We have carried out the tank experiment in the low-temperature room to clarify the CO₂ gas exchange mechanism between the sea and the overlying air during the sea-ice formation process. The air CO₂ concentration in the headspace of the tank began to increase simultaneously with the sea-ice formation and growth. The CO₂ flux was with in the range from 2.1x10^-4 to 4.5x10^-4 g-C m^-2 hour^-1 at ice thickness of 5cm. The CO₂ flux was mainly dependent on the brine salinity in the upper layer of sea-ice, which suggests that CO₂ was released from the brine in the sea-ice, and transported to the atmosphere.

Exchange rates within the Global Carbon Cycle, between oceans, atmosphere and terrestrial biosphere – including the anthropogenic CO₂ production – are being traced by concentration and isotope ratio measurements of atmospheric CO₂. The background value of the stable isotope ratio of oxygen in atmospheric CO₂ is determined by oxygen exchange with the ocean surface waters. During contact with leaf water, the signature of this then evaporation-enriched groundwater (the extent still being dependent on plant physiological and environmental parameters), will be imprinted on CO₂ diffusing back out of the leaf stomata. From water cycle studies the continental effect (Rayleigh-distillation) is known, leading to precipitation strongly depleted in d¹⁸O over e.g. Siberia. This signal is also transferred into plant material. These main mechanisms within the ¹⁸O-cycle are known or under investigation. The d¹⁸O source term for atmospheric CO₂ derived from biomass burning and anthropogenic fossil fuel combustion, however, is less well-known.

In order to elucidate seasonal and interannual variations of oceanic CO₂ uptake in the Greenland Sea and the Barents Sea, partial pressures of CO₂ in the surface ocean (pCO₂^sea) were measured from 1992 to 2001. The values of pCO₂^sea were lower than the partial CO₂ pressures in the atmosphere (pCO₂^air) throughout the year, and the annual net air-sea CO₂ fluxes in the Greenland Sea and the Barents Sea were evaluated to be 52 ± 31 and 46 ± 27 gC m^-2 yr^-1, respectively, yielding a total oceanic CO₂ uptake of 0.050 ± 0.030 GtC yr^-1. We also found that the annual mean CO₂ uptake was positively correlated with the North Atlantic Oscillation Index (NAOI) via wind strength, but was negatively correlated with DpCO₂ (pCO₂^air-pCO₂^sea) and the sea ice coverage. The results also indicate that the wind speed and sea ice coverage play a major role in determining the interannual variation of CO₂ uptake, with DpCO₂ playing a minor role.

We estimated the production of bomb radiocarbon using available information on atmospheric nuclear bomb tests, the simple (radio-)carbon cycle model GRACE (Global RadioCarbon Exploration Model) and atmospheric observations as constraints. Subsequent forward simulations of the bomb radiocarbon inventory in the different carbon reservoirs turned out to be in very good agreement with recent observation-based estimates, therewith for the very first time allowing to close the global bomb radiocarbon budget. Besides confirming original stratospheric bomb ¹⁴C data published in the reports of the Health and Safety Laboratories [Telegadas, 1971, and references therein], our results confirm recent observation-based ocean bomb radiocarbon inventory estimates for the time of GEOSECS (1970s) and WOCE (1990s) from Peacock [2004] and Key et al. [2004], but refute the GEOSECS ocean inventory estimates from Broecker et al. [1985, 1995].

We used recent ocean bomb radiocarbon inventory estimates for the time of GEOSECS (mid-1970s) and WOCE (mid-1990s) from Peacock [2004] and Key et al. [2004], corrected for missing ocean areas [Naegler 2005], to develop a new parameterisation of the piston velocity – wind speed relationship of CO₂ air-sea gas exchange. For monthly mean climatological winds on a 1°x1° grid, this results in a gas exchange parameter a_q,660 of 0.32±0.04 (in cm hr^-1 m^-2 s²) and a net oceanic CO₂ uptake of 1.53±0.18 PgC/yr for the mid-1990s, when using the Takahashi et al. [2002] pCO₂ data.

Using high-quality data for the CO₂-system and related properties obtained 10-year apart, we estimated decadal increases of anthropogenic CO₂ along the A10 section of the World Ocean Circulation Experiment (WOCE) Hydrographic Program (WHP). Increases of anthropogenic CO₂ were found down to an isopycnal surface of 27.3σ_θ (approx. 1000 dbar). In the sub-Antarctic Mode Water (SAMW), the increase was 6.9 ± 2.0 μmol kg^-1 on average, while in the Antarctic Intermediate Water (AAIW), it was 4.2 ± 1.9 μmol kg^-1. The increase in SAMW was larger in the west than that in the east of the section. No significant increases were detected in North Atlantic Deep Water (NADW) and Antarctic Bottom Water (AABW).

The coupled feedback processes of energy and carbon cycles are an essential mechanism for understanding global environmental change. We developed a simplified one-dimensional carbon and energy cycle coupled model to quantify the feedback processes between energy and carbon cycles. The model was calibrated to reproduce the historical variations in temperature and atmospheric CO₂ concentration. The model results of vertical ocean temperature profiles, and latitudinal NPP and NEP distributions were in good agreement with the observation data and terrestrial biosphere model results. The regional difference of terrestrial ecosystem response by climate feedback appeared in the middle and high latitudes. The north-south distribution is important to investigate the terrestrial ecosystem because the opposite response appeared in the middle and high latitude. The future change of carbon cycle and climate was also simulated up to the year 2100 based on the IPCC scenario. The atmospheric CO₂ concentration reaches 735 ppmv in 2100 and global average temperature increases 1.9 K for 2000-2100.

Systematic observations of the atmospheric CO₂ concentration, and carbon and oxygen isotope ratios of CO₂ (d¹³C and d¹⁸O) have been maintained at Japanese Arctic Observatory in Ny Ålesund (79°N, 12°E) and Antarctic station, Syowa (69°S, 40°E). The interannual variations of the CO₂ concentration and d¹³C in association with the occurrence of ENSO event were clearly observed at the both sites. The d¹⁸O values observed at Syowa Station showed significant increasing trend after 1999.

Begur-Pals site (41,58ºN, 3,14ºE, Catalonia, Spain) is weekly sampled for CO₂ and other GHG (CH₄, CO, N₂O, SF₆) since January 2000. This CO₂ serial data shows at the middle of each summer a sudden increase and decrease of the CO₂ peak. It is a process that can be either attributed to a highest transpiration rate than ecosystem production due to the lack of summer precipitation, to biomass burning from Mediterranean forest fires, to tourist activities in the coast, or to CO₂ pumping from waters in the Western Mediterranean sea (according to wind backtrajectories). A sampling strategy using sites with high towers with continuous measurements has been developed. Sites are placed at the vortexes of a rhombus: two extremes are continental sites in the center of the Ebro’s watershed and a marine site is located in the Menorca Island. The other two are high towers in the Catalonian coast.