What does radiocarbon tell about the ocean ventilation?

The rate at which the ocean can sequester excess heat and carbon from the atmosphere is determined by its ventilation. Ventilation is defined as the renewal of interior waters by seawater that has been in contact with the atmosphere [1]. In models this rate is obtained by computing the ideal age. On the other hand, radiocarbon anomalies (Δ14C) measured either in the water column or in deep sea cores are one of the only ways to get estimates of present or past ocean ventilation rates. However, the interpretation of field data is hindered by the complexity of processes controlling radiocarbon in the ocean. While ventilation age is immediately set to zero at the sea surface, radiocarbon is characterized by a slow air-sea exchange rate. The interplay between slow sea surface adjustment and transit pathways in the ocean interior results in significant differences between radiocarbon and ventilation ages [2].

By means of modeling studies with idealized age tracers we investigate the implications of such departures for the use of Δ14C as a ventilation proxy. Splitting the global ocean into several sub-domains allows computing partial ages [3] which provide information about ventilation pathways. By tagging surface water we also obtain the ventilation fraction associated with any ocean surface region.

In contrast to the ventilation age, the radiocarbon signal at depth displays significant contributions from the entire ocean surface. The roles of the different ocean basins in setting the local age are also totally different when reported by radiocarbon rather than by the ventilation age. This work illustrates that the direct interpretation of radiocarbon anomalies from deep-sea cores as ventilation ages might be problematic.

Radiocarbon is represented in the model as a purely physical tracer [4]. Differences between radiocarbon and ventilation ages are positive and non-uniform (Fig. 1). In contrast to the ventilation age, radiocarbon is characterized by a slow air-sea exchange rate. This rate exerts the main control in setting radiocarbon ages at depth. Indeed, the age of a conservative tracer which exchanges with the atmosphere at the same global rate is very similar to that of the ''real'' radiocarbon (Fig. 2). We take advantage of these similarities to investigate the impact of piston velocity on tracer ages.

In this purpose we add to the model 7 tracers tagging ocean surface regions. This allows computing the mixing ratio and ventilation fraction due to any of the surface region. Two experiments which differ by the air-sea exchange rate are performed:

  • FAST: large restoring to surface values;

  • SLOW: radiocarbon-like rate (piston velocity ~7.5 m/yr).

These experiments present contrasted results (Fig. 3). With FAST air-sea exchange, deep waters (>1400 m) are mostly composed and ventilated by two sources: the North Atlantic and the Southern Ocean. This corresponds to the classical view of deep ocean ventilation. When air-sea exchange is SLOW, all surface regions contribute significantly to the deep water composition and ventilation. The contrast between the deep Atlantic and Pacific is significantly reduced. Further analysis, performed with partial ages (not illustrated), shows that for a radiocarbon-like tracer, significant contributions to the local deep age result from a larger life span in surface waters. This is a direct consequence of the slow air-sea exchange rate which reduces the probability of the radiocarbon clock being reset to zero in surface layers. These results put strong limitations on the practice consisting in interpreting Δ14C signals in terms of a limited number of end-members.




All results presented here are obtained with a version of the Max Planck Institute Ocean Model (MPIOM-GR30L40) which has been optimized for deep ocean radiocarbon, salinity, and temperature. Experiments were performed over 10 kyr with climatological (OMIP [5]) forcing.


  1. England (1995) Journal of Physical Oceanography, doi: 10.1175/1520-0485(1995)025<2756:TAOWAV>2.0.CO;2.
  2. Campin & al. (1999) Earth and Planetary Science Letters, doi: 10.1016/S0012-821X(98)00255-6.
  3. Mouchet & al. (2016) Ocean Dynamics, doi: 10.1007/s10236-016-0922-6.
  4. Toggweiler & al. (1989) Journal of Geophysical Research, doi: 10.1029/JC094iC06p08243.
  5. Marsland & al. (2003) Ocean Modelling, doi: 10.1016/S1463-5003(02)00015-X.