Uncertainty analysis in data processing on the estimation of net carbon exchanges at different forest ecosystems in China

Information about the uncertainties associated with eddy covariance observations of surface-atmosphere CO₂ exchange is of importance for model-data fusion in carbon cycling studies and the accurate evaluation of ecosystem carbon budgeting. In this paper, a comprehensive analysis was conducted to investigate the influence of data processing procedures, focusing especially on the nocturnal data correction and three procedures in nonlinear regression method of gap filling [i.e., the selection of respiration model (REM), light-response model (LRM) and parameter optimization criteria (POC)], on the annual net ecosystem CO₂ exchange estimation at three forest ecosystems in ChinaFLUX with three yearly datasets for each site. The results showed that uncertainties caused from four methodological uncertainties were between 61 and 108?g?C?m^?2?year^?1, with 61?C93?g?C?m^?2?year^?1 (21?C30%) in a temperate mixed forest, 80?C107?g?C?m^?2?year^?1 (19?C21%) in a subtropical evergreen coniferous plantation and 77?C108?g?C?m^?2?year^?1 (16?C19%) in a subtropical evergreen broad-leaved forest. Factorial analysis indicated that the largest uncertainty was associated with the choice of POC in the regression method across all sites in all years, while the influences of the choice of models (i.e., REM and LRM) varied with climate conditions at the measurement station. Furthermore, the uncertainty caused by data processing procedures was of approximately the same magnitude as the interannual variability in the three sites. This result stressed the importance to understand the uncertainty caused by data processing to avoid the introduction of artificial between-year and between-site variability that hampers comparative analysis.     

A statistical thinning model for initialising large-scale ecosystem models

Abstract

Large-scale ecosystem models are important tools for carbon assessment at national scales. Many of these models are not initialised with known field data from any particular time, but simulate the growth of each stand from its estimated germination year up to the present or future. The models will overestimate current-day standing volume or biomass unless historic stand management (biomass removal due to thinning) is taken into account. The full management history of each stand is rarely known, and must be somehow estimated. One possibility is to build statistical thinning models based on data in a National Forest Inventory, which could then be integrated into the ecosystem models. If the harvesting model is constructed using only variables that are also used within the ecosystem model, then the management impacts can be included in the ecosystem model for the entire simulated life of the stand. In the case of most flux dynamics models, this precludes the use of the tree-level data that harvesting models have traditionally relied on. In this article, we develop a novel means to interrogate a subset of the Austrian National Forest Inventory based on deriving probability density functions for particular combinations of stand and site variables. We determine the parameters of a probabilistic model to estimate historic patterns of timber removals and validate it against inventory estimates. Our procedure can establish supportable estimates of historic management regimes suitable as input data for subsequent modelling of national-scale forest carbon stocks, sources and sinks.     

Dataset of CarboEastAsia and uncertainties in the CO2 budget evaluation caused by different data processing

The datasets of net ecosystem CO₂ exchange (NEE) were acquired from 21 forests, 3 grasslands, and 3 croplands in the eastern part of Asia based on the eddy covariance measurements of the international joint program, CarboEastAsia. The program was conducted by three networks in Asia, ChinaFLUX, JapanFlux, and KoFlux, to quantify, synthesize, and understand the carbon budget of the eastern part of Asia. An intercomparison was conducted for NEE estimated by three gap-filling procedures adopted by ChinaFLUX, JapanFlux, and KoFlux to test the range of uncertainty in the estimation of NEE. The overall comparison indicated good agreement among the procedures in the seasonal patterns of NEE, although a bias was observed in dormant seasons depending on the different criteria of data screening. Based on the gap-filled datasets, the magnitude and seasonality of the carbon budget were compared among various biome types, phenology, and stress conditions throughout Asia. The annual values of gross primary production and ecosystem respiration were almost proportional to the annual air temperature. Forest management, including clear-cutting, plantation, and artificial drainage, was significant and obviously affected the annual carbon uptake within the forests. Agricultural management resulted in notable seasonal patterns in the crop sites. The dataset obtained from a variety of biome types would be an essential source of knowledge for ecosystem science as well as a valuable validation dataset for modeling and remote sensing to upscale the carbon budget estimations in Asia.     

Uncertainty analysis of modeled carbon fluxes for a broad-leaved Korean pine mixed forest using a process-based ecosystem model

The uncertainty in the predicted values of a process-based terrestrial ecosystem model is as important as the predicted values themselves. However, few studies integrate uncertainty analysis into their modeling of carbon dynamics. In this paper, we conducted a local sensitivity analysis of the model parameters of a process-based ecosystem model at the Chaibaishan broad-leaved Korean pine mixed forest site in 2003?C2005. Sixteen parameters were found to affect the annual net ecosystem exchange of CO₂ (NEE) in each of the three?years. We combined a Monte Carlo uncertainty analysis with a standardized multiple regression method to distinguish the contributions of the parameters and the initial variables to the output variance. Our results showed that the uncertainties in the modeled annual gross primary production and ecosystem respiration were 5?C8% of their mean values, while the uncertainty in the annual NEE was up to 23?C37% of the mean value in 2003?C2005. Five parameters yielded about 92% of the uncertainty in the modeled annual net ecosystem exchange. Finally, we analyzed the sensitivity of the meteorological data and compared two types of meteorological data and their effects on the estimation of carbon fluxes. Overestimating the relative humidity at a spatial resolution of 10?km?×?10?km had a larger effect on the annual gross primary production, ecosystem respiration, and net ecosystem exchange than underestimating precipitation. More attention should be paid to the accurate estimation of sensitive model parameters, driving meteorological data, and the responses of ecosystem processes to environmental variables in the context of global change.     

Plant ecophysiology and forest response to global change

There are many ways of studying forest responses to global change. Most current national and international programs focus on net gas exchange of the terrestrial biosphere and are typically interdisciplinary, multi-scale projects. Key objectives of these programs are surprisingly similar to those of classical plant ecophysiology studies, i.e., to explore functional relationships of plant or plant community responses to environmental change. Thus, common research questions that link plant ecophysiology to ecosystem functioning can be identified for both research communities, promising complementarity and synergism for joint research projects. Although some well-established ecophysiological relationships, such as light responses or stomatal limitations of photosynthetic gas exchange, are currently employed in many ecosystem-scale net flux studies for gap-filling or modeling, only 14% (n = 27) of all eddy covariance flux studies in forests (n = 196; published between 1992 and April 2002) include plant ecophysiological measurements (n = 24) or biomass and growth estimates (n = 8). Generally, emphasis is on CO2 exchange measurements at various scales (foliage, shoots, branches; n = 14) and water relations measurements (n = 11). These measurements do not fully support the typical parameterization of stand and regional models, which often need information on canopy architecture and nitrogen nutrition. By means of a complementary research approach, valuable information can be acquired that is unobtainable by means of a single approach. This additional information is important for the identification of underlying biotic and environmental drivers, for the regulation of net ecosystem fluxes and their partitioning, and the independent validation of measured net ecosystem fluxes. Thus, combining micrometeorology and ecophysiology at flux sites is strongly recommended for ecosystem functioning studies.     

What the towers don't see at night: nocturnal sap flow in trees and shrubs at two AmeriFlux sites in California

At the leaf scale, it is a long-held assumption that stomata close at night in the absence of light, causing transpiration to decrease to zero. Energy balance models and evapotranspiration equations often rely on net radiation as an upper bound, and some models reduce evapotranspiration to zero at night when there is no solar radiation. Emerging research is showing, however, that transpiration can occur throughout the night in a variety of vegetation types and biomes. At the ecosystem scale, eddy covariance measurements have provided extensive data on latent heat flux for a multitude of ecosystem types globally. Nighttime eddy covariance measurements, however, are generally unreliable because of low turbulence. If significant nighttime water loss occurs, eddy flux towers may be missing key information on latent heat flux. We installed and measured rates of sap flow by the heat ratio method (Burgess et al. 2001) at two AmeriFlux (part of FLUXNET) sites in California. The heat ratio method allows measurement and quantification of low rates of sap flow, including negative rates (i.e., hydraulic lift). We measured sap flow in five Pinus ponderosa Dougl. ex Laws. trees and three Arctostaphylos manzanita Parry and two Ceanothus cordulatus A. Kellog shrubs in the Sierra Nevada Mountains, and in five Quercus douglasii Hook and Arn. trees at an oak savanna in the Central Valley of California. Nocturnal sap flow was observed in all species, and significant nighttime water loss was observed in both species of trees. Vapor pressure deficit and air temperature were both well correlated with nighttime transpiration; the influence of wind speed on nighttime transpiration was insignificant at both sites. We distinguished between storage-tissue refilling and water loss based on data from Year 2005, and calculated the percentage by which nighttime transpiration was underestimated by eddy covariance measurements at both sites.     

Error propagation in stock-difference and gain–loss estimates of a forest biomass carbon balance

According to the United Nations International Panel on Climate Change good practice guidance, an annual forest biomass carbon balance (AFCB) can be estimated by either the stock-difference (SD) or the gain–loss (GL) method. An AFCB should be accompanied by an analysis and estimation of uncertainty (EU). EUs are to be practicable and supported by sound statistical methods. Sampling and model errors both contribute to an EU. As sample size increases, the sampling error decreases but not the error due to errors in model parameters. Uncertainty in GL AFCB estimates is dominated by model-parameter errors. This study details the delta technique for obtaining an EU with the SD and the GL method applicable to the carbon in aboveground forest biomass. We employ a Brownian bridge process to annualize the uncertainty in SD AFCBs. A blend of actual and simulated data from three successive inventories are used to demonstrate the application of the delta technique to SD- and GL-derived AFCBs during the years covered by the three inventories (SD) and rescaled national wood volume harvest statistics (GL). Examples are limited to carbon in live trees with a stem diameter of 7 cm or greater. We confirm that a large contribution to the uncertainty in an AFCB comes from models used to estimate biomass. Application of the delta technique to summary statistics can significantly underestimate uncertainty as some sources of uncertainty cannot be quantified from the available information. We discuss limitations and problems with the Monte Carlo technique for quantifying uncertainty in an AFCB.     

Uncertainty in eddy covariance measurements and its application to physiological models

Flux data are noisy, and this uncertainty is largely due to random measurement error. Knowledge of uncertainty is essential for the statistical evaluation of modeled and measured fluxes, for comparison of parameters derived by fitting models to measured fluxes and in formal data-assimilation efforts. We used the difference between simultaneous measurements from two towers located less than 1 km apart to quantify the distributional characteristics of the measurement error in fluxes of carbon dioxide (CO2) and sensible and latent heat (H and LE, respectively). Flux measurement error more closely follows a double exponential than a normal distribution. The CO2 flux uncertainty is negatively correlated with mean wind speed, whereas uncertainty in H and LE is positively correlated with net radiation flux. Measurements from a single tower made 24 h apart under similar environmental conditions can also be used to characterize flux uncertainty. Uncertainty calculated by this method is somewhat higher than that derived from the two-tower approach. We demonstrate the use of flux uncertainty in maximum likelihood parameter estimates for simple physiological models of daytime net carbon exchange. We show that inferred model parameters are highly correlated, and that hypothesis testing is therefore possible only when the joint distribution of the model parameters is taken into account.     

Carbon dioxide exchange in Norway spruce at the shoot, tree and ecosystem scale

Net CO2 exchange in a 35-year-old boreal Norway spruce (Picea abies (L.) Karst.) forest in northern Sweden was measured at the shoot (NSE), tree (NTE) and ecosystem levels (NEE) by means of shoot cuvettes, whole-tree chambers and the eddy covariance technique, respectively. We compared the dynamics of gross primary production (GPP) at the three levels during the course of a single week. The diurnal dynamics of GPP at each level were estimated by subtracting half-hourly or hourly model-estimated values of total respiration (excluding light-dependent respiration) from net CO(2) exchange. The relationship between temperature and total respiration at each level was derived from nighttime measurements of NSE, NTE and NEE over the course of 1 month. There was a strong linear relationship (r2 = 0.93) between the hourly estimates of GPP at the shoot and tree levels, but the correlation between shoot- and ecosystem-level GPP was weaker (r2 = 0.69). However, the correlation between shoot- and ecosystem-level GPP was improved (r2 = 0.88) if eddy covariance measurements were restricted to periods when friction velocity was > or = 0.5 m s(-1). Daily means were less dependent on friction velocity, giving an r2 value of 0.94 between shoot- and ecosystem-level GPP. The correlation between shoot and tree levels also increased when daily means were compared (r2 = 0.98). Most of the measured variation in carbon exchange rate among the shoot, tree and ecosystem levels was the result of periodic low coupling between vegetation and the atmosphere at the ecosystem level. The results validate the use of measurements at the shoot and tree level for analyzing the contribution of different compartments to net ecosystem CO2 exchange.     

Effects of mutual shading of tree crowns on prediction of photosynthetic light-use efficiency in a coastal Douglas-fir forest

Gross primary production (GPP) is often expressed as the product of absorbed photosynthetically active radiation and the efficiency (epsilon) with which a plant community uses absorbed radiation in biomass production. Light-use efficiency is affected by environmental stresses, and varies diurnally and seasonally. Uncertainty about epsilon can be a serious limitation when modeling GPP. An important determinant of epsilon is the amount and type of solar radiation incident on a canopy, because an abundance of light can trigger a photo-protective reaction, diminishing GPP. The radiation regime in a forest canopy is determined by the predominant sky conditions and by mutual shading of tree crowns. Shading effects, producing shifts in the amount of incident direct and diffuse solar radiation, have been largely ignored, however, because they depend on forest structure and are difficult to measure. We describe a new approach for estimating changes in mutual canopy shading throughout the day and year based on a canopy structure model derived from light detection and ranging (LiDAR). Proportions of canopy shading were then combined with eddy covariance data to assess the explanatory power for variance in epsilon by regression tree analysis over half-hourly, daily and weekly time scales. The approach explained between 75 and 97% of variance in epsilon, representing an increase of between 5 and 16% compared with models driven solely by meteorological variables.     

Photosynthesis and respiration of black spruce at three organizational scales: shoot, branch and canopy

To gain insight into the function of photosynthesis and respiration as processes operating within a global ecosystem, we measured gas exchange of mature black spruce (Picea mariana (Mill.) B.S.P.) trees at three organizational scales: individual shoots, whole branches and a forest canopy. A biochemical model was fitted to these data, and physiological parameters were extracted. Pronounced seasonal variation in the estimated model parameters was found at all three organizational scales, highlighting the need to make physiological measurements throughout the year. For example, it took over 100 days for physiological activity to increase from zero during the springtime thaw to its yearly maximum. Good agreement was found between parameter values estimated for the different organizational scales, suggesting that, in the case of aerodynamically rough, largely mono-specific forest canopies, physiological parameters can be estimated from eddy covariance flux measurements. The small differences between photosynthetic parameters estimated at the different scales suggest that the overall spatial organization of photosynthetic capacity is nearly optimized for carbon uptake at each scale.     

Soil respiration in a mixed temperate forest and its contribution to total ecosystem respiration

Soil respiration (SR) was measured with an infrared gas analyzer in nine plots representative of the heterogeneous vegetation in a mixed coniferous-deciduous forest in the Belgian Campine region. Selected plots included the two most representative overstory species (Pinus sylvestris L. and Quercus robur L.) in combination with the most representative understory species of the forest. A model that includes temperature and water as the main controlling variables was fitted to the data. We found large spatial variability in SR among plots, with typically lower fluxes under the coniferous overstory than under the deciduous overstory (means of 4.8 +/- 0.4 and 8.8 +/- 0.5 Mg C ha(-1) year(-1), respectively). Total annual soil carbon (C) emissions were estimated by weighting fluxes from different types of vegetation according to their relative contribution to the footprint area of the eddy covariance flux measurement. The relative contribution of the two main tree species to the footprint-weighted total SR varied among seasons with the more abundant coniferous overstory contributing the most to total SR during most of the year. Nonetheless, during summer, the contribution of deciduous plots to total SR was disproportionally high because of the more pronounced seasonality of belowground metabolic activity. Net ecosystem carbon dioxide exchange was measured by eddy covariance, and we estimated total ecosystem respiration (TER) with footprint-constrained nighttime fluxes. Mean total annual SR and TER were 6.1 +/- 0.11 and 9.1 +/- 1.15 Mg C ha(-1) year(-1), respectively. The 95% confidence interval of the ratio of annual SR:TER ranged from 0.58 to 0.76, with a mean of 0.67. The contribution of SR to TER tended to vary seasonally, with minimum contributions during summer (less than 50% of TER) and maximum contributions during winter (about 94% of TER).     

Quantifying the sources of epistemic uncertainty in model predictions of insect disturbances in an uncertain climate

Key message

Natural disturbance can disrupt the anticipated delivery of forest-related ecosystem goods and services. Model predictions of natural disturbances have substantial uncertainties arising from the choices of input data and spatial scale used in the model building process, and the uncertainty of future climate conditions which are a major driver of disturbances. Quantifying the multiple contributions to uncertainty will aid decision making and guide future research needs.

Context

Forest management planning has been able, in the past, to rely on substantial empirical evidence regarding tree growth, succession, frequency and impacts of natural disturbances to estimate the future delivery of goods and services. Uncertainty has not been thought large enough to warrant consideration. Our rapidly changing climate is casting that empirical knowledge in doubt.

Aims

This paper describes how models of future spruce budworm outbreaks are plagued by uncertainty contributed by (among others): selection of data used in the model building process; model error; and uncertainty of the future climate and forest that will drive the future insect outbreak. The contribution of each to the total uncertainty will be quantified.

Methods

Outbreak models are built by the multivariate technique of reduced rank regression using different datasets. Each model and an estimate of its error are then used to predict future outbreaks under different future conditions of climate and forest composition. Variation in predictions is calculated, and the variance is apportioned among the model components that contributed to the epistemic uncertainty in predictions.

Results

Projections of future outbreaks are highly uncertain under the range of input data and future conditions examined. Uncertainty is not uniformly distributed spatially; the average 75% confidence interval for outbreak duration is 10 years. Estimates of forest inventory for model building and choice of climate scenario for projections of future climate had the greatest contributions to predictions of outbreak duration and severity.

Conclusion

Predictions of future spruce budworm outbreaks are highly uncertain. More precise outbreak data with which to build a new outbreak model will have the biggest impact on reducing uncertainty. However, an uncertain future climate will continue to produce uncertainty in outbreak projections. Forest management strategies must, therefore, include alternatives that present a reasonable likelihood of achieving acceptable outcomes over a wide range of future conditions.

     

Estimates of the distributions of forest ecosystem model inputs for deciduous forests of eastern North America

Techniques for evaluating uncertainties in process-based, computer simulation models are evolving in response to the proliferation of such models and the demand for their use in the management of forest ecosystems. Many evaluation techniques require precise statements of the uncertainties associated with each model input. Statements of uncertainty are typically formulated as probability density functions (pdfs). Here, pdfs are developed for 29 inputs of the process-based, forest ecosystem, computer simulation model PnET-II, many of which are inputs to other well-known forest ecosystem models. The inputs considered describe vegetation characteristics of forests typical of the Eastern Deciduous Forest biome of North America. Data were compiled largely from published literature to estimate pdfs. The compiled distributions can be used to conduct various model evaluations including uncertainty assessment, calibration, and sensitivity analysis.     

Component carbon fluxes and their contribution to ecosystem carbon exchange in a pine forest: an assessment based on eddy covariance measurements and an integrated model

We used a combination of eddy flux, canopy, soil and environmental measurements with an integrated biophysical model to analyze the seasonality of component carbon (C) fluxes and their contribution to ecosystem C exchange in a 50-year-old Scots pine forest (Pinus sylvestris L.) in eastern Finland (62 degrees 47' N, 30 degrees 58' E) over three climatically contrasting years (2000-2002). Eddy flux measurements showed that the growing Scots pine forest was a sink for CO2, with annual net C uptakes of 131, 210 and 258 g C m-2> year-1 in 2000, 2001 and 2002, respectively. The integrated process model reproduced the annual course of daily C flux above the forest canopy as measured by the eddy covariance method once the site-specific component parameters were estimated. The model explained 72, 66 and 68% of the variation in daily net C flux in 2000, 2001 and 2002, respectively. Modeled annual C loss by respiration was 565, 629 and 640 g C m-2 year-1, accounting for 77, 77 and 65% of annual gross C uptake, respectively. Carbon fluxes from the forest floor were the dominant contributors to forest ecosystem respiration, with the fractions of annual respiration from the forest floor, foliage and wood being 46-62, 27-44 and 9-10%, respectively. The wide range in daily net C uptake during the growing season was largely attributable to day-to-day fluctuations in incident quantum irradiance. During just a few days in early spring and late autumn, ecosystem net C exchange varied between source and sink as a result of large daily changes in temperature. The forest showed a greater reduction in gross C uptake by photosynthesis than in C loss by respiration during the dry summer of 2000, indicating that interannual variability in ecosystem net C uptake at this site was modified mostly by summer rainfall and vapor pressure deficit.     

Ecosystem-scale biosphere-atmosphere interactions of a hemiboreal mixed forest stand at Järvselja, Estonia

During two measurement campaigns, from August to September 2008 and 2009, we quantified the major ecosystem fluxes in a hemiboreal forest ecosystem in Järvselja, Estonia. The main aim of this study was to separate the ecosystem flux components and gain insight into the performance of a multi-species multi-layered tree stand. Carbon dioxide and water vapor fluxes were measured using the eddy covariance method above and below the canopy in conjunction with the microclimate. Leaf and soil contributions were quantified separately by cuvette and chamber measurements, including fluxes of carbon dioxide, water vapor, nitrogen oxides, nitrous oxide, methane, ozone, sulfur dioxide, and biogenic volatile organic compounds (isoprene and monoterpenes). The latter have been as well characterized for monoterpenes in detail. Based on measured atmospheric trace gas concentrations, the flux tower site can be characterized as remote and rural with low anthropogenic disturbances.Our results presented here encourage future experimental efforts to be directed towards year round integrated biosphere-atmosphere measurements and development of process-oriented models of forest-atmosphere exchange taking the special case of a multi-layered and multi-species tree stand into account. As climate change likely leads to spatial extension of hemiboreal forest ecosystems a deep understanding of the processes and interactions therein is needed to foster management and mitigation strategies.     

Using a chain of models to simulate the distribution between quality clusters of a set of boards coming from a sessile oak (<Emphasis Type="Italic">Quercus petraea</Emphasis> Liebl.) resource. Validation of the simulations

Taking sessile oak as an example, this paper initially presents a method to predict the final production (quantity and quality) coming from a forest resource when two sets of data are available. The data sets are from two models: measured or simulated ring width profiles from pith to bark of the constituent trees as well as a mixed model for the basic wood properties which are used to grade the boards into quality clusters. The second part of the paper contains a validation for the proposed method. Simulations are used to predict two basic wood properties (volumetric swelling coefficient and wood density) in the trees of a forest resource in relation to the ring width profile of each tree. The simulations are used to compute a map of these two basic properties in each plank derived from the trees. A quality index derived from this map of basic wood properties in the boards is then used to allocate the planks to quality clusters. The basic wood properties considered in this paper are modelled with linear mixed models. Since computation of the plank properties or definition of the grading rule can use several properties simultaneously, the models used to simulate the basic properties are joint models. Modelling jointly several properties with a mixed model consists of defining a covariance structure between the random effects of the model. Such a model can be substantial in terms of parameters and computational resources required, thus we compared three kinds of joint models. The simplest one is not quite a joint model but is simply obtained from the juxtaposition of independent models, one for each of the two properties taken into consideration. We also defined a model with a moderate covariance structure between the two properties, and lastly, we used a third model with a full covariance structure. Simulations of volumetric swelling coefficient, wood density and the resulting board grading were carried out with each of these three models. All give results roughly in accordance with the observations, but the two truly joint models give better results than the simplest model.     

A Process for Improving Wildlife Habitat Models for Assessing Forest Ecosystem Health

Vertebrate wildlife will probably continue to be a primary surrogate for assessing biological diversity in forested ecosystems. However, assessment tools such as wildlife-habitat models generally have proved to be poor predictors of wildlife population responses to landscape-scale changes in forest ecosystems. Forest ecosystem assessment therefore will require improved models. To improve modeling capabilities, scientists must clarify the primary determinants of wildlife habitat selection, which is a behavioral process that links wildlife populations with ecosystem processes. Wildlife populations respond to functional redundancies caused by multiple interactions among landforms, soils, and vegetation. Therefore, probing wildlife habitat selection responses to attributes of landforms, soils, and vegetation should result in improved wildlife-habitat models. In this paper, radiotelemetry data from a study on northern spotted owls (Strix occidentalis caurina) are used to illustrate how remote sensing and geographic information systems (GIs) analysis might clarify basic determinants of habitat selection.     

湖南会同杉木人工林生态系统CO_2通量特征

利用开路式涡动相关系统与自动气象梯度观测系统2008年12个月的观测数据,研究会同13年生杉木人工林CO2通量特征。结果表明:13年生杉木人工林生态系统CO2通量日变化存在明显的季节差异,晴天平均碳汇持续时间表现为夏>春>秋>冬,平均日较差表现为夏>秋>春>冬,最大碳汇出现时间由早到晚依次为夏、秋、春和冬;1年中,月累积碳通量除1和2月为碳源外,其他各月均表现为碳汇,碳汇最大值出现在6月(-53.0gC·m-2);13年生杉木林的年碳汇总量为-255.3gC·m-2。白天CO2通量与光合有效辐射的关系可用Michaelis-Menten模型模拟(P<0.05),但模型参数随温度而异;夜间CO2通量与5cm土壤温度呈指数关系(P<0.05)。