市井 和仁

We present a synthesis of the land-atmosphere carbon flux from land use and land cover change (LULCC) in Asia using multiple data sources and paying particular attention to deforestation and forest regrowth fluxes. The data sources are quasi-independent and include the U.N. Food and Agriculture Organization-Forest Resource Assessment (FAO-FRA 2015 country-level inventory estimates), the Emission Database for Global Atmospheric Research (EDGARv4.3), the 'Houghton' bookkeeping model that incorporates FAO-FRA data, an ensemble of 8 state-of-the-art Dynamic Global Vegetation Models (DGVM), and 2 recently published independent studies using primarily remote sensing techniques. The estimates are aggregated spatially to Southeast, East, and South Asia and temporally for three decades, 1980-1989, 1990-1999 and 2000-2009. Since 1980, net carbon emissions from LULCC in Asia were responsible for 20%-40% of global LULCC emissions, with emissions from Southeast Asia alone accounting for 15%-25% of global LULCC emissions during the same period. In the 2000s and for all Asia, three estimates (FAO-FRA, DGVM, Houghton) were in agreement of a net source of carbon to the atmosphere, with mean estimates ranging between 0.24 to 0.41 Pg C yr-1, whereas EDGARv4.3 suggested a net carbon sink of -0.17 Pg C yr-1. Three of 4 estimates suggest that LULCC carbon emissions declined by at least 34% in the preceding decade (1990-2000). Spread in the estimates is due to the inclusion of different flux components and their treatments, showing the importance to include emissions from carbon rich peatlands and land management, such as shifting cultivation and wood harvesting, which appear to be consistently underreported.

The latitudinal gradient of the start of the growing season (SOS) and the end of the growing season (EOS) were quantified in Alaska (61 degrees N to 71 degrees N) using satellite-based and ground-based datasets. The Alaskan evergreen needleleaf forests are sparse and the understory vegetation has a substantial impact on the satellite signal. We evaluated SOS and EOS of understory and tundra vegetation using time-lapse camera images. From the comparison of three SOS algorithms for determining SOS from two satellite datasets (SPOT-VEGETATION and Terra-MODIS), we found that the satellite-based SOS timing was consistent with the leaf emergence of the forest under story and tundra vegetation. The ensemble average of SOS over all satellite algorithms can be used as a measure of spring leaf emergence for understory and tundra vegetation. In contrast, the relationship between the ground-based and satellite-based EOSs was not as strong as that of SOS both for boreal forest and tundra sites because of the large biases between those two EOSs (19 to 26 days). The satellite-based EOS was more relevant to snowfall events than the senescence of understory or tundra. The plant canopy radiative transfer simulation suggested that 84-86% of the NDVI seasonal amplitude could be a reasonable threshold for the EOS determination. The latitudinal gradients of SOS and EOS evaluated by the satellite and ground data were consistent and the satellite derived SOS and EOS were 3.5 to 5.7 days degree(-1) and -2.3 to -2.7 days degree(-1), which corresponded to the spring (May) temperature sensitivity of -2.5 to -3.9 days degrees C-1 in SOS and the autumn (August and September) temperature sensitivity of 3.0 to 4.6 days degrees C-1 in EOS. This demonstrates the possible impact of phenology in spruce forest understory and tundra ecosystems in response to climate change in the warming Artic and sub-Arctic regions. (C) 2016 The Authors. Published by Elsevier Inc.

Greenhouse gases Observing SATellite (GOSAT) is the operational satellite dedicated to atmospheric CO2 observations. Assimilation of data provided by GOSAT is expected to yield reliable CO2 fluxes in semi-arid regions because of frequent observations owing to clear skies. Here we estimated net CO2 flux over semi-arid regions of the Southern Hemisphere using the GOSAT column averaged CO2 (X-CO2) and surface CO2 measurements. Assimilation of GOSAT X-CO2 indicated that semi-arid regions are integral components of recent terrestrial CO2 uptake, accounting for 44% globally. Compared with estimates assimilated from surface measurements, estimates by GOSAT X-CO2 suggest a 50% reduction in the semi-arid CO2 uptake, amounting to 1.1 Pg C yr(-1). Significant estimation differences occurred for South America and South Africa, where the GOSAT makes frequent measurements but where surface CO2 measurements are limited. In comparison, the two estimates varied less in Australia, where more surface measurements are available. These results suggest that GOSAT X-CO2 is effective at regulating excess estimates of semi-arid CO2 uptake in regions that are less constrained by surface CO2 measurements. To promote understanding of climate change effects in semi-arid regions, it is important to continue monitoring trends in CO2 uptake with GOSAT.

Spatio-temporal fields of land-atmosphere fluxes derived from data-driven models can complement simulations by process-based land surface models. While a number of strategies for empirical models with eddy-covariance flux data have been applied, a systematic intercomparison of these methods has been missing so far. In this study, we performed a cross-validation experiment for predicting carbon dioxide, latent heat, sensible heat and net radiation fluxes across different ecosystem types with 11 machine learning (ML) methods from four different classes (kernel methods, neural networks, tree methods, and regression splines). We applied two complementary setups: (1) 8-day average fluxes based on remotely sensed data and (2) daily mean fluxes based on meteorological data and a mean seasonal cycle of remotely sensed variables. The patterns of predictions from different ML and experimental setups were highly consistent. There were systematic differences in performance among the fluxes, with the following ascending order: net ecosystem exchange (R-2 < 0.5), ecosystem respiration (R-2 > 0.6), gross primary production (R-2 > 0.7), latent heat (R-2 > 0.7), sensible heat (R-2 > 0.7), and net radiation (R-2 > 0.8). The ML methods predicted the across-site variability and the mean seasonal cycle of the observed fluxes very well (R 2 > 0.7), while the 8-day deviations from the mean seasonal cycle were not well predicted (R-2 < 0.5). Fluxes were better predicted at forested and temperate climate sites than at sites in extreme climates or less represented by training data (e.g., the tropics). The evaluated large ensemble of ML-based models will be the basis of new global flux products.

The Great East Japan Earthquake and Tsunami on 11 March 2011 led to the Fukushima Daiichi nuclear disaster. The Japanese government subsequently outlined an evacuation zone around the power plant, and all residents were evacuated. In the absence of cropland or urban vegetation management, land cover was expected to change. The changes in vegetation cover following the nuclear disaster are presented using long-term time series data obtained from Moderate Resolution Imaging Spectroradiometer (MODIS) sensors. Utilizing MODIS 250m spatial resolution observations, clear signals of vegetation changes were detected following the disaster in 2011. The areas affected were non-forest regions (mostly paddy fields) within the 20km radius evacuation zone around the power plant. Multi-year comparisons of vegetation seasonality indicated that the changes can be explained by the natural succession of abandoned cropland. The affected area within the 20km radius is equivalent to about 20% of the total area affected by the tsunami.

We examined the consistency between terrestrial biosphere fluxes (terrestrial CO2 exchanges) from data-driven top-down (GOSAT CO2 inversion) and bottom-up (empirical eddy flux upscaling based on a support vector regression (SVR) model) approaches over 42 global terrestrial regions from June 2009 to October 2011. Seasonal variations of the biosphere fluxes by the two approaches agreed well in boreal and temperate regions across the Northern Hemisphere. Both fluxes also exhibited strong anomalous signals in response to contrasting anomalous spring temperatures observed in North America and boreal Eurasia. This indicates that the CO2 concentration data integrated in the GOSAT inversion and the meteorological and vegetation data in the SVR models are equally effective in producing spatiotemporal variations of biosphere flux. Meanwhile, large differences in seasonality were found in subtropical and tropical South America, South Asia, and Africa. The GOSAT inversion showed seasonal variations that pivoted around CO2 neutral, while the SVR model showed seasonal variations that tended toward CO2 sink. Thus, a large difference in CO2 budget was identified between the two approaches in subtropical and tropical regions across the Southern Hemisphere. Examination of the integrated data revealed that the large tropical sink of CO2 by the SVR model was an artifact due to the underrepresented biosphere fluxes predicted by limited eddy flux data for tropical biomes. Because of the global coverage of CO2 concentration data, the GOSAT inversion provides better estimates of continental CO2 flux than the SVR model in the Southern Hemisphere.

We investigated carbon allocation in a cool-temperate forest in central Japan in years of contrasting climate anomalies: an early spring warming induced by the El Nino Southern Oscillation in 2002 and a low summer photosynthetic photon flux density (PPFD) induced by a stationary rain front in 2003. Observations of eddy flux, biometric variables, and chamber measurements from 1999 to 2006 and interannual variations in fine root net primary production (frNPP) were analyzed in conjunction with a terrestrial biosphere model simulation with multiple biometric constraints. Compared to the annual means (excluding 2002 and 2003), the low summer PPFD in 2003 reduced the annual gross primary productivity (GPP; 6%), soil respiration (SR; -11%), and ecosystem respiration (RE; -5%). Under the low summer PPFD, CO2 fluxes commonly decreased but components of the NPP were not affected. The low variation in NPP is explained by previous findings that NPP is more sensitive to climate conditions before or during the early stage of the growing season. The early spring warming in 2002 increased the GPP (8%) and woody tissue NPP (wNPP; 55%) and decreased the frNPP (-33%) and SR (-6%). Although early spring warming prolonged the growing season, the foliage NPP (fNPP) and litterfall were invariant. The increase in wNPP and the decrease in frNPP imply that the forest decreased frNPP in favor of wNPP under the high spring temperature. Although the frNPP was estimated by model-data integration, we argue that the decrease in frNPP is plausible because the decrease in SR cannot be explained without the contribution from fine root respiration. These results suggest that increasing or decreasing patterns of component fluxes cannot necessarily be inferred from the GPP. Factors such as the nature and duration of climate anomalies and allocation shift between components of the NPP may need to be considered when characterizing carbon allocation under anomalous climate events. (C) 2014 Elsevier B.V. All rights reserved.

Larch (Larix spp.) forests are predominantly distributed across high latitudes of Eurasia. They potentially have a strong influence on the terrestrial carbon and energy cycles, because of their vast area and the large carbon stocks in their peat soils in the permafrost. In this study, we elucidated intersite variation of ecosystem photosynthetic and respiratory parameters of eight larch forests in East Asia using the CarboEastAsia carbon flux and micrometeorology dataset. These parameters were determined using the empirical relationship between the carbon fluxes (photosynthesis and respiration) and micrometeorological variables (light and temperature). In addition, we examined leaf area index (LAI) determined by Moderate Resolution Imaging Spectroradiometer (MODIS) remote sensing data to explain the intersite variation. Linear or exponential relationships with annual mean temperature or seasonal maximum LAI at the study sites were found for the annual carbon fluxes (gross primary production [GPP] and total ecosystem respiration [RE]) as well as for four of the five seasonal maximum values of determined photosynthetic and respiratory parameters (maximum GPP at light saturation, initial slope of the light-response curve, daytime respiration, and RE at the reference temperature of 10 degrees C). Phenological indices, such as start day of the growing season, growing season length and growing season degree days explained much of the intersite variation of GPP and RE of the studied larch forests; however, the relationship between MODIS LAI and photosynthetic or respiratory parameters implies that the intersite variation in GPP and RE was caused not only by the temperature variation (abiotic factor), but also by the variation in the photosynthetic and respiration activity by vegetation (biotic factor) through the change in leaf (or whole vegetation) biomass. Our analysis shows that MODIS LAI serves as a good index to explain the variation of the ecosystem photosynthetic and respiratory characteristics of East Asian larch forests.

As part of the terrestrial branch of the Japan-funded Arctic Climate Change Research Project (GRENE-TEA), which aims to clarify the role and function of the terrestrial Arctic in the climate system and assess the influence of its changes on a global scale, this model intercomparison project (GTMIP) is designed to (1) enhance communication and understanding between the modelling and field scientists and (2) assess the uncertainty and variations stemming from variability in model implementation/design and in model outputs using climatic and historical conditions in the Arctic terrestrial regions. This paper provides an overview of all GTMIP activity, and the experiment protocol of Stage 1, which is site simulations driven by statistically fitted data created using the GRENE-TEA site observations for the last 3 decades. The target metrics for the model evaluation cover key processes in both physics and biogeochemistry, including energy budgets, snow, permafrost, phenology, and carbon budgets. Exemplary results for distributions of four metrics (annual mean latent heat flux, annual maximum snow depth, gross primary production, and net ecosystem production) and for seasonal transitions are provided to give an outlook of the planned analysis that will delineate the inter-dependence among the key processes and provide clues for improving model performance.

Warming in northern high latitudes has changed the energy balance between terrestrial ecosystems and the atmosphere. This study evaluated changes in regional surface energy exchange in Alaska from 2000 to 2011 when substantial declines in spring snow cover due to spring warming and large-scale fire events were observed. Energy fluxes from a network of 20 eddy covariance sites were upscaled using a support vector regression (SVR) model, by combining satellite remote sensing data and global climate data. Based on site-scale analysis, SVR reproduced observed net radiation, sensible heat flux, latent heat flux, and ground heat flux; 8 day root-mean-square errors for these variables were 15, 10, 9, and 3Wm(-2), respectively. Based on upscaled fluxes, decreases in spring snow cover induced an increase in surface net radiation, a net heating effect, of 0.56Wm(-2) decade(-1). This heating effect was comparable to the net cooling effect due to increased fire extent during the study period (up to 0.59Wm(-2) decade(-1)). These land cover effects were larger than the change in the energy forcing associated with CO2 balance for the Alaska region. Spring warming and postfire land cover change increased the regional latent heat flux. The regional sensible heat flux decreased with the postfire land cover change. Our results highlight the importance of positive spring snow albedo feedback to climate and a postfire negative feedback under the expected warming climate in the Arctic.

Evapotranspiration (E) in the Amazon connects forest function and regional climate via its role in precipitation recycling However, the mechanisms regulating water supply to vegetation and its demand for water remain poorly understood, especially during periods of seasonal water deficits In this study, we address two main questions: First, how do mechanisms of water supply (indicated by rooting depth and groundwater) and vegetation water demand (indicated by stomatal conductance and intrinsic water use efficiency) control evapotranspiration (E) along broad gradients of climate and vegetation from equatorial Amazonia to Cerrado, and second, how do these inferred mechanisms of supply and demand compare to those employed by a suite of ecosystem models? We used a network of eddy covariance towers in Brazil coupled with ancillary measurements to address these questions With respect to the magnitude and seasonality of E, models have much improved in equatorial tropical forests by eliminating most dry season water limitation, diverge in performance in transitional forests where seasonal water deficits are greater, and mostly capture the observed seasonal depressions in E at Cerrado However, many models depended universally on either deep roots or groundwater to mitigate dry season water deficits, the relative importance of which we found does not vary as a simple function of climate or vegetation In addition, canopy stomatal conductance (g's) regulates dry season vegetation demand for water at all except the wettest sites even as the seasonal cycle of E follows that of net radiation In contrast, some models simulated no seasonality in gs, even while matching the observed seasonal cycle of E. We suggest that canopy dynamics mediated by leaf phenology may play a significant role in such seasonality, a process poorly represented in models Model bias in gs and E, in turn, was related to biases arising from the simulated light response (gross primary productivity, GPP) or the intringic water use efficiency of photosynthesis (iWUE). We identified deficiencies in models which would not otherwise be apparent based on a simple comparison of simulated and observed rates of E. While some deficiencies can be remedied by parameter tuning, in most models they highlight the need for continued process development of belowground hydrology and in particular, the biological processes of root dynamics and leaf phenology, which via their controls on E, mediate vegetation-climate feedbacks in the tropics. (C) 2014 Elsevier B.V. All rights reserved.

A fundamental question connecting terrestrial ecology and global climate change is the sensitivity of key terrestrial biomes to climatic variability and change. The Amazon region is such a key biome: it contains unparalleled biological diversity, a globally significant store of organic carbon, and it is a potent engine driving global cycles of water and energy. The importance of understanding how land surface dynamics of the Amazon region respond to climatic variability and change is widely appreciated, but despite significant recent advances, large gaps in our understanding remain. Understanding of energy and carbon exchange between terrestrial ecosystems and the atmosphere can be improved through direct observations and experiments, as well as through modeling activities. Land surface/ecosystem models have become important tools for extrapolating local observations and understanding to much larger terrestrial regions. They are also valuable tools to test hypothesis on ecosystem functioning. Funded by NASA under the auspices of the LBA (the Large-Scale Biosphere-Atmosphere Experiment in Amazonia), the LBA Data Model Intercomparison Project (LBA-DMIP) uses a comprehensive data set from an observational network of flux towers across the Amazon, and an ecosystem modeling community engaged in ongoing studies using a suite of different land surface and terrestrial ecosystem models to understand Amazon forest function. Here an overview of this project is presented accompanied by a description of the measurement sites, data, models and protocol. (C) 2013 Elsevier B.V. All rights reserved.

This study analyzes the inter-annual variability (IAV) of simulations of 21 different land surface model formulations, driven by meteorological conditions measured at 8 flux towers, located in rain forest, forest-savanna ecotone and pasture sites in Amazonia, and one in savanna site in Southeastern Brazil. Annual totals of net ecosystem exchange (NEE) of carbon and evapotranspiration (ET), measured and simulated by each model for each site-year, were compared in terms of year-to-year variability and possible relation to climate drivers. Results have shown that most of models simulations for annual totals of NEE and ET, and IAV of these fluxes, are frequently different from measurements. The average of the model simulations of annual fluxes tend to respond to climatic drivers similarly to the observations, but with noticeable discrepancies. Annual measurements of NEE are negatively correlated to annual rainfall in the forest sites group. Although the ensemble of all models yields a similar result, only three model formulations reproduce a significant negative correlation of simulated NEE with rainfall. For the IAV of ET, tower measurements are controlled by annual variations of radiation and this feature is captured by the ensemble of the models, both at individual sites and when all forest sites are grouped. However, simulated ET values are also significantly correlated to the amount of precipitation in many models and in the model ensemble, while there is no significant correlation in the observations. In general, the surface models are able to reproduce the responses of fluxes to climatic drivers, but improvements are still needed to better capture their inter-annual variability. (C) 2013 Elsevier B.V. All rights reserved.

Past changes in gross primary productivity (GPP) were assessed using historical satellite observations based on the Normalized Difference Vegetation Index (NDVI) from the Advanced Very High Resolution Radiometer (AVHRR) onboard the National Oceanic and Atmospheric Administration (NOAA) satellite series and four terrestrial biosphere models to identify the trends and driving mechanisms related to GPP and NDVI in Asia. A satellite-based time-series data analysis showed that approximately 40% of the area has experienced a significant increase in the NDVI, while only a few areas have experienced a significant decreasing trend over the last 30 years. The increases in the NDVI are dominant in the sub-continental regions of Siberia, East Asia, and India. Simulations using the terrestrial biosphere models also showed significant increases in GPP, similar to the results for the NDVI, in boreal and temperate regions. A modeled sensitivity analysis showed that the increases in GPP are explained by increased temperature and precipitation in Siberia. Precipitation, solar radiation and CO2 fertilization are important factors in the tropical regions. However, the relative contributions of each factor to GPP changes are different among the models.

The Tibetan plateau plays an important role in energy and carbon cycles by providing an elevated heat source and by storing a large amount of soil carbon due to low temperature. The main vegetation of the plateau is alpine grassland. This study evaluates performance of Community Land Model 3.5 with carbon and nitrogen cycles (CLM3.5CN) over a alpine grassland in the Tibetan plateau in terms of energy and carbon fluxes in conditions of reasonable phenology and initial carbon pool comparable to observations. Comparison between model and observation shows following features. The model captures the magnitude of maximum leaf area index (LAI) but underestimats leaf mass. Net ecosystem exchange (NEE) is significantly underestimated during the growing season and soil temperature is also underestimated throughout a year with higher negative bias in winter than in other seasons. In order to examine the cause of the model deficiencies, we design four sensitivity tests: seasonal mulch; shallow rooting depth; reduction of critical soil moisture to limit the decomposition rate; smaller specific leaf area (SLA). Considering seasonal mulch improves the negative bias of soil temperature during dormant season has little effect on the NEE during the growing seasson. Underestimation of NEE during the growing season is partly due to underestimated decomposition rate which results from underestimated soil temperature and deep root placement in the soil column. Underestimation of latent heat flux during summer is partly due to use of large SLA in the model. Other deficiencies are also discussed.

The process of confining unnecessary freedom is a step toward advanced ecosystem modeling. This study demonstrates the importance of carbon flux and biometric observation in constraining a terrestrial ecosystem model with a simple optimization scheme. At the selected sites from AsiaFlux network, a simultaneous optimization scheme for both carbon flux and biomass was compared with carbon flux-oriented and biomass-oriented optimization schemes using the Biome-BGC model. The optimization scheme oriented to either carbon flux or biomass provided simulation results that were consistent with observations, but with reduced performance in unconstrained variables. The simultaneous optimization scheme yielded results that were consistent with observations for both carbon flux and biomass. By comparing long-term projections simulated by three schemes, it was found that the optimization oriented only to carbon flux has limited performance because misrepresented biomass significantly affected a projection of carbon exchange through heterotrophic respiration. From these experiments, we found that (1) a process model like Biome-BGC is capable of reproducing both carbon flux and biomass within acceptable proximity, (2) constraining biomass is importance not just because it is one of carbon cycle components, but also it significantly affects simulations of carbon flux. Thus, it is important to invest more effort to improve simulation of biomass as well as carbon flux.

Carbon dioxide (CO2) fluxes from a network of 21 eddy covariance towers were upscaled to estimate the Alaskan CO2 budget from 2000 to 2011 by combining satellite remote sensing data, disturbance information, and a support vector regression model. Data were compared with the CO2 budget from an inverse model (CarbonTracker). Observed gross primary productivity (GPP), ecosystem respiration (RE), and net ecosystem exchange (NEE) were each well reproduced by the model on the site scale; root-mean-square errors (RMSEs) for GPP, RE, and NEE were 0.52, 0.23, and 0.48g C m(-2) d(-1), respectively. Landcover classification was the most important input for predicting GPP, whereas visible reflectance index of green ratio was the most important input for predicting RE. During the period of 2000-2011, predicted GPP and RE were 36922 and 36212 Tg C yr(-1) (meaninterannual variability) for Alaska, respectively, indicating an approximately neutral CO2 budget for the decade. CarbonTracker also showed an approximately neutral CO2 budget during 2000-2011 (growing season RMSE=14g C m(-2) season(-1); annual RMSE=13g C m(-2) yr(-1)). Interannual CO2 flux variability was positively correlated with air temperature anomalies from June to August, with Alaska acting as a greater CO2 sink in warmer years. CO2 flux trends for the decade were clear in disturbed ecosystems; positive trends in GPP and CO2 sink were observed in areas where vegetation recovered for about 20 years after fire.

Based on the model-data comparison at the eddy-covariance observation sites from CarboEastAsia datasets, we report the current status of the terrestrial carbon cycle modeling in monsoon Asia. In order to assess the modeling performance and discuss future requirements for both modeling and observation efforts in Asia, we ran eight terrestrial biosphere models at 24 sites from 1901 to 2010. By analyzing the modeled carbon fluxes against the CarboEastAsia datasets, the strengths and weaknesses of terrestrial biosphere modeling over Asia were evaluated. In terms of pattern and magnitude, the carbon fluxes (i.e., gross primary productivity, ecosystem respiration, and net ecosystem exchange) at the temperate and boreal forest sites were simulated best, whereas the simulation results from the tropical forest, cropland, and disturbed sites were poor. The multi-model ensemble mean values showed lower root mean square errors and higher correlations, suggesting that composition of multiple terrestrial biosphere models would be preferable for terrestrial carbon budget assessments in Asia. These results indicate that the current model-based estimation of terrestrial carbon budget has large uncertainties, and future research should further refine the models to permit re-evaluation of the terrestrial carbon budget.

The role of disturbance and climate factors in determining the forest carbon balance was investigated at a Japanese cypress forest in central Japan with eddy flux measurements, tree-ring analyses, and a terrestrial biosphere model. The forest was established as a plantation after intermittent harvesting and replanting between 1959 and 1977, and acted as a strong carbon sink of approximately 500 g C m(-2) year(-1) for the measurement years between 2001 and 2007. A terrestrial biosphere model. BIOME-BGC, was validated using the eddy flux data at daily to interannual timescales, and the tree-ring width data at interannual to decadal timescales. According to the model simulation, during the observation period 270 +/- 55 g C m(-2) year(-1) was additionally sequestered due to the indirect effects of the harvesting and planting, whereas the increase of CO(2) concentration and the change in climate increased the sink of 110 +/- 40 and 30 +/- 80 g C m(-2) year(-1), respectively. The model simulation shows that the forest is now recovering from harvesting, and that harvesting is a more important determinant of the current carbon sink than either interannual climate anomalies or increased atmospheric CO(2) concentration. We found that harvesting with long rotation length could be effective management activity in order to increase carbon sequestration, if the harvested timber is converted into products with long lifecycles. (C) 2011 Elsevier B.V. All rights reserved.

The terrestrial carbon cycle is strongly affected by natural phenomena, terrain heterogeneity, and human-induced activities that alter carbon exchange via vegetation and soil activities. In order to accurately understand terrestrial carbon cycle mechanisms, it is necessary to estimate spatial and temporal variations in carbon flux and storage using process-based models with the highest possible resolution. We estimated terrestrial carbon fluxes using a biosphere model integrating eco-physiological and mechanistic approaches based on satellite data (BEAMS) and observations with 1-km grid resolution. The study area is the central Far East Asia region, which lies between 30 degrees and 50 degrees north latitude and 125 and 150 east longitude. Aiming to simulate terrestrial carbon exchanges under realistic land surface conditions, we used as many satellite-observation datasets as possible, such as the standard MODIS, TRMM, and SRTM high-level land products. Validated using gross primary productivity (GPP), net ecosystem production (NEP), net radiation and latent heat with ground measurements at six flux sites, the model estimations showed reasonable seasonal and annual patterns. In extensive analysis, the total GPP and NPP were determined to be 2.1 and 0.9 PgC/year, respectively. The total NEP estimation was + 5.6 TgC/year, meaning that the land area played a role as a carbon sink from 2001 to 2006. In analyses of areas with complicated topography, the 1-km grid estimation could prove to be effective in evaluating the effect of landscape on the terrestrial carbon cycle. The method presented here is an appropriate approach for gaining a better understanding of terrestrial carbon exchange, both spatially and temporally. (C) 2011 Elsevier Inc. All rights reserved.

This paper examines carbon stocks and their relative balance in terrestrial ecosystems simulated by Biome-BGC, LPJ, and CASA in an ensemble model experiment conducted using the Terrestrial Observation and Prediction System. We developed the Hierarchical Framework for Diagnosing Ecosystem Models to separate the simulated biogeochemistry into a cascade of functional tiers and examine their characteristics sequentially. The analyses indicate that the simulated biomass is usually two to three times higher in Biome-BGC than LPJ or CASA. Such a discrepancy is mainly induced by differences in model parameters and algorithms that regulate the rates of biomass turnover. The mean residence time of biomass in Biome-BGC is estimated to be 40-80 years in temperate/moist climate regions, while it mostly varies between 5 and 30 years in CASA and LPJ. A large range of values is also found in the simulated soil carbon. The mean residence time of soil carbon in Biome-BGC and LPJ is similar to 200 years in cold regions, which decreases rapidly with increases of temperature at a rate of similar to 10 yr degrees C-1. Because long-term soil carbon pool is not simulated in CASA, its corresponding mean residence time is only about 10-20 years and less sensitive to temperature. Another key factor that influences the carbon balance of the simulated ecosystem is disturbance caused by wildfire, for which the algorithms vary among the models. Because fire emissions are balanced by net ecosystem production (NEP) at steady states, magnitudes, and spatial patterns of NEP vary significantly as well. Slight carbon imbalance may be left by the spin-up algorithm of the models, which adds uncertainty to the estimated carbon sources or sinks. Although these results are only drawn on the tested model versions, the developed methodology has potential for other model exercises.

We conducted an ensemble modeling exercise using the Terrestrial Observation and Prediction System (TOPS) to evaluate sources of uncertainty in carbon flux estimates resulting from structural differences among ecosystem models. The experiment ran public-domain versions of biome-bgc, lpj, casa, and tops-bgc over North America at 8 km resolution and for the period of 1982-2006. We developed the Hierarchical Framework for Diagnosing Ecosystem Models (HFDEM) to separate the simulated biogeochemistry into a cascade of three functional tiers and sequentially examine their characteristics in climate (temperature-precipitation) and other spaces. Analysis of the simulated annual gross primary production (GPP) in the climate domain indicates a general agreement among the models, all showing optimal GPP in regions where the relationship between annual average temperature (T, degrees C) and annual total precipitation (P, mm) is defined by P=50T+500. However, differences in simulated GPP are identified in magnitudes and distribution patterns. For forests, the GPP gradient along P=50T+500 ranges from similar to 50 g C yr-1 m-2 degrees C-1 (casa) to similar to 125 g C yr-1 m-2 degrees C-1 (biome-bgc) in cold/temperate regions; for nonforests, the diversity among GPP distributions is even larger. Positive linear relationships are found between annual GPP and annual mean leaf area index (LAI) in all models. For biome-bgc and lpj, such relationships lead to a positive feedback from LAI growth to GPP enhancement. Different approaches to constrain this feedback lead to different sensitivity of the models to disturbances such as fire, which contribute significantly to the diversity in GPP stated above. The ratios between independently simulated NPP and GPP are close to 50% on average; however, their distribution patterns vary significantly between models, reflecting the difficulties in estimating autotrophic respiration across various climate regimes. Although these results are drawn from our experiments with the tested model versions, the developed methodology has potential for other model exercises.

Improvement of terrestrial submodels in Earth system models (ESMs) is important to reduce uncertainties in future projections of global carbon cycle and climate. Since these submodels lack detailed validation, evaluation of terrestrial submodels using networks of field observations is necessary. The purpose of this study is to improve an ESM by refining a terrestrial submodel using eddy covariance observations. We evaluated the terrestrial submodel (MOSES2/TRIFFID) included in the UVic Earth System Climate Model (UVic-ESCM) and tested the effects of terrestrial submodel improvements on future projection of carbon cycle and climate. First, we evaluated the terrestrial submodel as an off-line mode at point scales using 48 eddy covariance observation data, and improved it through fixing model parameters and structures. The terrestrial submodel was improved with the reduction of the root mean square error and the closer simulation of the seasonal carbon fluxes. Second, using the UVic-ESCM with the improved terrestrial submodel, we confirmed model improvement at most observation sites. The terrestrial submodel refinement also affected future projections; the UVic-ESCM with the improved terrestrial submodel simulated 100 ppmv lower atmospheric CO(2) concentration in 2100 compared with the default UVic-ESCM. Our study underscores the importance of refinement of terrestrial submodels in ESM simulations.

Scaling up of observed point data to estimate regional carbon fluxes is an important issue in the context of the global terrestrial carbon cycle. In this study, the authors proposed a new model to scale up the eddy covariance data to estimate regional carbon fluxes using satellite-derived data. Gross primary productivity (GPP) and ecosystem respiration (RE) were empirically calculated using the normalized difference vegetation index (NDVI) and land surface temperature (LST) from the Moderate Resolution Imaging Spectroradiometer (MODIS). First, the model input is evaluated by comparing with the field data, then established and tested the model at the point scale, and then extended it into a regional scale. At the point scale, the empirical model could reproduce the seasonal and interannual variations in the carbon budget of the mature black spruce forests in Alaska and Canada sites, suggesting that seasonality of the NDVI and LST could explain the carbon fluxes and that the model is robust within mature black spruce forests in North America. Regional-scale analysis showed that the total GPP and RE between 2003 and 2006 were 1.76 +/- 0.28 and 1.86 +/- 0.26 kg CO(2) m-(2) yr(-1), respectively, in mature black spruce forests in Alaska, indicating that these forests were almost carbon neutral. The authors' model analysis shows that the proposed method is effective in scaling up point observations to estimate the regional-scale carbon budget and that the mature black spruce forests increased in sink strength during spring warming and decreased in sink strength during summer and autumn warming.

We used terrestrial ecosystem models to estimate spatial and temporal variability in and uncertainty of estimated soil carbon dioxide (CO2) efflux, or soil respiration, over the Japanese Archipelago. We compared five carbon-cycle models to assess inter-model variability: Biome-BGC, CASA, LPJ, SEIB, and VISIT. These models differ in approaches to soil carbon dynamics, root respiration estimation, and relationships between decomposition and environmental factors. We simulated the carbon budget of natural ecosystems over the archipelago for 2001-2006 at 1-day time steps and 2-min (latitude and longitude) spatial resolution. The models were calibrated using measured flux data to accurately represent net ecosystem CO2 exchange. Each model successfully reproduced seasonal changes and latitudinal gradients in soil respiration. The five-model average of estimated total soil respiration of Japanese ecosystems was 295 Tg C year(-1), with individual model estimates ranging from 210 to 396 Tg C year(-1) (1 Tg = 10(12) g). The differences between modeled estimates were more evident in summer and in warmer years, implying that they were mainly attributable to differences in modeling the temperature dependence of soil respiration. There was a large discrepancy between models in the estimated contribution of roots to total soil respiration, ranging from 3.9 to 48.4%. Although model calibration reduced the uncertainty of flux estimates, substantial uncertainties still remained in estimates of underground processes from these terrestrial carbon-cycle models.

Larch forests are widely distributed across many cool-temperate and boreal regions, and they are expected to play an important role in global carbon and water cycles. Model parameterizations for larch forests still contain large uncertainties owing to a lack of validation. In this study, a process-based terrestrial biosphere model, BIOME-BGC, was tested for larch forests at six AsiaFlux sites and used to identify important environmental factors that affect the carbon and water cycles at both temporal and spatial scales. The model simulation performed with the default deciduous conifer parameters produced results that had large differences from the observed net ecosystem exchange (NEE), gross primary productivity (GPP), ecosystem respiration (RE), and evapotranspiration (ET). Therefore, we adjusted several model parameters in order to reproduce the observed rates of carbon and water cycle processes. This model calibration, performed using the AsiaFlux data, substantially improved the model performance. The simulated annual GPP, RE, NEE, and ET from the calibrated model were highly consistent with observed values. The observed and simulated GPP and RE across the six sites were positively correlated with the annual mean air temperature and annual total precipitation. On the other hand, the simulated carbon budget was partly explained by the stand disturbance history in addition to the climate. The sensitivity study indicated that spring warming enhanced the carbon sink, whereas summer warming decreased it across the larch forests. The summer radiation was the most important factor that controlled the carbon fluxes in the temperate site, but the VPD and water conditions were the limiting factors in the boreal sites. One model parameter, the allocation ratio of carbon between belowground and aboveground, was site-specific, and it was negatively correlated with the annual climate of annual mean air temperature and total precipitation. Although this study substantially improved the model performance, the uncertainties that remained in terms of the sensitivity to water conditions should be examined in ongoing and long-term observations.

Northern Eurasia experienced anomalous weather conditions in the 2003 summer. We examined how forest ecosystems responded to the meteorological anomalies during the period using the dataset collected at flux monitoring sites in Asia, including a boreal forest in Mongolia, temperate forests in China and Japan, and a sub-tropical forest in China, as well as the dataset from satellite remote sensing. From July to August 2003, an active rain band stayed in the mid-latitude in East Asia for an unusually long period. Under the influence of the rain band, the Gross Primary Production (GPP), of temperate forests was 20-30% lower in the 2003 summer than in other years due to significant reduction in the Photosynthetic Photon Flux Density (PPFD). The GPP of a cool-temperate forest in the north of the rain band was slightly enhanced by the higher PPFD; however, the GPP of a sub-tropical forest located in the south of the rain band was reduced by drought stress due to extremely hot and dry conditions. The correlation coefficients for the year-to-year changes in the PPFD and GPP during mid-summer were calculated, and the spatial distribution was examined. The spatial pattern of the PPFD was calculated by satellite data, and that of the GPP was estimated by a regression-type model, which was trained and tested by ground observation data. The correlation was positive in the mid- and high-latitudes since light was an essential factor of the summer GPP. On the other hand, a negative correlation appeared in the lower latitudes, suggesting that the water limitation was much more important than the PPFD in the region. Our study illustrated that the integration of flux data from wide areas by combining satellite remote sensing data can help us gain an understanding of the ecosystem responses to large-scale meteorological phenomena.

Forest inventories from the intact rainforests of the Amazon indicate increasing rates of carbon gain over the past three decades. However, such estimates have been questioned because of the poor spatial representation of the sampling plots and the incomplete understanding of purported mechanisms behind the increases in biomass. Ecosystem models, when used in conjunction with satellite data, are useful in examining the carbon budgets in regions where the observations of carbon flows are sparse. The purpose of this study is to explain observed trends in normalized difference vegetation index (NDVI) using climate observations and ecosystem models of varying complexity in the western Amazon basin for the period of 1984-2002. We first investigated trends in NDVI and found a positive trend during the study period, but the positive trend in NDVI was observed only in the months from August to December. Then, trends in various climate parameters were calculated, and of the climate variables considered, only shortwave radiation was found to have a corresponding significant positive trend. To compare the impact of each climate component, as well as increasing carbon dioxide (CO2) concentrations, on evergreen forests in the Amazon, we ran three ecosystem models (CASA, Biome-BGC, and LPJ), and calculated monthly net primary production by changing a climate component selected from the available climate datasets. As expected, CO2 fertilization effects showed positive trends throughout the year and cannot explain the positive trend in NDVI, which was observed only for the months of August to December. Through these simulations, we demonstrated that the positive trend in shortwave radiation can explain the positive trend in NDVI observed for the period from August to December. We conclude that the positive trend in shortwave radiation is the most likely driver of the increasing trend in NDVI and the corresponding observed increases in forest biomass.

Accurate determination of rooting depths in terrestrial biosphere models is important for simulating terrestrial water and carbon cycles. In this study, we developed a method for optimizing rooting depth using satellite-based evapotranspiration (ET) seasonality and an ecosystem model by minimizing the differences between satellite-based and simulated ET. We then analyzed the impacts of rooting depth optimization on the simulated ET and gross primary production (GPP) seasonality in California, USA. First, we conducted a point-based evaluation of the methods against flux observations in California and tested the sensitivities of the simulated ET seasonality to the rooting depth settings. We then extended it spatially by estimating spatial patterns of rooting depth and analyzing the sensitivities of the simulated ET and GPP seasonalities to the rooting depth settings. We found large differences in the optimized and soil survey (STATSGO)-based rooting depths over the northern forest regions. In these regions, the deep rooting depths (>3 m) estimated in the study successfully reproduced the satellite-based ET seasonality, which peaks in summer, whereas the STATSGO-based rooting depth (<1.5 m) failed to sustain a high ET in summer. The rooting depth refinement also has large effects on simulated GPP; the annual GPP in these regions is increased by 50-100% due to sufficient soil water during the summer. In the grassy and shrubby regions of central and southern California, the estimated rooting depths are similar to those of STATSGO, probably due to the shallow rooting depth in these ecosystems. Our analysis suggests that setting a rooting depth is important for terrestrial ecosystem modeling and that satellite-based data could help both to estimate the spatial variability of rooting depths and to improve water and carbon cycle modeling. (C) 2009 Elsevier B.V. All rights reserved.

The increasing complexity of ecosystem models represents a major difficulty in tuning model parameters and analyzing simulated results. To address this problem, this study develops a hierarchical scheme that simplifies the Biome-BGC model into three functionally cascaded tiers and analyzes them sequentially. The first-tier model focuses on leaf-level ecophysiological processes; it simulates evapotranspiration and photosynthesis with prescribed leaf area index (LAI). The restriction on LAI is then lifted in the following two model tiers, which analyze how carbon and nitrogen is cycled at the whole-plant level (the second tier) and in all litter/soil pools (the third tier) to dynamically support the prescribed canopy. in particular, this study analyzes the steady state of these two model tiers by a set of equilibrium equations that are derived from Biome-BGC algorithms and are based on the principle of mass balance. Instead of spinning-up the model for thousands of climate years, these equations are able to estimate carbon/nitrogen stocks and fluxes of the target (steady-state) ecosystem directly from the results obtained by the first-tier model. The model hierarchy is examined with model experiments at four AmeriFlux sites. The results indicate that the proposed scheme can effectively calibrate Biome-BCC to simulate observed fluxes of evapotranspiration and photosynthesis; and the carbon/nitrogen stocks estimated by the equilibrium analysis approach are highly consistent with the results of model simulations. Therefore, the scheme developed in this study may serve as a practical guide to calibrate/analyze Biome-BGC; it also provides an efficient way to solve the problem of model spin-up, especially for applications over large regions. The same methodology may help analyze other similar ecosystem models as well. (C) 2009 Elsevier B.V. All rights reserved.

The AmeriFlux network of eddy covariance towers has played a critical role in the analysis of terrestrial water and carbon dynamics. It has been used to understand the general principles of ecosystem behaviors and to scale up those principles from sites to regions. To support the generalization from individual sites to large regions, it is essential that all major ecoregions in North America are represented in the AmeriFlux network. In this study, we examined the representativeness of the AmeriFlux network by comparing the climate and vegetation across the coterminous United States in 2004 with those at the AmeriFlux network in 2000-2004 on the basis of remote sensing products. We found that the AmeriFlux network generally captured the climatic and vegetation characteristics in the coterminous United States with under-representations in the Rocky Mountain evergreen needleleaf forest, the Sierra Nevada Mountains, the Sonora desert, the northern Great Plains, the Great Basin Desert, and New England. In terms of site representativeness, our analysis suggested that Indiana Morgan Monroe State Forest, Indiana, and Harvard Forest, Massachusetts, were among the forest sites with high representativeness extents; while Audubon Research Ranch, Arizona, and Sky Oaks Young Chaparral were among the nonforest sites with high representativeness extents.

The spatial and temporal dynamics of tropical forest functioning are poorly understood, partly attributed to a weak seasonality and high tree species diversity at the landscape scale. Recent neotropical rainforest studies with local tower flux measurements have revealed strong seasonal carbon fluxes that follow the availability of sunlight in intact forests, while in areas of forest disturbance, carbon fluxes more closely tracked seasonal water availability. These studies also showed a strong seasonal correspondence of satellite measures of greenness, using the Enhanced Vegetation index (E-VI) with ecosystem carbon fluxes in both intact and disturbed forests, which may enable larger scale extension of tower flux measurements. In this study, we investigated the seasonal patterns and relationships of local site tower flux measures of gross primary productivity (P-g) with independent Moderate Resolution Imaging Spectroradiometer (MODIS) satellite greenness measures across three Monsoon Asia tropical forest types, encompassing drought-deciduous, dry evergreen, and humid evergreen secondary tropical forests. In contrast to neotropical forests, the tropical forests of Monsoon Asia are more extensively degraded and heterogeneous due to intense land use pressures, and therefore, may exhibit unique seasonal patterns of ecosystem fluxes that are more likely water-limited and drought-susceptible. Our results show significant phenologic variability and response to moisture and light controls across the three tropical forest sites and at the regional scale. The drier tropical forests were primarily water-limited, while the wet evergreen secondary forest showed a slight positive trend with light availability. Satellite E-VI greenness observations were generally synchronized and linearly related with seasonal and inter-annual tower flux P-g measurements at the multiple sites and provided better opportunities for tower extension of carbon fluxes than other satellite products, such as the MODIS P-g product. Satellite E-VI-derived P-g images revealed strong seasonal variations in photosynthetic activity throughout the Monsoon Asia tropical region. (C) 2008 Elsevier B.V. All rights reserved.

Snow is important for water management, and an important component of the terrestrial biosphere and climate system. In this study, the snow models included in the Biome-BGC and Terrestrial Observation and Prediction System (TOPS) terrestrial biosphere models are compared against ground and satellite observations over the Columbia River Basin in the US and Canada and the impacts of differences ill snow models oil simulated terrestrial ecosystem processes are analysed. First, a point-based comparison of ground observations against model and satellite estimates of snow dynamics are conducted. Next, model and satellite snow estimates for the entire Columbia River Basin are compared. Then, using two different TOPS simulations, the default TOPS model (TOPS with TOPS snow model) and the TOPS model with the Biome-BGC snow model, the impacts of snow model selection oil runoff and gross primary production (GPP) are investigated. TOPS snow model predictions were consistent with ground and satellite estimates of seasonal and interannual variations in snow cover, snow water equivalent, and snow season length; however, in the Biome-BGC snow model, the snow pack melted too early, leading to extensive Underpredictions of snow season length and snow covered area. These biases led to earlier simulated peak runoff and reductions ill summer GPP, underscoring the need for accurate snow models within terrestrial ecosystem models. Copyright (c) 2007 John Wiley & Sons, Ltd.

Remote sensing is a potentially powerful technology with which to extrapolate eddy covariance-based gross primary production (GPP) to continental scales. In support of this concept, we used meteorological and flux data from the AmeriFlux network and Support Vector Machine (SVM), an inductive machine learning technique, to develop and apply a predictive GPP model for the conterminous U.S. In the following four-step process, we first trained the SVM to predict flux-based GPP from 33 AmeriFlux sites between 2000 and 2003 using three remotely-sensed variables (land surface temperature, enhanced vegetation index (EVI), and land cover) and one ground-measured variable (incident shortwave radiation). Second, we evaluated model performance by predicting GPP for 24 available AmeriFlux sites in 2004. In this independent evaluation, the SVM predicted GPP with a root mean squared error (RMSE) of 1.87 gC/m(2)/day and an R-2 of 0.71. Based on annual total GPP at 15 AmeriFlux sites for which the number of 8-day averages in 2004 was no less than 67%(30 out of a possible 45), annual SVM GPP prediction error was 32.1% for non-forest ecosystems and 22.2% for forest ecosystems, while the standard Moderate Resolution Imaging Spectroradiometer GPP product (MOD17) had an error of 50.3% for non-forest ecosystems and 21.5% for forest ecosystems, suggesting that the regionally tuned SVM performed better than the standard global MOD 17 GPP for non-forest ecosystems but had similar performance for forest ecosystems. The most important explanatory factor for GPP prediction was EVI, removal of which increased GPP RMSE by 0.85 gC/m2/day in a cross-validation experiment. Third, using the SVM driven by remote sensing data including incident shortwave radiation, we predicted 2004 conterminous U.S. GPP and found that results were consistent with expected spatial and temporal patterns. Finally, as an illustration of SVM GPP for ecological applications, we estimated maximum light use efficiency (e(max)), one of the most important factors for standard light use efficiency models, for the conterminous U.S. by integrating the 2004 SVM GPP with the MOD17 GPP algorithm. We found that emax varied from similar to 0.86 gC/MJ in grasslands to similar to 1.56 gC/MJ in deciduous forests, while MOD17 emax was 0.68 gC/MJ for grasslands and 1.16 gC/MJ for deciduous forests, suggesting that refinements of MOD17 emax may be beneficial. 2007 Elsevier Inc. All rights reserved.

Despite early speculation to the contrary, all tropical forests studied to date display seasonal variations in the presence of new leaves, flowers, and fruits. Past studies were focused on the timing of phenological events and their cues but not on the accompanying changes in leaf area that regulate vegetation-atmosphere exchanges of energy, momentum, and mass. Here we report, from analysis of 5 years of recent satellite data, seasonal swings in green leaf area of approximate to 25% in a majority of the Amazon rainforests. This seasonal cycle is timed to the seasonality of solar radiation in a manner that is suggestive of anticipatory and opportunistic patterns of net leaf flushing during the early to mid part of the light-rich dry season and net leaf abscission during the cloudy wet season. These seasonal swings in leaf area may be critical to initiation of the transition from dry to wet season, seasonal carbon balance between photosynthetic gains and respiratory losses, and litterfall nutrient cycling in moist tropical forests.

Interannual variations in terrestrial carbon cycle over tropical rainforests affect the global carbon cycle. Terrestrial ecosystem models show the interannual relationship between climate changes due to El Nino-Southern Oscillation (ENSO) and net primary production over tropical rainforests. However, we need an independent analysis using satellite-based vegetation index and climate parameters. In the present study, we extracted the ENSO-related interannual variations from time-series in Normalized Difference Vegetation Index (NDVI) and climate data from 1981 to 2000, and analysed their relevance. We detected relationships among NDVI, ENSO, and climate parameters from long-term data with negative NDVI-ENSO, NDVI-temperature, and positive NDVI-precipitation relations. These correlations suggest that interannual variability in vegetation activities over tropical rainforests could be extracted from NDVI time-series despite noise components in NDVI data, and that interannual changes in precipitation and temperature caused by ENSO play a more important role in vegetation activities over tropical rainforests than in incoming surface solar radiation.

Accurate parameterization of rooting depth is difficult but important for capturing the spatio-temporal dynamics of carbon, water and energy cycles in tropical forests. In this study, we adopted a new approach to constrain rooting depth in terrestrial ecosystem models over the Amazon using satellite data [moderate resolution imaging spectroradiometer (MODIS) enhanced vegetation index (EVI)] and Biome-BGC terrestrial ecosystem model. We simulated seasonal variations in gross primary production (GPP) using different rooting depths (1, 3, 5, and 10 m) at point and spatial scales to investigate how rooting depth affects modeled seasonal GPP variations and to determine which rooting depth simulates GPP consistent with satellite-based observations. First, we confirmed that rooting depth strongly controls modeled GPP seasonal variations and that only deep rooting systems can successfully track flux-based GPP seasonality at the Tapajos km67 flux site. Second, spatial analysis showed that the model can reproduce the seasonal variations in satellite-based EVI seasonality, however, with required rooting depths strongly dependent on precipitation and the dry season length. For example, a shallow rooting depth (1-3 m) is sufficient in regions with a short dry season (e.g. 0-2 months), and deeper roots are required in regions with a longer dry season (e.g. 3-5 and 5-10 m for the regions with 3-4 and 5-6 months dry season, respectively). Our analysis suggests that setting of proper rooting depths is important to simulating GPP seasonality in tropical forests, and the use of satellite data can help to constrain the spatial variability of rooting depth.

Application of remote sensing data to extrapolate evapotranspiration (ET) measured at eddy covariance flux towers is a potentially powerful method to estimate continental-scale ET. In support of this concept, we used meteorological and flux data from the AmeriFlux network and an inductive machine learning technique called support vector machine (SVM) to develop a predictive ET model. The model was then applied to the conterminous U.S. In this process, we first trained the SVM to predict 2000-2002 ET measurements from 25 AmeriFlux sites using three remotely sensed variables [land surface temperature, enhanced vegetation index (EVI), and land cover] and one ground-measured variable (surface shortwave radiation). Second, we evaluated the model performance by predicting ET for 19 flux sites in 2003. In this independent evaluation, the SVM predicted ET with a root-mean-square error (rmse) of 0.62 mm/day (approximately 23% of the mean observed values) and an R-2 of 0.75. The rmse from SVM was significantly smaller than that from neural network and multiple-regression approaches in a cross-validation experiment. Among the explanatory variables, EVI was the most important factor. Indeed, removing this variable induced an rmse increase from 0.54 to 0.77 mm/day. Third, with forcings; from remote sensing data alone, we used the SVM model to predict the spatial and temporal distributions of ET for the conterminous U.S. for 2004. The SVM model captured the spatial and temporal variations of ET at a continental scale.

In this study we present a new biosphere model called the Biosphere model integrating Eco-physiological And Mechanistic approaches using Satellite data (BEAMS). BEAMS provides a new method of calculating the environmental stress affecting plant growth (Stress). Stress is calculated eco-physiologically using a photosynthesis model and stomatal conductance formulation, providing a more realistic result than previous models. Stress values are used to estimate Gross Primary Production (GPP) estimates via the light use efficiency concept. We used BEAMS, including our new Stress approach, to investigate global spatial and temporal patterns of net primary production (NPP) and net ecosystem production (NEP). BEAMS was run for the years 1982-2000 using global scale satellite and climate data. Comparison of model results with observational measurements at flux sites reveals that GPP values predicted by BEAMS agree with measured GPP. Obtained Stress values were compared with those of MOD17 and CASA; the three methods produce contrasting spatial patterns. Upon comparing predicted and observed NPP, the pattern of NPP for each plant functional type can be adequately estimated. In terms of trend analysis, NPP increased for the years 1982-2000 in most regions. Different NPP trends were observed in Europe, Russia, and northeast Canada than those proposed by Nemani et al. (2003); we attribute these differences to climate-related processes. Simulated interannual variations in global NEP are similar to results from inverse modeling. A sensitivity study of obtained NEP shows that the interannual variability in NEP is strongly controlled by air temperature, precipitation, CO(2), and the fraction of absorbed photosynthetically active radiation.

The role of tropical ecosystems in global carbon cycling is uncertain, at least partially due to an incomplete understanding of climatic forcings of carbon fluxes. To reduce this uncertainty, we simulated and analyzed 1982-1999 Amazonian, African, and Asian carbon fluxes using the Biome-BGC prognostic carbon cycle model driven by National Centers for Environmental Prediction reanalysis daily climate data. We first characterized the individual contribution of temperature, precipitation, radiation, and vapor pressure deficit to interannual variations in carbon fluxes and then calculated trends in gross primary productivity (GPP) and net primary productivity (NPP). In tropical ecosystems, variations in solar radiation and, to a lesser extent, temperature and precipitation, explained most interannual variation in GPP. On the other hand, temperature followed by solar radiation primarily determined variation in NPP. Tropical GPP gradually increased in response to increasing atmospheric CO2. Confirming earlier studies, changes in solar radiation played a dominant role in CO2 uptake over the Amazon relative to other tropical regions. Model results showed negligible impacts from variations and trends in precipitation or vapor pressure deficits on CO2 uptake. (c) 2005 Elsevier B.V. All rights reserved.

Deforestation in Rondonia state in the south-western part of the Brazilian Legal Amazon was analysed using Landsat Multi-Spectral Scanner (MSS), Thematic Mapper (TM), Enhanced Thematic Mapper Plus (ETMz), National Oceanic & Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) and hydrological data. The Landsat sensor data coverage was supplemented with Pathfinder AVHRR Land ( PAL) Normalized Difference Vegetation Index (NDVI) datasets. The results from the Landsat-based analysis show that more than 30% of the natural vegetation in the study area was subject to deforestation between 1973 and 1999, a finding reinforced by analysis of the PAL NDVI data. In addition, it was established that trends in the PAL NDVI datasets were coincident with the pattern of deforestation. Apart from imagery analysis, time variations in the hydrological data between 1982 and 1988 were used to estimate the evapotranspiration. A decreasing trend was identified in the rate of evapotranspiration, suggesting that deforestation has a significant impact on the local hydrological cycle.

A simple Earth system model, the Four-Spheres Cycle of Energy and Mass (4-SCEM) model, has been developed to simulate global warming due to anthropogenic CO2 emission. The model consists of the Atmosphere-Earth Heat Cycle (AEHC) model, the Four Spheres Carbon Cycle (4-SCC) model, and their feedback processes. The AEHC model is a one-dimensional radiative convective model, which includes the greenhouse effect of CO2 and H2O, and one cloud layer. The 4-SCC model is a box-type carbon cycle model, which includes biospheric CO2 fertilization, vegetation area variation, the vegetation light saturation effect and the HILDA oceanic carbon cycle model. The feedback processes between carbon cycle and climate considered in the model are temperature dependencies of water vapor content, soil decomposition and ocean surface chemistry. The future status of the global carbon cycle and climate was simulated up to the year 2100 based on the "business as usual" (IS92a) emission scenario, followed by a linear decline in emissions to zero in the year 2200. The atmospheric CO2 concentration reaches 645 ppmv in 2100 and a peak of 760 ppmv approximately in the year 2170, and becomes a steady state with 600 ppmv. The projected CO2 concentration was lower than those of the past carbon cycle studies, because we included the light saturation effect of vegetation. The sensitivity analysis showed that uncertainties derived from the light saturation effect of vegetation and land use CO2 emissions were the primary cause of uncertainties in projecting future CO2 concentrations. The climate feedback effects showed rather small sensitivities compared with the impacts of those two effects. Satellite-based net primary production trends analyses can somewhat decrease the uncertainty in quantifying CO2 emissions due to land use changes. On the other hand, as the estimated parameter in vegetation light saturation was poorly constrained, we have to quantify and constrain the effect more accurately.

The relationship between the Normalized Difference Vegetation Index (NDVI) and climatic variables was analysed on a global scale using the Pathfinder AVHRR Land NDVI data set, and observed climate data for the period 1982-1990. A significant correlation between interannual NDVI and temperature variation was recognized in the northern mid- to high latitude areas between spring and autumn. A significant correlation was also identified between the NDVI, temperature and precipitation in northern and southern semiarid regions. A comparison of global NDVI trends show that NDVI increases in the northern mid- and high latitudinal zones are related to temperature rise, and NDVI decreases in southern semiarid regions are due to a precipitation decrease in the survey period. Although the cause of NDVI increases in the equatorial regions remains unclear, the combined effects of forest regrowth, deforestation and fertilization may impact on the NDVI trend.