Spatial measurement, inversion and mapping of carbon flows

_CO2 emissions are highly spatially heterogeneous, not least because fossil fuels are mainly used in the industrialized areas of the northern hemisphere. An international effort has led to the development of a worldwide network of several dozen stations measuring _pCO2 and other gases and associated parameters (notably ^δ13C). The longest continuous record (since 1958) is at the Mauna Loa station in Hawaii, where _pCO2 is measured by an infrared spectroscopy system.

Each _pCO2 record is characterized by a more or less marked seasonality, depending on the natural seasonal cycle of plant growth. This seasonality is particularly wide (> 20 ppm for _pCO2 and 1 ‰ for ^δ13C) for the high latitudes of the northern hemisphere characterized by large continental forests.

For the same latitude, _pCO2 shows a much more marked high-frequency variability (of 20-30 ppm and 1 ‰ for ^δ13C) for sites on the continent compared with those over the ocean. These fluctuations illustrate the rapid dynamics associated with the terrestrial biosphere and air mass mixing over the continents.

In addition to the seasonal cycle and high frequencies, all stations show a common long-term trend, with an increase rate of around 1 to 2 ppm per year. Time tracking reveals systematic variations, with maxima and minima corresponding to certain years (e.g. + 3.4 ppm in 1998 and + 0.6 ppm in 1992). This interannual modulation affects all latitudes, but variations are more marked in the northern hemisphere.

Spatial and temporal analysis of atmospheric _pCO2 has made recent progress with the launch of satellites (GOSAT, OCO-2) whose spectrometers quantify _pCO2 by measuring the absorption of solar radiation reflected by the Earth in certain infrared wavelengths. The advantage of remote sensing is that it can cover the Earth's surface in a matter of weeks, and map regional _pCO2 anomalies linked to natural and anthropogenic sources, as well as to pumping by oceanic and biospheric sinks.

Numerical models, including spatialization of the atmosphere and carbon cycle reservoirs, enable unknown fluxes to be calculated from _pCO2 measurements and known fluxes, notably anthropogenic_CO2 emissions. An emblematic example is the interpretation of the inter-hemispheric _pCO2 gradient, which today stands at around 4 ppm. This difference is linked to the fact that air exchanges between the hemispheres are not fast enough to homogenize the anthropogenic_CO2 emitted mainly in the northern hemisphere. Simple numerical modelling allows us to deduce that the terrestrial biosphere of the Northern Hemisphere constitutes a carbon sink of around 1 GtC/year.

The same principle is applied to the spatial and temporal range of atmospheric _pCO2, using 3D models to simulate atmospheric transport. At each time step, the model is perturbed by anthropogenic_CO2 emissions whose transport, after each time increment, should enable _pCO2 observations to be reproduced. The differences are linked to the existence of other sources, or sinks, with their specific intensities and spatial distributions. These other sources are calculated and cumulated at each time step of the calculation.

Such flux inversions have been carried out by a dozen modeling groups, whose maps can be compared and compiled (TRANSCOM international project). The continental biospheric flux shows marked interannual variations, mainly in the intertropical fringe, with a negative correlation with the ENSO oscillation (decrease in the biospheric sink during El Nino events, notably that of 1998). These inversion calculations are completed and corrected by taking into account ^δ13Crecords and upper-air _pCO2 gradients measured by aircraft sensors.

For the last two decades, the inversion of carbon fluxes indicates that the biospheric sink is mainly located in the temperate latitudes of the Northern Hemisphere (North America, boreal and temperate Asia, Europe and, to a lesser extent, in the temperate zones of South America and South Africa). This biospheric sink is partially offset by source zones located in the intertropical fringe (South and Central America, tropical Africa and Asia).

For the decade 2000-2010, the total gross biospheric sink is around 2.5 GtC/year, of which 1.5 corresponds to the high latitudes of the northern hemisphere and 1 GtC/year to the intertropical zone. Over the same period, the pumping of the terrestrial biosphere is therefore equivalent to the oceanic sink of around 2.5 GtC/year (see previous years' lecture). Furthermore, changes in atmospheric _pCO2 are equivalent to an average increase in atmospheric stock of around 4 GtC/year.

The total increase in carbon is therefore around 9 GtC/year (2.5 + 2.5 + 4), which corresponds to anthropogenic emissions over the same decade (8 GtC/year of fossil carbon and 1 GtC/year from deforestation and soil degradation). In the end, the net flow from the atmosphere to the terrestrial biosphere is therefore of the order of 1.5 GtC/year (2.5 from the gross biospheric sink minus the 1 GtC/year from deforestation and soil degradation).

The evolution of the terrestrial carbon balance should be clarified in the future by inversions of _pCO2 fields measured at high spatial resolution by satellites. Their initial data even make it possible to locate " " point sources, such as large urban areas, and to use inversion to calculate the carbon emissions of a given region or country, and to compare them with declared emissions.