1. Resolving the seasonal cycle of mixed layer physics and phytoplankton biomass in the SAZ using high- resolution glider data Seb Swart, Sandy Thomalla & Pedro Monteiro

2. Chl-a seasonal cycle Thomalla et al., 2011 Complex balance between light and nutrient limitation that drives higher production in SAZ >> Sub-seasonal variability of MLD modulates this balance Joubert et al., submitted Recent work highlights importance of seasonal to sub- seasonal forcing of ML on PP (Levy, Klein, 2009; Thomalla et al., 2011; Fauchereau et al., 2011) The overall Chl variance that is explained by the seasonal cycle (0-100%) was computed as the variance explained by the regression of Chl onto a repetition of the mean seasonal cycle. SAZ

3. High res in situ MLD summer progression and variability 17 transects of XBTs to derive MLD Light - Fe threshold (Jourbert et al., submitted) SAZ Underway chl-a south of Africa

4.  Well stratified, punctuated by short winter mix  Summer highly reproducible but winter not  Dominated by heat fluxes. STZ APZ  MLDs are deep (±100m)  MLD is seasonal = 57% 57 % 57% 17 % 17% What do gridded datasets tell us? Monthly EN3-derived Brunt- Vaisala Frequency and MLD 14 % 14%  Weak seasonal cycle = 14%  Variable MLDs & weak strat.  Assoc with high wind var = 2.5 m.s-1 SAZ

5. HYPOTHESIS High rates of PP in SAZ are a direct result of MLD variability at submeso-subseasonal scales (around a threshold depth) that allows for alleviation of both light and Fe limitations at appropriate time scales for phytoplankton growth Swart et al., 2012 …At present we cannot do this without continuously sampling autonomous platforms! Unless these time scales (sub-seasonal) are correctly defined in terms physical – biogeochemical coupling, models will not accurately reproduce the seasonal cycle and hence predict future climate states

6. Gough Is. STF SAF APF Gough/Tristan Transect GoodHope Line Cape Town Bathymetry (meters) = Glider deployment & ship CTD station = ship based underway measurements ±2000 nm away… September 2012 – March 2013

7. SG573 SG574 SG543 SG575 SG542 Surface – 1000m 1.4 km horiz res 2532 dives = 5064 profiles 537 days of sampling + ship process study

8. FLUOR TEMP SPRING BLOOM PRIMING PERIOD SUMMER BLOOM SUSTAINING PERIOD

9. FLUOR BVF T f f

10. Cyclone Cyclone edge Front edge Intrusion Submeso filament -eddy Strat. (BVF) 0-100m & 100-300m MLD Fluor Temp Density Poster by S. Nicholson et al: PP sensitivities to submeso dyn and subseas atm forcing

11. FLUOR BVF

12. Wind R=0.52 Density MLD Fluor Strat. (BVF)

13. Spring – Summer MLD progression… a reminder of scales 5-hrly Glider Monthly EN3 CFSR 7-day

14. 1.Bloom initiations vary depending on the criteria used to define them. Different bloom initiations can be explained by different mechanisms (e.g. Sverdrup’s critical depth, Taylor and Ferrari’s turbulent convection, Mahadevan’s eddy driven stratification) >> The response of the bloom onset to interannual and climatic change will depend strongly on which mechanism prevails, eg. wind/features 2. In Spring, feature driven changes to the mixed drives early stratification and bloom initiations >> If climate models don’t include lateral processes they will overestimate bloom initiation dates 3. In Summer, wind driven adjustments to the mixed layer plays an important role in sustaining the summer phytoplankton bloom by relieving Fe and light limitation at the appropriate time scales >> Highlight the importance of interplay between meso-submesoscale features versus wind-buoyancy processes in characterising the ML, productivity, timing of the bloom and carbon export Conclusions

15. Many thanks to the following people and collaborators!  Geoff Shilling, Craig Lee &Eric D’Asaro at APL, UW  Derek &Andre at STS / SOERDC  Grant Pitcher & Andre Du Randt, DAFF  Stewart Bernard, Marjo Krug & Andy Rabagliati, CSIR  Gavin Tutt, DEA  IMT, SANAP, UCT, DEA & DAFF