covered with a canopy (Ned Patton, pers. comm.). The results presented in this work can be used to infer the existence of advective fluxes of CO2 in ecosystems located on similar topographies. The agreement of measurements with modelling work supports the theoretical approach about the influence of the flow passing over a low hill covered with a vegetation layer. These two sets of results help to expand the current knowledge of advection of CO2 over complex topographies.
B31A-0282
Is the Interannual Variability of NEE Controlled by Dryness?
The global rate of fossil fuel combustion continues to rise, but the amount of CO2 accumulating in the atmosphere has not increased accordingly. The relative magnitudes of carbon sinks are widely debated. In particular, the locations and mechanisms that drive interannual variability of atmospheric CO2 are highly uncertain. Terrestrial carbon reservoirs are believed to cause more interannual variability of atmospheric CO2 than oceanic carbon reservoirs (Bousquet et al., 2001). Determining controlling factors of interannual variability of terrestrial carbon sequestration is a key to understanding of essential processes of terrestrial carbon sinks and sources. We pose a hypothesis that the interannual variability of the ecosystem-atmosphere exchange (NEE) of CO2 is controlled by dryness. We use the data from the regional and global networks of flux towers to test this hypothesis. The dryness is a dimensionless parameter defined by Budyko (1974) as the ratio of the potential evapotranspiration to precipitation (P), Dryness = Rn/(LP) Where Rn is annual sum of net radiation, and L is the enthalpy of vaporization. Therefore, Rn/L is the potential evapotranspiration. Budyko used the dryness parameter to successfully classify geobotanic zones globally. Our initial analysis demonstrated that the annual NEE numbers are well organized by dryness parameter based on the data from several flux towers in the AmeriFlux network. We are using improved methodologies for filling missing flux, Rn and P data and are extending this analysis to include a larger number of sites. Although this model is oversimplified, dryness may play an important role in determination of annual variability of NEE of CO2 because this parameter has two features: (1) annual water balance (potential evapotranspiration to actual precipitation; and (2) energy balance (available radiation energy to latent heat). The fundamental understanding of the link between dryness and terrestrial carbon sequestration is that terrestrial exchanges of carbon and water are physiologically controlled by the same opening/closing mechanisms of plant stomata.
B31A-0283
Photosynthetic activity of a black spruce tree under cold environment and the contribution to annual CO2 budget
We have measured tower-based CO2 flux at a sub-arctic black spruce forest interior Alaska during past six years, and found interesting CO2 sink during early spring when the temperature was below zero and the ground was still frozen. There is a hot discussion on flux measurement under cold environment, that is, the self-heated open-path sensor causes apparent downward flux. To verify the downward flux during cold season, we measured CO2 exchange by attaching a leaf-chamber to a stem of black spruce. CO2 exchange by the tree was detected on DOY 67 at first, which was triggered by temperature. When the daily degree-hour of air temperature above zero (dayDH) raised over 50 degree–hour, the respiration started. However, the respiration level was quite low and returned negligible when dayDH fell to zero. The CO2 accumulation, namely photosynthesis of the tree occurred on DOY 88 when the dayDH raised again over zero while daily averages of air t and soil temperature were still below zero. Spring break was DOY 121 and soil temperature at 0.1m depth was below zero by DOY 120. From the leaf-chamber measurement, we decided that the CO2 sink by the black spruce forest started after DOY 88 in 2008, which was about three weeks earlier than ordinal green-up of the forest. If we include the CO2 sink in cold season into the annual CO2 budget, the tower-based NEE increased about 10% of its sink strength.
B31A-0284
Understanding Biophysical Regulations of Ecosystem and Soil Respiration Across Multiple Vegetation Types
A major ecological question is how biophysical factors influence terrestrial CO2 fluxes among different vegetation types. To understand how ecosystems loss CO2 to the atmosphere it is crucial to study the biophysical mechanisms that regulate ecosystem respiration (CO2 losses to the atmosphere by plants and soils) and soil respiration (CO2 losses to the atmosphere by soils). We present hypothetical mechanisms on how biophysical factors regulate ecosystem and soil respiration among different vegetation types. For ecosystem respiration we propose that environmental factors (temperature, water and light) influence photosynthesis and through this process they influence ecosystem respiration. For soil respiration we propose that environmental factors (temperature and water) influence the production of CO2 in the soil and through this process they influence soil respiration. Finally, we identified that soil CO2 production is influence by an unmeasured factor other than temperature and water and we tested how photosynthesis regulates soil respiration at different vegetation types.
B31A-0285
Atmospheric-Ecosystem CO2 Exchange in Sparse Arid Shrublands Across the Great Basin USA Over Multiple Years: Identifying Patterns and Mechanisms
Up to recently, desert ecosystems have essentially been ignored with respect to their influence on global carbon cycling and their potential role in modulating atmospheric CO2 levels. Because deserts, defined here as ecosystems receiving <280 mm of precipitation annually, cover 35% of Earth's surface, even small positive or negative net ecosystem CO2 exchange (NEE=fluxes) can have globally meaningful effects on atmospheric CO2. Since 2003 we have been measuring NEE and annual NEP at 10 arid shrubland sites around the Great Basin in Nevada, USA using eddy covariance and large static chamber "domes" with the objectives of quantifying seasonal, annual and interannual fluxes and the environmental and ecological factors that may be modulating these fluxes. Surprisingly, annual NEP measured in Mojave Desert creosote bush (Larrea tridentata)-dominated ecosystems, high desert sagebrush steppe (Aremesia tridentata) ecosystems, and greasewood (Sarcobatus vermiculatus) ecosystems have been largely positive (net C uptake by ecosystems; range of zero to 90 g C m-2 yr-1) and often large (as high 100 to 180 g C m-2 yr-1). Thus, the data from these arid shrublands suggest a much larger arid land C sink than has been previously assumed and call for closer tracking of the CO2 fluxes in these ecosystems.
B31A-0286
Deriving "Daily" Variables from the Ameriflux Standard Eddy Covariance Data Set
A gap-filled, quality assessed eddy covariance dataset has recently become available for the AmeriFlux
network. This dataset uses standard processing and produces commonly used science variables. This
shared dataset enables robust comparisons across different analyses. Of course, there are many remaining
questions. One of those is how to define "during the day" which is an important concept for many analyses.
Some studies have used local time – for example 9am to 5pm; others have used thresholds on
photosynthetic active radiation (PAR). A related question is how to derive quantities such as the Bowen ratio.
Most studies compute the ratio of the averages of the latent heat (LE) and sensible heat (H). In this study, we
use different methods of defining "during the day" for GPP, LE, and H. We evaluate the differences between
methods in two ways. First, we look at a number of statistics of GPP, LE, and H. Second, we look at
differences in the derived Bowen ratio and correlations between GPP and LE. Our goal is not science per
se, but rather informatics in support of the science.
http://www.fluxdata.org
B31A-0287
Exceptional Carbon Uptake in European Forests During the 2007 Warm Spring: a Data- Model Analysis.
Temperate to boreal forests undergo drastic springtime functional changes, shifting within a few weeks from net carbon sources to net carbon sinks. Most of these changes are mediated by temperature. The autumn 2006-winter 2007 record warm period was followed by an exceptionally warm spring in Europe, which makes spring 2007 a good candidate for advances of the onset of the photosynthetically active period, hence high net carbon uptake. The analysis of a decade of eddy covariance data at six European forests stands, encompassing a wide range of functional types (broadleaf evergreen, broadleaf deciduous, needleleaf evergreen) and latitudinal band (from 44N to 62N), revealed exceptional fluxes during spring 2007. Gross Primary Productivity (GPP) top ranked the decade record for all sites but a Mediterranean evergreen forest (with a +40 to +120 gC m-2 anomaly compared to the decadal mean over the January-May period). Total Ecosystem Respiration (TER) was also promoted during spring 2007, though relatively less than GPP (with a +10 to +60 gC m-2 anomaly over five months), leading to higher than the long-term mean net uptake at all sites (+10 to +100 gC m-2 anomaly over five months). A correlative analysis relating springtime carbon fluxes to simple phenological indices suggested the exceptional net carbon uptake and temperatures to be related. The CASTANEA process-based SVAT/growth model was used to disentangle the seasonality of climatic drivers (incoming radiation, air and soil temperature) and biological drivers (canopy dynamics, thermal acclimation of photosynthesis to low temperatures) on spring C fluxes along the latitudinal gradient. A sensitivity analysis of model simulations further evidenced the role of (i) an exceptional early budburst combined to elevated air temperature in deciduous sites, (ii) an early relief of winter thermal acclimation in evergreen sites for the promotion of 2007 spring assimilation. The 2007 warm spring and 2003 hot-dry summer anomalies appeared of comparable absolute magnitudes, which is illustrative of a possible compensation if both extreme events were to be combined in the future. A model analysis revealed that earlier resumption of photosynthetic processes during a warm spring might extent the period of soil water deficit, depending on the spring precipitation regime.
B31A-0288
Simultaneous Carbon Dioxide and Methane Eddy-Covariance Flux Measurements Using a High-Speed WS-CRDS Analyzer: Field Comparisons to Conventional AmeriFlux Systems
The well-known eddy-covariance flux method is commonly employed to study biogeochemical scalar fluxes of greenhouse gases between the land surface and the atmosphere. Its use has so far been limited mainly to flux measurements of carbon dioxide (CO2) and water vapor due to the lack of adequate instrumentation for other greenhouse gases. The carbon exchange of gas species between terrestrial ecosystems and the atmosphere, particularly in areas with stagnant water or poorly drained and heavy soils, represents a substantial portion of the atmospheric budget of both CO2 and methane (CH4), and so quantifying such exchanges of both gases is critical in adding to our understanding of the carbon cycle. Recently, a novel closed-path analyzer has been evaluated against conventional instrumentation for performing not only CO2 flux measurements, but for making simultaneous CO2 and CH4 flux measurements. We present field data evaluating the performance of a high-frequency fast-response (10Hz) gas analyzer based on Wavelength-Scanned Cavity Ring Down Spectroscopy (WS-CRDS) that was developed for simultaneous, dual-gas eddy-covariance flux measurements. The performance of this analyzer was evaluated in terms of what is required for appropriately high-quality measurements involving the exchange of both CO2 and CH4 between terrestrial ecosystems and the atmosphere. The WS- CRDS analyzer was deployed at Hyslop Crop Science Field Research Laboratory outside of Corvallis, OR and was compared in real-time with the AmeriFlux Portable Eddy Covariance System for measuring CO2 fluxes using conventional infrared open- and closed path CO2 and water vapor analyzers. Data comparing the performance of this WS-CRDS analyzer against this currently-used AmeriFlux instrumentation is presented.
B31A-0289
Evapotranspiration Partitioning and its Effects on Terrestrial Ecosystem Dynamics in the GISS Land Surface Model
We explore the partitioning of evapotranspiration into canopy evaporation, transpiration, and soil evaporation and its effects on terrestrial net primary productivity with the Goddard Institute for Space Studies (GISS) land-surface model coupled to the Ent Dynamic Global Terrestrial Ecosystem Model (DGTEM). Limited observations and most land surface models indicate that transpiration is the dominant component at the global scale, but the current GISS land-surface model and Ent DGTEM predict that canopy evaporation significantly outweighs both transpiration and soil evaporation. Low estimates of transpiration have implications for photosynthetic productivity, because water and carbon fluxes are highly coupled. The effects of climate on evapotranspiration partitioning are also analyzed through its relationship with the dryness index, the ratio of available energy (net radiation) to available water (precipitation). The GISS land-surface model and the Ent DGTEM are used to simulate evapotranspiration and net primary productivity at both the canopy and global scales. Fluxnet meteorological data force the models for canopy- scale simulations, which allow evaluation of evapotranspiration partitioning and net primary productivity with observed soil moisture, latent heat, and CO2 fluxes. The analysis is conducted for several plant functional types including broadleaf forest, needleleaf forest, and grasses. For the global-scale analysis, the models are first run with meteorological forcings from a global dataset for 1951 to 2000 (Sheffield et al., 2006). The GISS land-surface model and the Ent DGTEM are then incorporated into the GISS global climate model (GCM) to improve understanding of the effects of evapotranspiration partitioning on climate (e.g. surface temperature, precipitation). For both types of global simulation, the relationships of evapotranspiration partitioning and net primary productivity with the dryness index are explored and the distributions of annual and seasonal dryness indices are computed for each plant functional type. We propose that dryness-index statistics for each plant functional type are important measures of the impact of GCM climate biases on terrestrial ecosystems.
B31A-0290
Assessing the Performance of the Simple Biosphere 3 (SiB3) Model in the Amazon Basin
Interactions between biosphere and atmosphere are extremely important when characterizing climatic forcing over land surfaces. These interactions have been extensively examined in several large-scale experiments, most recently in the case of the Amazon Basin by the Large-Scale Biosphere-Atmosphere Experiment in Amazonia (LBA). Such experiments seek to improve understanding of the biosphere-atmosphere system by measuring and analyzing different aspects of the ecosystem exchange, including soil water dynamics and below- and above-canopy energy and gas exchanges, etc., and incorporating these findings into the land surface parameterization (LSP) schemes used in regional or global climate models. The present study is a contribution to the LBA Model Intercomparison Project (LBA-MIP), which seeks to identify potential differences between simulated and observed behavior over the Amazon basin and to investigate the differences in simulations provided by a range of LSPs. In general, LSPs contain several parameters that can be calibrated to improve model performance relative to observations. The preferred approach is to use automatic optimization algorithms to estimate model parameters from field data. This study evaluates the performance of the Simple Biosphere 3 (SiB3) LSP when model parameters are estimated using the Multi- Objective Shuffled Complex Evolution Metropolis (MOSCEM) with different levels of data availability, the objective being to understand how sensitive SiB3 is to different field data with a view to improving the experimental design of future hydrometeorological experiments.
B31A-0291
Spatial Structure in North American Regional Carbon Dioxide Fluxes Evaluated With a Simple Land Surface Model
We evaluate spatial structure in Ameriflux CO2 flux observations using a simple diagnostic land surface model. The Vegetation Photosynthesis Respiration Model (VPRM) calculates NEE using locally observed temperature and PAR, and satellite-derived phenology and moisture. We use observed NEE from a group of 65 Ameriflux eddy covariance tower sites spanning North America to estimate VPRM parameters for these sites. We use the spatial structure of VPRM errors to investigate spatial coherence in regional CO2 fluxes at several different time scales. We find that strong spatial structure does exist in the data-model residuals. Preliminarily, this structure is not seen unless model parameters are estimated on a local basis.
B31A-0292
Evaluation of the Advanced-Canopy-Atmosphere-Surface Algorithm (ACASA Model) Using Eddy Covariance Technique Over Sparse Canopy
Land surface models are usually used to quantify energy and mass fluxes between terrestrial ecosystems and atmosphere on micro- and regional scales. One of the most elaborate land surface models for flux modelling is the Advanced Canopy-Atmosphere-Soil Algorithm (ACASA) model, which provides micro-scale as well as regional-scale fluxes when imbedded in a meso-scale meteorological model (e.g., MM5 or WRF). The model predicts vegetation conditions and changes with time due to plant responses to environment variables. In particular, fluxes and profiles of heat, water vapor, carbon and momentum within and above canopy are estimated using third-order equations. It also estimates turbulent profiles of velocity, temperature, humidity within and above canopy, and CO2 fluxes are estimated using a combination of Ball-Berry and Farquhar equations. The ACASA model is also able to include the effects of water stress on stomata, transpiration and CO2 assimilation. ACASA model is unique because it separates canopy domain into twenty atmospheric layers (ten layers within the canopy and ten layers above the canopy), and the soil is partitioned into fifteen layers of variable thickness. The model was mainly used over dense canopies in the past, so the aim of this work was to test the ACASA model over a sparse canopy as Mediterranean maquis. Vegetation is composed by sclerophyllous species of shrubs that are always green, with leathery leaves, small height, with a moderately sparse canopy, and that are tolerant at water stress condition. Eddy Covariance (EC) technique was used to collect continuous data for more than 3 years period. Field measurements were taken in a natural maquis site located near Alghero, Sardinia, Italy and they were used to parameterize and validate the model. The input values were selected by running the model several times varying the one parameter per time. A second step in the parameterization process was the simultaneously variation of some parameters. ACASA simulations were compared with measured fluxes of net radiation (Rn), sensible heat (H), latent heat (LE), soil heat (G), and CO2 fluxes at half-hourly time scale. Statistical analysis was made to evaluate model performance. Comparisons between simulated and measured values were evaluated using linear regression, the root mean squared error (RMSE), mean absolute error (RA), and mean bias error (MBE). Modeled data showed a good energy balance closure. ACASA estimates of net radiation were excellent. Sensible (H) and latent heat (LE) flux predictions exhibited only small differences between modeled and observed data. The ACASA model was able to capture the seasonal variation in CO2 flux. Net Ecosystem Exchange (NEE) showed the typical summer decrease due to drought induced water stress, and the simulations predicted the lower CO2 flux. Differences between simulated and observed fluxes were significant at 0.001 probability. ACASA simulations, therefore, are considered good. So, we can say that the use of ACASA to predict energy and mass fluxes between the vegetation and atmosphere is promising, and it could greatly improve our ability to estimate fluxes over natural ecosystems at both local and regional scales.