Southern Ocean Surface Measurements and the Upper Ocean Heat Balance Janet Sprintall Sarah Gille Shenfu Dong Scripps Institution of Oceanography, UCSD Challenges in the Southern Ocean Western Boundary Currents Southern Ocean data rich data poor sampling possible all year round very few winter observations heat transport from low to high latitudes circumpolar SST probably means little net heat transport ocean heat transport primarily geostrophic strong westerlies drive strong meridional Ekman transport fairly reliable, validated surface heat flux products heat flux products with very large uncertainties Talk Outline: 1. Status of shipboard observations in the Southern Ocean 2. Science Applications:a. Variability in the Antarctic Polar Front b. The upper ocean heat budget in the Southern Ocean 3. Conclusions: Implications for data sampling requirements in the Southern Ocean Southern Ocean HR-XBT Measurements USA-SIO; Aust-CSIRO; NZ-NIWA; France www-hrx.ucsd.edu NOAA-AOML www.aoml.noaa.gov/phod/hdenxbt/high_density_home.html IX21 IX15 PX50 PX14 AX18 AX25 PX08 IX28 Italian CLIMA PX81 AX22 U.S-Chinese moon.ldgo.columbia.edu/~xiaojun/xbt/ Drake Passage Measurements Depth averaged ADCP velocity wind speed (m/s) pCO2 (atm) salinity (psu) XBT temperature PIs: Sprintall (XBT); Takahashi, Sweeney (pCO2); Chereskin, Firing (ADCP) Science Application: 1. Variability in the Antarctic Polar Front* MODIS (Infrared) Lots of cloudy or bad data AMSR-E (Microwave) AMSR-E: cloud penetration *Dong, Sprintall & Gille, Location of the Antarctic Polar Front from AMSR-E Satellite Sea Surface Temperature measurements, J. Phys. Oceanogr., in press, 2006. Comparison of the PF Location from XBT and AMSR-E SST Subsurface Polar Front from XBT (northern extent of the 2°C isotherm at 100300m depth) Surface Polar Front from AMSR-E (southernmost location of an SST gradient above 1.5x10-2°C km-1) Mean Polar Front Location AMSR-E (2002-05) Dong et al. (2006) AVHRR SST (1987 – 1993) Moore et al. (1999) Deep ocean basin with weak bottom slope: large PF variability What controls the Polar Front Variability? 1. Response of PF location to meridional shifts in the wind field (∂PF/∂t ~ ∂(x)/∂t) wind wind PF coherence 60% > 95%CI phase PF phase negative phase: shift in latitude of maximum zonal wind stress leads meridional shift in PF histogram of phase Science Application: 2. Southern Ocean Upper Ocean Heat Budget* Domain Averaged Surface Layer Heat Balance (weekly resolution on 1°x1° grid) Tm Qnet q(hm) weT 2 um Tm Tm t ocphm hm * Dong, Gille and Sprintall, Heat budget of the Southern Ocean, in prep, 2006 Horizontal Advection Tm Qnet q(hm) weT 2 um Tm Tm t ocphm hm geostrophic advection (AVISO SSHa plus GRACE) Ekman advection (COAPS wind stress) Tm mixed layer temperature AMSR-E SST Imbalance of the Heat Budget Analysis Tm Qnet q(hm) weT um Tm 2Tm t ocphm hm Largest imbalance in winter (~100 W m-2) “Best Case” rms of the imbalance is 146 Wm-2 (0.031°C/day) (NCEP1 air-sea heat fluxes; ARGO density MLD; diffusion =500 m2s-2; spatially-variable ∆T from ARGO) Sensitivity: 1. Surface heat flux products Spatial rms of Qnet (Wm-2) (1 Jan 2000 - 31 August 2002) NCEP1-NCEP2 RMS of heat balance: NCEP1: 146 Wm-2 NCEP2: 148 Wm-2 (June 02 - Dec 05) NCEP1-ECMWF NCEP1-SOC (monthly clim) Tm Qnet q(hm) weT um Tm 2Tm t ocphm hm Sensitivity: 2. Mixed Layer Depth (hm) Tm Qnet q(hm) weT um Tm 2Tm t ocphm hm hm from ARGO float data (∆=0.03kg/m3) Variable hm Time-Mean hm Sensitivity: 2. Mixed Layer Depth (hm) Tm Qnet q(hm) weT um Tm 2Tm t ocphm hm RMS of Heat Balance for Various MLD: 1. Argo Floats climatology (Jun 02 - Dec 05): - density diff 0.03kgm-3: 0.031°C/day (146 Wm-2) “best” case - temp diff 0.2°C: 0.033°C/day (157 Wm-2) 2. de Boyer Montegut et al. (2004) climatology: - density diff 0.03kgm-3: insufficient data - temp diff 0.2°C: 0.037°C/day (151 Wm-2) 3. WOA 2001 climatology: - density diff 0.125kgm-3: 0.038°C/day (235 Wm-2) NB: Density MLD criteria uses variable ∆T in entrainment term; temperature MLD criteria uses ∆T=0.02°C. T Qnet q(h ) w T Sensitivity: u T 2T t ch h 3. ∆T across the base of the MLD Annual Average ∆T from Argo floats m m e m o p m m m m • many studies use ∆T=0.2°C (eg. Qiu and Kelly, 1993) • for ∆T=0.2, rms of imbalance is 159 Wm-2 cf “best” case of 146 Wm-2. • largest improvement in Indian Ocean where ∆T can be negative: salinity matters! • differences in ∆T are largest in fall and winter when entrainment is strongest Spatial Variability in the Imbalance of the Heat Budget • Largest imbalance north of ACC: hm is large & temperature gradient is strong • complex upper ocean processes not well resolved by existing measurements Regional Heat Budget: The Agulhas Retroflection March Average (x10-6 °Cs-1) cooling warming • strong SST gradient and large meanders in Agulhas region • reanalysis Qnet has long length scales, while we & advection have smaller scales: net effect on imbalance is small scale structure. • large imbalances related to small-scale coupling of the wind field and SST? (e.g O’Neill et al. (2003) find wind stress curl (divergence) related to cross (down) wind components of SST gradient) Conclusions: Implications for data sampling requirements 1. Shipboard Measurements - met & pCO2 sampling opportunity: need identified PIs. 2. Science Application: Polar Front Variability - weekly and daily SST fields resolve similar PF locations - winds are important! Global forcing fields from satellite. - winds more energetic at higher frequencies 3. Science Application: Upper Ocean Heat Balance - large uncertainties in all terms! - Qnet varies enormously. Need validation with in situ data and winter time measurements - salinity matters! Need Argo floats for MLD and ∆T - need spatial resolution ~0.25° for small-scale coupling - weekly and daily SST fields give similar heat balance What controls the Polar Front Variability? 2. Response of PF Transport to changes in zonal wind stress du/dt (~ u==dT/dy) and x wind coherence 70% > 95%CI ACC phase wind ACC histogram of phase negative phase: change in zonal wind stress leads to changes in PF transport Southern Ocean Heat Balance by Basin