UNDERSTANDING OCEAN SALINITY Stephen Riser Jessica Anderson University of Washington riser@ocean.washington.edu NASA SST Science Team Meeting October 29, 2013 [ Acknowledgements to NASA, NOAA, NSF, ONR ]

Ocean salinity: what is it, and why are we interested? (Hartmann, 1994) (Peixoto and Oort, 1992)

Ocean salinity: classical and modern salinity is tied to the hydrological cycle classical definition of salinity [ / = parts per thousand ] the chemical composition of seawater Perhaps surprisingly, the proportions of the major constituents of seawater are approximately constant nearly everywhere in the deep ocean.

the modern equation of state for seawater modern definition of salinity; the 1978 Practical Salinity Scale the modern equation of state for seawater the density (mass) distribution determines the flow, via Newton’s Laws 1.023 1.024 Data from UW float 6889 (WMO 5903283) in the subtropical N. Atlantic 1.025 1.026 1.027 1.028 1.029

The annual mean global sea surface salinity (PSU)

classical definition of salinity average this over a surface mixed layer of thickness h : “ocean rain gauge” advection of vertically-averaged salinity by the average flow advection of salinity by the sheared part of the horizontal flow entrainment/detrainment and/or obduction/subduction of salinity through the base of the mixed layer change of salinity at the sea surface through evaporation and/or precipitation small scale mixing

canonical Argo mission: 0-2000 m; T, S, p t = 10 days The Argo Float Array canonical Argo mission: 0-2000 m; T, S, p (0.005 C; 0.01 PSU ; 2.5 dbar) t = 10 days 6-7 years 250-300 profiles 32 countries, 800 floats/year deployed UW: 120 floats/year All data are publicly available with 24 hours of collection

SBE41: T  0.002C; S  0.005 PSU; p  2.4 dbars SBE CTD performance in Argo: generally excellent SBE41: T  0.002C; S  0.005 PSU; p  2.4 dbars Indian Ocean, 32S: 142 floats/shipboard CTD (1000 m) [UW laboratory check, N=380] Results from recovered floats DMQC results:  10 % adjusted

WOCE section P1 (dash/orange) 1994

Integrals along 24N: the upper ocean has freshened Argo minus WOCE Argo, 2007 WOCE, 1994 Integrals along 24N: the upper ocean has freshened [Ren and Riser, 2010]

T S  Upper ocean (250 m), changes in the past 25 yr [Roemmich and Gilson, 2009]

Float with auxiliary STS sensor and PAL hydrophone Argo float with CTD sensor (cuts off at 3 m) Float with auxiliary STS sensor and PAL hydrophone

T and S profiles (0-30 m) from the same float T and S profiles (0-2000 m) from UW float 6117, in the western equatorial Pacific T S S T T and S profiles (0-30 m) from the same float

Barrier Layer Thickness = Barrier Layers – Definition and Importance Mixed Layer MLDσ Barrier Layer MLDt Barrier Layer Thickness = MLDt – MLDσ MLDσ : 0.01 kg/m4 MLDT : 0.05 °C / m MLDS : 0.02 PSU/m Lukas and Lindstrom (1991)

Locations of 69 UW floats equipped with STS sensors

T and S from the upper 10 m of the ocean from float 6117 during a 3 week period when it was profiling 0-200 m at intervals of 2 hours

Float 6117 7/3/2009 Rainfall: 25 mm ∆S : 0.15 PSU ∆T : 0.18 °C TAO mooring 2°N, 147°E TRMM (3 hourly) STS Temperature 7/3/2009 Rainfall: 25 mm ∆S : 0.15 PSU ∆T : 0.18 °C   STS Salinity Hour (Local)

Comparisons of T and S (sea surface minus 15 m value) for the fast cycle period of float 6117 in the western equatorial Pacific

T and S differences between the sea surface and a depth of 4 m from float 6117, plotted as a function of wind speed measured at a nearby TAO mooring

T (C) S (PSU) Binned T and S anomalies as a function of local time, shown with binned colocated rainfall from TRMM. Diurnal cycles in T and S are clearly present.

Global variability in the amplitude of T and S diurnal cycles from all STS floats

Summary Ocean salinity is an indicator of the ocean’s role in the global hydrological cycle as well as a state variable for the ocean circulation. The salinity distribution is more than simply a measure of precipitation and evaporation. Argo data, and data from auxiliary STS sensors on some floats, has proven to be useful in understanding the global distribution of upper ocean salinity, especially very near the sea surface. The data can also be used in Aquarius/SAC-D validation exercises. Many storm events can be seen in the data, lasting several hours, changing the upper ocean temperature and salinity. There are measureable diurnal cycles in both T and S in the observations; T variability is clearly tied to diurnal heating in the atmosphere; S variability is related to diurnal signals in rainfall. Estimates of decadal-scale variability are only beginning; the longer that programs like Argo and Aquarius continue, the more we will understand long-term variability in the hydrological cycle and the ocean’s role in climate.

N = 16625 Global salinity difference (Aquarius minus Argo) for the period 9/2011-9/2013

Aquarius SSS October 2013