B42B-01 INVITED
Top-down and bottom-up carbon budgets of North America, Europe and Asia
The European, North Asian and North American continents have markedly different ecosystem characteristics, climate regimes, as well as forest management, land use and disturbances histories. Therefore, their carbon balance and controlling processes are different. We will analyze on an area basis, and on a continental average basis the differences and uncertainties in Net Primary Productivity (NPP) and in net carbon balance (NBP) between the two continents. The analysis of NPP relies on long term simulations from the SIB-2, ORCHIDEE and LPJ global process-oriented ecosystem models. Net carbon balance estimates are derived from diverse ensembles of atmospheric inversions over the past 20 years and from independent forest biomass inventories. An additional rigorously consistent estimation of NPP and NBP fluxes from top-down and bottom-up models constrained by atmospheric concentration and remote sensing observations is derived from a Carbon Cycle Data Assimilation System. The long-term trend in growing season onset and termination and in carbon uptake duration will be analyzed over both continents using ecosystem model output and remotely sensed vegetation properties, and related to changes in NBP
B42B-02
Regional CO2 Inversion Study for Europe and West Siberia
Recently, the number of stations to continuously measure atmospheric CO2 concentration increased. These stations bring us important information about regional and short-term variations of CO2 fluxes through atmospheric transport. To take advantage of this new information, we developed a Bayesian synthesis inverse scheme to infer daily CO2 flux at a grid cell resolution over a pre-defined region from measured daily mean CO2 concentration. This inverse method was applied to AEROCARB European network measurements for the year 2001. To calculate atmospheric transport, we used the global circulation model LMDZ with variable grid size. We used enhanced spatial grid resolution over Europe up to 50x50 km2, therefore improving atmospheric transport modeling over this region. For each grid cell the flux contribution to simulated CO2 concentration at observation sites was calculated by LMDZ running in adjoint mode. Whereas still under discussion, we believe the flux errors at these small spatial and temporal scales don't vary independently from one another. Then we developed and tested three different spatial correlation schemes on flux errors in the inversion. In a first attempt, we used synthetic data to analyze the potential of the European network. These data were simulated by forward atmospheric transport of daily CO2 fluxes from the biogeochemical model ORCHIDEE for the year 2001. This allowed us to determine over which area the fluxes are retrieved most accurately in case of a perfect atmospheric transport. We also estimated at which minimal scales, in space and in time, fluxes should be aggregated to compare favorably with true fluxes from ORCHIDEE model. With actual observations, flux estimates appear less realistic. Temporal and spatial aggregations, as well as proper data selection, are shown to improve the fluxes considerably. Moreover, by 2005, NIES has equipped five towers in West Siberia to continuously measure atmospheric CO2 concentrations, in addition to previously established airborne observations. These towers are located relatively close from one another, covering 786000 km2, and allow us to study CO2 fluxes at regional scale by inverse modeling.. We developed regional inverse model setup with LMDZ zoomed over West Siberia for 2005, and tested the forward simulations of CO2 transport with ORCHIDEE model fluxes against observations.
B42B-03
Continental Carbon Cycle Data Assimilation in SiB-RAMS
We have analyzed a suite of atmospheric CO2 observations for 2004 using a coupled land-atmosphere modeling system (SiB-RAMS), and estimated surface sources and sinks using the Maximum Likelihood Ensemble Filter (MLEF). Fossil fuel emissions and air-sea gas exchange fluxes were prescribed from inventory data, and an annually-balanced terrestrial flux due to photosynthesis and respiration was calculated globally on a 1x1.25 degree grid using analyzed weather and satellite vegetation data. Regional corrections to these fluxes were then estimated each month from the global flask network by batch synthesis inversion using the global Parameterized Chemistry and Transport Model (PCTM). The output from this global inversion was then used as a �background� field to provide lateral and initial conditions on a mesoscale grid over North America. SiB-RAMS was used to generate high-resolution regional variations of CO2 corresponding to the background field, and also derive influence functions for each flask and continuous tower measurement site in the domain. These influence functions were then used to derive gridded biases in the background estimates of photosynthesis and respiration using the MLEF. Results show most of the domain to be poorly constrained by the observations, but good reduction of uncertainty over distances of several hundred km in the vicinity of each site. Spatial covariance of the model error is estimated explicitly by the MLEF, and improves the spatial coverage of the observational constraint. Model results are evaluated against aircraft observations that were not used in the assimilation.
http://biocycle.atmos.colostate.edu
B42B-04
The state of the carbon cycle in the Northeastern United States & Canada 1982-present: results from a constrained, dynamic terrestrial biosphere model
This study analyzed the state of the carbon cycle from 1982-present in the Northeastern United States and Canada. In contrast to previous analyses using simplified biosphere models, the ED2 terrestrial biosphere model used in this study is a complete dynamic vegetation model capable of predicting both short-term carbon fluxes and long-term changes in above- and below-ground carbon stocks. Prior to the analysis, the model's parameterization had been optimized against multiple data constraints, including eddy-flux measurements of fast time-scale ecosystem carbon fluxes and forest inventory measurements of long-term above-ground carbon dynamics. Subsequent regional- and site-level validation against independent datasets showed that a model that accurately predicted observed seasonal, annual and decadal-scale patterns of carbon fluxes within the Northeastern region. The mean annual carbon flux over the region during the simulation period was uptake of 0.57 tC ha-1 y-1, with substantial interannual variability: annual carbon fluxes ranged from a spatially-averaged mean uptake of 1.3 tC ha-1 y-1 in 1991 to a near-neutral biosphere in 1996. The three dominant causes of interannual variability in terrestrial carbon uptake during the period, in order of importance, were: (i) summer-time precipitation anomalies, (ii) spring-time temperature anomalies, and (iii) fall temperature anomalies. Above average temperature and precipitation conditions during all 3 of these periods caused increases in net primary productivity (NPP), but drove even larger increases in soil respiration, resulting in overall negative impacts on the rate of carbon uptake by terrestrial biosphere. The analysis also indicated a significant long-term decadal-scale trend in NPP of 0.05 tC ha-1 y-1 resulting from an increasing difference between regional biomass growth and regional carbon losses due to mortality and forest harvesting. However, due to high degree of interannual variability in soil respiration rates, this increasing long-term trend in NPP did not translate into a significant long-term trend in net carbon uptake.
B42B-05
Linking Forest Carbon Monitoring with Management Decisions
Managing forests to increase carbon stocks or reduce emissions requires knowledge of how management practices effect carbon pools over time, and inexpensive techniques to monitor activities. Here we discuss our approach to integrate the multi-tier monitoring data from the North American Carbon Program (NACP) with management decisions by linking bottom-up and top-down ecosystem models with decision-support tools. Monitoring carbon stocks and fluxes in the NACP involves a multi-tier hierarchy of observation methods: remote sensing, inventories, landscape biometrics, and flux towers. We use the GIS version of PnET-CN to scale up and map observations from flux towers, landscape biometrics, and inventories to areas of approximately 50 km2 around flux tower sites. The NASA-CASA model is used to scale down remote sensing observations from the MODIS sensor and biophysical maps to the same areas. Mapped estimates of productivity and biomass that embed consequences of land disturbances and forest age structure are used to compare and reconcile the top- down and bottom-up approaches, and to provide input to decision-support tools. Key information for the decision-support tools includes (1) estimates of carbon stocks and quantified impacts of management activity; (2) estimates of net ecosystem production (NEP) and changes in carbon pools; and (3) estimates of forest/atmosphere carbon fluxes and relevant effects from various environmental controls. To demonstrate the relevance of this work to land managers, we illustrate how this information can be used for estimating and reporting carbon stocks and changes in carbon stocks to the national greenhouse gas registry.
B42B-06
Are US Croplands a Source or Sink of Atmospheric Carbon Dioxide Based on Changes in Soil Organic Carbon Stocks?
US Croplands could potentially sequester from 275 to 750 Tg CO$_{2}$ equivalants per year according to analyses based on widespread adoption of alternative management practices that increases carbon input or reduces carbon losses from soils, relative to traditional practices. However, potentials are not necessarily representative of current trends. Our objectve was to determine the national trends by combining simulation modeling with datasets that provide input driving variables, including land use and management statistics, weather data, and soil characteristics. Century was used to simulate trends during the 1990s for US Croplands in a Monte Carlo Analysis, assessing uncertainty in model input data as well as the structural uncertainty associated with parameterization and model algorithms. Probability distribution functions were developed for N fertilization, tillage practices, organic amendments, while structural uncertainty was assessed based on an empirical estimator derived from 47 long-term agricultural experiments. A total of 100 iterations were simulated for each NRI during the analysis. Using this simulation modeling approach, US Croplands were estimated to sequester from 60 to 70 Tg CO$_{2}$ equivalents per yr in soil organic carbon pools, which is well below the estimated potentials. While agricultural practices have created a modest sink for carbon, sequestration rates are below the potentials due to limited adoption of alternative practices.
B42B-07
Spatially Explicit Full Carbon and Greenhouse Gas Accounting for the Midwestern and Continental US: Modeling and Decision Support for Carbon Management
Full carbon accounting for terrestrial ecosystems is intended to quantify changes in net carbon emissions caused by changes in land management. On agricultural lands, changes in land management can cause changes in CO2 emissions from fossil fuel use, agricultural lime, and decomposition of soil carbon. Changes in off-site emissions can occur from the manufacturing of fertilizers, pesticides, and agricultural lime. We are developing a full carbon accounting framework that can be used for estimates of on-site net carbon flux or for full greenhouse gas accounting at a high spatial resolution. Estimates are based on the assimilation of national inventory data, soil carbon dynamics based on empirical analyses of field data, and Landsat-derived remote sensing products with 30x30m resolution. We applied this framework to a mid-western region of the US that consists of 679 counties approximately centered around Iowa. We estimate the 1990 baseline soil carbon for this region to be 4,099 Tg C to a 3m maximum depth. Soil carbon accumulation of 57.3 Tg C is estimated to have occurred in this region between 1991-2000. Without accounting for soil carbon loss associated with changes to more intense tillage practices, our estimate increases to 66.3 Tg C. This indicates that on-site permanence of soil carbon is approximately 86% with no additional economic incentives provided for soil carbon sequestration practices. Total net carbon flux from the agricultural activities in the Midwestern US in 2000 is estimated at about - 5 Tg C. This estimate includes carbon uptake, decomposition, harvested products, and on-site fossil fuel emissions. Therefore, soil carbon accumulation offset on-site emissions in 2000. Our carbon accounting framework offers a method to integrate new inventory and remote sensing data on an annual basis, account for alternating annual trends in land management without the need for model equilibration, and provide a transparent means to monitor changes soil carbon. Our method of integration is capable of estimating regional or national changes in soil carbon while still representing heterogeneity at the sub-county level. Future research will include predictive changes in soil carbon and net carbon flux based on socio-economic drivers, and a sensitivity analysis using high-resolution remote sensing products.
B42B-08
Evaluating the Contribution of Soil Carbon to Global Climate Change Mitigation in an Integrated Assessment
Assessing the contribution of terrestrial carbon sequestration to national and international climate change mitigation requires integration across scientific and disciplinary boundaries. In a study for the US Climate Change Technology Program, site based measurements and geographic data were used to develop a three- pool, first-order kinetic model of global agricultural soil carbon (C) stock changes over 14 continental scale regions. This model was then used together with land use scenarios from the MiniCAM integrated assessment model in a global analysis of climate change mitigation options. MiniCAM evaluated mitigation strategies within a set of policy environments aimed at achieving atmospheric CO$_{2}$ stabilization by 2100 under a suite of technology and development scenarios. Adoption of terrestrial sequestration practices is based on competition for land and economic markets for carbon. In the reference case with no climate policy, conversion of agricultural land from conventional cultivation to no tillage over the next century in the United States results in C sequestration of 7.6 to 59.8 Tg C yr$^{-1}$, which doubles to 19.0 to 143.4 Tg C yr$^{-1}$ under the most aggressive climate policy. Globally, with no carbon policy, agricultural C sequestration rates range from 75.2 to 18.2 Tg C yr$^{-1}$ over the century, with the highest rates occurring in the first fifty years. Under the most aggressive global climate change policy, sequestration in agricultural soils reaches up to 190 Tg C yr$^{-1}$ in the first 15 years. The contribution of agricultural soil C sequestration is a small fraction of the total global carbon offsets necessary to reach the stabilization targets (9 to 20 Gt C yr$^{-1}$) by the end of the century. This integrated assessment provides decision makers with science-based estimates of the potential magnitude of terrestrial C sequestration relative to other greenhouse gas mitigation strategies in all sectors of the global economy. It also provides insight into the behavior of terrestrial C mitigation options in the presence and absence of climate change mitigation policies.