Current predictions of future CO₂ sink strength vary widely as a result of different model representations of the carbon cycle. A sound characterization of these prediction uncertainties is crucial for the design of economically efficient carbon management strategies. We use a mechanistically sound and statistically tractable model of the global carbon cycle to (1) assimilate historical observations of atmospheric CO₂ concentrations and oceanic CO₂ fluxes, (ii) derive probabilistic predictions of future CO₂ concentrations and fluxes, and (iii) compare the utility of terrestrial and oceanic observations to constrain predictive uncertainties. We found that terrestrial and oceanic flux observations have nearly equal ability to constrain these uncertainties, if terrestrial observations include both net primary productivity (NPP) and respiration. Model predictions are dependent on the choice of historical land use emissions dataset. The probability density function (PDFs) of model parameter estimates are not normally distributed, and neglecting autocorrelation in the CO₂ concentration signal during model calibration causes overconfident results.

Regular vertical profiles over Europe were set up in 2001 as part of the AEROCARB and Carboeurope-IP projects at five locations: Griffin (56°36'N, 3°47'W, Scotland), Orléans (47°50'N, 2°30'E, France), Schauinsland (47°55'N, 7°55'E, Germany), Hegyhatsal (46°57'N, 16°39'E, Hungary), and Bialystok (53.20°N, 22.75°E, Poland). The objective of the program is to measure CO₂, CH₄, N₂O, SF₆, CO, ¹³C and ¹⁸O in CO₂ vertical profiles at a bi-weekly frequency using air samples taken up at several levels from 100m up to 3000 m above the ground surface. One liter flasks are sampled on board small aircraft using a standardised protocol. The samples are analysed at three laboratories (LSCE, MPI-BGC, IUP-UHEI) which are linked through regular intercomparison exercises. We have characterised for each site the CO₂ seasonal cycles within the atmospheric boundary layer (ABL: 14 to 20 ppm) and the free troposphere (FT: 10 to 13 ppm). From these signals we have calculated the difference between ABL and FT, known as the CO₂ 'jump', which will be compared to the simulations from atmospheric transport models. We have also calculated the offset between each airborne sampling site and the time series from Mace Head observatory, used as a maritime reference. For CO₂, the wintertime offsets at the lowest level of the average vertical profiles are ranging from 0 ppm in Scotland up to 10 ppm in all continental sites. Depending of the site the positive offset due to emissions from anthropogenic and biospheric processes may extend up to 300 to 1500 m agl. In summertime we observe a negative gradient in most of the sites with a typical decrease of 5 ppm between 2000m and 100m agl. The average vertical gradients will be compared to the ouput of atmospheric models, and will be analysed with regards to the other trace gas (CO, CH₄, and CO₂ isotopes).

We carried out airborne campaigns over Europe in order to analyze atmospheric CO₂ variability at the regional scale. Data reveal a higher standard variation in the planetary boundary layer (PBL) against a lower one in the free troposphere (FT), where the air is more well mixed. Ground data generally agree well with airborne measurements when done in the FT, but not in the PBL where they are exposed to local disturbances. Ground stations located in the FT are shown to be representative of a regional scale while PBL observatories provide only locally representative measurements.

CHIOTTO – Continuous HIgh-precisiOn Tall Tower Observations of greenhouse gases – is a European Union-funded project which has as objective to build an infrastructure for the continuous monitoring of greenhouse gas concentrations across Europe above the surface layer using tall towers (~300m height). For this purpose a new analysis system for continuous atmospheric measurements was built and tested at Max Planck Institute for Biogeochemistry, Jena, Germany and was recently installed at a 300 m tower close to Bialystok, Poland (Lat 53°14'N, Long 23°01'E, Alt 180m), as part of the “CHIOTTO” tall tower network. Since July 2005 this system is measuring quasi-continuously the atmospheric concentration of CO₂, CH₄, CO, N₂O, SF₆ and the O₂/N₂ ratio as well as meteorological parameters (atmospheric pressure, temperature, humidity; wind speed and direction) from 5 heights on the tower ranging from 5 to 300 m. The measurement devices are: an Oxzilla O₂ fuel cell analyzer, a LiCor 7000 NDIR CO₂ analyzer, an Agilent gas chromatograph (GC) with flame ionization detector (FID) and electronic capture detector (ECD) for CH₄, CO, N₂O, SF₆. The challenge was to build a reliable automatic system which can run continuously with very little maintenance and to fulfill at the same time the high precision requirements for all the measured species prescribed by the CHIOTTO project goals. The high temporal resolution achieved will capture short term events and diurnal variability. In addition, the system is planned to run for at least several years in order to observe long-term trends as well. We describe the technical setup of the measurement system, the region of influence of the station and present the first months of data if available: correlations between species, observed short term variability patterns and their relation to meteorology and air parcel paths.

We present preliminary results of a modeling experiment that compares observed vertical profiles of CO₂ with those generated by an atmospheric transport model (ATM). The ATM is driven by CO₂ flux fields generated from the inversion of monthly averaged CO₂ surface data (GLOBALVIEW). We note large differences between the best fit to the observations produced in the inversion and the same quantity simulated by the forward model. This difference arises from the nonlinearity of the advection scheme used in the transport model. When comparing with vertical profiles, we note that much of the difference between simulated and observed concentration has the same structure as the impact of this nonlinearity. Inversion schemes must therefore take nonlinearity into account. Despite these differences, the profiles are able to distinguish among inversions that fit subsets of the surface data, suggesting they are a useful validation dataset.

We present an analysis of terrestrial net CO₂ fluxes from North America for the period 2000-2004. These fluxes consist of hourly maps at ~70km×100km resolution that are consistent with observed atmospheric CO₂ mixing ratios, as well as with varying climatic conditions across different ecosystems as observed from space. The flux maps are created in a newly developed ensemble data assimilation system that consists of the atmospheric Transport Model v5 (TM5), the Vegetation Photosynthesis Respiration Model (VPRM), and an efficient Bayesian least-squares algorithm to optimize the fluxes from different biomes in VPRM against CO₂ mixing ratios from the NOAA-CMDL observing network. The stochastic nature of the ensemble data assimilation system allows us to consistently include uncertainty on net CO₂ fluxes from the neighboring oceans and more distant continents in the flux estimates for North America.

The inventory of nuclear bomb produced ¹⁴C (bomb ¹⁴C) in the ocean is a major constraint of CO₂ exchange between the atmosphere and ocean in numerical models and analytical estimates of gas exchange. New ¹⁴C data in the ocean, improved methods of separating the bomb ¹⁴C from the natural background of ¹⁴C in the ocean, and reassessment of previous inventories are challenging the canonical estimates of the air-sea gas transfer. An improved method of separating natural ¹⁴C from the observed ¹⁴C distribution is being used to estimate the bomb ¹⁴C distribution and inventory. We use GEOSECS ¹⁴C data to represent the global distribution in 1975, and the new WOCE dataset for 1995 to get two time representations of inventory. To reduce the bias error for averaging zonal bomb ¹⁴C inventories from limited observation stations during the GEOSECS times, we use zonal averages given by Peacock [2004] for re-evaluation of 1975 air-sea CO₂ exchange rates. Zonal inventories for 1995 will be from GLODAP mapping results using WOCE data [Key et al. 2004]. Lateral transport models developed by Broecker et al. [1985] are used to assess the regional air-sea CO₂ exchange rates as well as an appropriately weighted global mean. Four independent methods of estimating bomb ¹⁴C inventory in the ocean show that the original estimate by Broecker et al. [1995] could be about 25% too high, the air-sea CO₂ exchange rates derived from this original bomb ¹⁴C inventory could also be too high by a similar amount. Results of this assessment will be presented.

We present a new version of the global atmospheric tracer transport model driven by analyzed meteorology with diurnally varying mixing in the boundary layer capable of running globally at resolutions up to quarter degree longitude-latitude or higher. The impact of the higher resolution model can be visible in resolving city plumes, airmass boundaries, diurnal cycle, fronts and synoptic scale events often observed in continuous CO₂ monitoring site data.

This paper presents estimates of the ¹³C Suess effect and anthropogenic carbon concentration increase in the Nordic Seas since 1981.

The formation of Antarctic Intermediate Water is investigated in a state of the art numerical model. Results are compared with a previous, lower resolution version of the model, and with data from the World Ocean Circulation Experiment.