Present: Ric Williams, Rhiannon Mather (Liv), Pete Brown, Marie-Jo Messias, Ute Schuster, Andy Watson (UEA), Elaine McDonagh , Brian King, Richard Sanders, Harry Bryden (NOCS).
Apologies: Hannah Corney, Tim Jickells, Tim Lesworth and Kay Thorne.
Presentations on Friday afternoon and poster session on Wednesday of the Orlando Meeting. For those attending, propose an informal meeting perhaps after the poster session, but will be arranged when the programme is known.
Overall: the above set of planned journals are strong and providing good outcomes to the study. Obviously, the really important aspect of the study involves the final synthesis, which is on track.
Discussion about the progress of the data returns. Elaine wishes to compile a complete bottle data set, which is internally consistent. Discussion of the units for the tracer concentrations, whether per kg or per litre, which is delaying this task being completed. Per litre is standard BODC practice, but this choice is not apparently applied for other data centres.
Report on DON signals measured via UV oxidation and HTCO for 36N; same data sets exist for 26N. There is a greater recovery for the HTCO, but possibly more noise. In some regions, there are contrasting vertical profiles.
Action: Rhiannon to liaise with Eric to check the HTCO data, which in principle should be more accurate. Richard/Sinhue/Rhiannon to liaise in providing checked data sets to Elaine.
Recommendation to include both DON data sets in the final box inversion (neither are used to constrain the inversion) in order to provide two estimates for the supply of DON to the box.
Discussion of SF6 and transient tracer ages for 36N. SF6 has a delayed uptake. Discussion of surface saturation for 36N data, compared with previous German data that had lower surface saturations over the upper 500m. Either there are methodological issues or different winter mixed layer history for each year.
Thorny problem of C changes between repeated sections for 26N addressed. Initially see contrasting uptakes for 2 different 6 year periods. By careful analysis of crossing points, oxygen for 1998 section was corrected, which then altered the inferred C changes. Final fields looked much more improved. This work to update a JGR manuscript in the revision stage. If possible, it would be very valuable for this analysis to be applied to 36N as well.
Briefly reported on work addressing transport variability using ARGO float data and assumptions of western or eastern boundary condition.
Update on the progress with the inverse model, reporting on the inversion for the combined sections, which is the central part of the 36N programme.
At the previous 2 meetings, there has been discussion of the western boundary transport and eastern boundary transport. Both these problems have now been solved in the latest analysis of the initial state.
Discussion of the Ekman correction to account for the 26N and 36N sections being in the late spring, temperature offset applied, needs to be applied in a similar manner for the other tracers.
When the box inverse model is now applied, there is a greater surface heat loss over the subtropics, which reduces the heat transport at 36N from 1.2 PW to 1.1 PW.
The inversion struggles with the constraints linked to the diapycnal transfer for layers associated with the Mediterranean outflow. Discussion ranged on whether to incorporate an additional gyre circulation, allow higher diapycnal transfers for these layers or allow a greater storage and expansion for these layers. Recommendation was to compare with likely estimates of double diffusion for that region.
Discussion of when the box inverse is to be melded with tracer fields. For the tracer fields, need the carbon and nutrient teams to
While need more confidence for the final inversion, plan to proceed with incorporating these tracer budgets in January/February (for both the Orlando meeting in early March and to help with theses being written up; Brown and Mather in March).
Overall: progress in the programme looks good: analyses of the single sections is near completion, the inverse model is broadly working, and further fine tuning and melding with tracer fields planned for the early spring.
Action: tracer groups to finalise their tracer distributions and estimate seasonal signals, so that can be incorporated into the box inverse.
In Orlando for those participants there in March 2008. Subsequent UK meeting subject to progress of papers and the inverse model.
© 2015 NERC.
Last updated: June 8, 2015
