Comparison of Sea Surface and Mixed Layer Temperatures

Comparison of sea surface and mixed layer temperatures.

Semyon A. Grodsky, James A. Carton, and Hailong Liu

March 28, 2008

To be submitted to the Journal of Geophysical Research, Oceans

Department of Atmospheric and Oceanic Science

University of Maryland, College Park, MD20742

Corresponding author:

Abstract

Mixed layer temperature, , and sea surface temperature, SST, are frequently used interchangeably or assumed to proportional in climate studies. This study examines the historical observational record for systematic differences between these variables in order to explore the appropriateness of these assumptions. The mixed layer depth in this study is defined to follow the seasonal mixed layer. Global and time mean mixed layer temperature is lower than SST by approximately 0.1 oC. This negative -SST difference is stronger in the upwelling zones (e.g. the Equatorial East Pacific) where an abundant net surface warming is compensated for by cooling across the base of the mixed layer. In the Equatorial East Pacific-SST varies seasonally, reaching a negative extreme in boreal spring when SST is warm, solar radiation is high, and winds areweak. In contrast, on interannual timescales-SST variesin phase with SST with small values in the equatorial Pacific during El Niños as a result of an eastwardshift of atmospheric convection and anomalously low solar heating. Moreover, during El Niños the intermittent cold SSTs and positive-SST differences occur in response to nocturnal cooling in presence of stabilizing near surface salinity gradient (barrier layer). Near surface fresheningalso producespersistent shallow (a few meters depth) warm layers in the northwestern Pacific during boreal summer when solar heating is strong. Surprisingly, in the Gulf Stream area of the northwestern Atlantic large positive -SST occur whenshallow cold and fresh layers develop in boreal winter in response to lateral interactions and abundant turbulent heat loss. An impact of shallow barrier layers on near surface temperature gradients is explored with 1-D mixed layer model.

1.Introduction

Although instantaneousthermodynamic fluxes across the ocean-atmosphere interface are affected by the temperature of the near surface ocean (< 1 m), many climate studies identify the vertically average temperature of the ocean mixed layer to be the most relevant parameter or even use a slab model as a proxy for the ocean mixed layer thermodynamics [see e.g. Manabe and Stouffer, 1996]. In general we may expect the depth average mixed layer temperature, , to be lower than SST by a few tenth of degree. This difference reflects the time average effect of turbulent suppression in the near surface layer during the daytime warming. In this study we evaluate the mixed layer temperature from the vertical temperature profiles provided by the World Ocean Database 2005 of Boyer et al. [2006] and analyze its spatial and temporal deviation, , from SST provided by the two major climate archives of Rayner et al. [2003] and Smith and Reynolds [2003].

The SST provided by these archives is based to a large extend on measurements collected by voluntary observing ships and is referred here as bulk SST or simply SST. It is widely acknowledged[1] that SST is a difficult parameter to define exactly because the ocean in the upper 10 m has complex and variable vertical temperature stratification.This transient shallow stratificationis more frequent when the ocean gains heat or freshwater. Its impact must be removed from raw measurements to allow the data ingest into consistent climate archives. The Global Ocean Data Assimilation Experiment Project suggests the foundation SST as a proxy for .The foundation SST is defined as the temperature at the top of water column free of diurnal temperature variability that normally doesn’t penetrate below the 10m depth. But the 10m depth temperature is provided only by scarce vertical temperature profiles. Routine operational in situ temperature measurements used in climate archives come fromships and buoys (both moored and drifting). Most ship observations weremade from insulated buckets, hull contact sensors, and engine room intakes at rather ill-defined depthvarying fromone to severalmeters.In general, much of operational in situ SST measurements are taken at depths of ~1-5 m where the diurnal warming impact is not vanishing.So we may expect the difference between and bulk SST due to impacts of diurnal warming and other of processes affecting shallow stratifications.

Recent observations using satellite day-minus-night SST differences[Stuart-Menteth et al., 2003; Gentemann et al., 2003] and simulations [Clayson and Weitlich, 2007]have revealed a large diurnal cycleof SST over wide areas where windsare weakand solar heating is strong. The combination of these conditions is more frequent in the tropics and midlatitudes.Besides that the diurnal variation may be similarly large at high latitudes in summer [Kawai and Wada, 2007]. In given latitude belt the strongest manifestation of diurnal warming is linked to the upwelling areas. In particular, Deser and Smith [1998] indicated from TAO buoydata that the mean diurnal amplitude of SST showed alocal maximum over the cold tongue in the eastern equatorialPacific.We expect that is colder than SST in the regions of major upwelling where isotherm outcroppingproduces a persistent stable temperature stratification up to the surface that indicates situations with essentially no mixed layer [Croninand Kessler,2002].

Although it is becoming clear that diurnal variability may affect longer period interactions in the ocean-atmosphere system, it is considered as the noise for the upper ocean heat content estimation, and special attempts are undertaken to remove this variability from SST products developed for climate studies. In particular, the Met Office Hadley Centre’s HadISST1 of Rayner et al. [2003] utilizes only the nighttime satellite SSTs and adjusts them to in-situ measurements taken from a few meters (depth of ship hull sensor or ship engine intake), while Smith and Reynolds [2003] use satellite information to evaluate the spatial correlations only.Data adjustment to measurements taken from a few meters depth (where the diurnal signal is relatively weak) effectively attenuatesbut doesn’t eliminate completely impacts of transient near surface processes.

The near surface processes that affect the ocean near surface stratification are dominated by (but not limited to) the diurnal warming that concentrates in the upper few meters of the water column and makes daily averaged SST warmer than . We will next see that in the presence of near surface salinity gradients a wide variety of near surface processes affecting the minus SST difference, , is possible.Tropical studies indicate that salinity gradients near thesurface also determine the spatial distribution of the surface warming [Soloviev and Lukas, 1997].Near surface freshening produces the barrier layers [Lukas and Lindstrom, 1991] that trap the heat near the surface by shoaling the penetration depth of wind stirring and nocturnal convection.Rains form shallow, buoyant layers that don’t mix with the water below except during strong wind events [Anderson et al., 1996].Moreover the stable salinity profiles may permit nocturnal temperature inversions [Anderson et al., 1996; Croninand Kessler,2002] with magnitudes comparable to those of diurnal warming.This strong nocturnal cooling takes place if a shallow surface layer cools but doesn’t overturn until the cooling is strong enough to overcome the stability introduced by salinity. Besides tropics, the barrier layers are observed over wide ocean areas, in particular, they are produced by an excess precipitation over the north Pacific and by lateral interactions of water masses in the western boundary currents [de Boyer Montégut et al., 2007]. In all these areas we also expect significant stratification of near surface layers that affect the differencebetween and SST because the latter is affected by this shallow stratification.

2.Data and Methods

This study defines the mixed layer temperature, , as the water temperature vertically averaged above the base of the mixed layer. The mixed layer properties are estimated from individual temperature profiles provided byWOD05 for the period 1975 through 2004. We use data from the mechanical bathythermographs (MBT), expendable bathythermographs (XBT), conductivity-temperature-depth casts (CTD), as well ocean station data (OSD), moored buoys (MRB), and drifting buoys (DRB). The final four years of the database contain an increasing number of profiles from the new Argo system (PFL). The Argo profiles for the period through 2007 are obtained from the Argo Project web site. For better characterizationof the tropical Pacific region the data provided by the TAO/TRITON moorings [McPhaden et al., 1998] are also used.

For this study the mixed layer temperature is evaluated based on the concept of isothermal mixed layer to take advantage of prevalence of the number of temperature only profiles over the number of profiles including both, temperature and salinity. The isothermal mixed layer depth (MLD) is defined following the methodology of de Boyer Montégut et al. [2004]who evaluate itfrom individual vertical profiles based on the temperaturedifference from the temperature at a reference depth of 10 m. This reference depth was shownto be sufficiently deep to avoid aliasing by the diurnal signal, but shallow enough to give a reasonable approximation of monthly .Here the isothermal MLD is defined as the depth at which temperature changes by || = 0.2oC relative to its value at 10m depth. Following Karaet al. [2000a],the isothermal MLD is defined by the absolute difference of temperature, ||, rather than only the negative difference of temperature to account for mixed layers with temperature inversions in salt-stratified situations (most common at high latitudes).

An alternative definition of the mixed layer depth (based on the dynamical stability criterion) defines it as the depth of a density uniform layer. Vertically average temperature of temperature uniform layer is the same as vertically average temperature of density uniform layer if the latter layer is not deeper than the former (barrier layer). If a density uniform layer is deeper than a temperature uniform layer (density compensation) their average temperatures may be different. Here we follow de Boyer Montégut et al. [2004]and define the mixed layer as a layer vertically uniform in both, temperature and salinity. Hence, the mean mixed layer temperature is the same as the mean temperature of isothermal layer. The latter is referred in this study as the mixed layer temperature, .

The mixed layer archive and the seasonal and interannual variability of mixed layer properties are described by Carton et al. [2008].They show that the temperature difference criterion works reasonably well even at high latitudes in the North Atlantic and provide further details on data quality control procedures. The mixed layer temperature is assessed as the water temperature vertically averaged above the base of the mixed layer assuming that water temperature above the reference depth is uniform,. After estimatingat each profile location we then apply subjective quality control to remove ‘bulls eyes’ and bin the data into 2ox2ox1mo bins with no attempt to fill in empty bins.

Bulk SST is provided by Met Office Hadley Centre sea ice and sea surface temperature(HadISST1)of Rayner et al. [2003] and by Smith and Reynolds [2003].Both products are globally complete monthly averaged grids spanning time period beginning the late 19-th century onward. HadISST1 combines a suite of historical and modern in situ near surface water temperature observations from ships and buoys with the recent satellite SST retrievals, while the Smith and Reynolds [2003] datais mostly based on in-situ measurements. Neither of these products usesthe vertical temperature profiles fromWOD05.

The local response of the mixed layer to the forcing from the atmosphere is simulated using the 1-D hybrid mixed layer model of Chen et al. [1994]. This model is based on the Kraus-Turner-type bulk mixed layer physics in which the depth of the mixed layer is determined by a turbulent energy balance equation, while the temperature and salinity of the mixed layer is determined by budget equations forced by surface fluxes and entrainment. These balances are augmented in the Chen et al. formulation by the addition of convection and Richardson Number-dependent mixing. The model is forced by 6-hour surface fluxes provided by the NationalCenter for Environmental Predictions/Department of Energy (NCEP/DOE) Reanalysis-2 of Kanamitsu et al. [2002].

3. Results

We begin by examining the time average in order to identify regions where it is significant. Time average (Figs. 1a, 1b) is evaluated during 1960-2004 using evaluated from the WOD05 vertical temperature profiles and bulk SST provided by the two climatic archives. At each grid point the temperature difference contributes to the time average during only those months when is available. On average, the mixed layer temperature is colder than SST by about 0.1 oC that may be related to the remaining contribution of the shallow diurnal warming that is not completely filtered out of the bulk SST data.The spatial patterns of time averaged are similar for both SST products and are persistent in time. Comparison of evaluated from the WOD05 profiles with (Fig.1c) evaluated from the Argo profiles (spanning the 1997-present with much of coverage coming during the 2004-present) shows similar patterns in the equatorial belt and the northern Hemisphere. Argo based also shows substantial variability in the Antarctic Circumpolar Current area where the WOD05 data coverage is not good. These patterns are not analyzed because they are evaluated from relatively short time records.

In Figs.1a and 1b the three regions are identified where is relatively strong. These regions are found in the equatorial eastern Pacific, the Gulf Stream, and the northwestern Pacific. It will be seen next that all these regions are distinguished by persistent shallow near surface stratification due to either upwelling or impact of the barrier layers that trap warming (cooling) in the near surface. On the other hand, the air-sea interactions are particularly strong over these regions. It is illustrated by seasonal maps of the net surface heat gain by the ocean. During the cold season (Fig. 2a) the turbulent heat lossin excess of 200 Wm-2occurs over the warm western boundary currents due to strong air-sea temperature contrast and enhanced evaporation over warm SSTs. In northern summer (Fig. 2b) the ocean gains heat in excess of 150 Wm-2 in the northwestern Pacific and over the shelf waters north of the Gulf Stream. In both these areas the local increase of the ocean heat gain is due to a decreased evaporation over cool SSTs. The ocean also gains heat at a rate exceeding 100 Wm-2 in the eastern equatorial Pacific cold tongue (Fig. 2b)due to abundant solar radiation and relatively weak latent heat loss.In the cold tongue the heat gain is compensated for by entrainment cooling. In the near surface it produces remarkable magnitudes of diurnal warming. We shall next analyze the origins of persistent shallow stratifications in these three regions.

3.1Eastern Equatorial Pacific

Although the diurnal warming is strong in the western Pacific warm pool [Soloviev and Lukas, 1997], the seasonal thermocline is rather deep in this area. Hence, impacts of nocturnal convection and wind stirring are also deep, and nighttime SST is close to the bulk SST. Under these conditions animpact of diurnal warming is eliminated from the bulk SST data that results in a good regional correspondence with (Fig.1). Further eastward the mixed layer shoals and the magnitude of diurnal warming increases [Deser and Smith, 1998]. Here in the area of the east equatorial Pacific cold tongue is persistent and averages approximatelyminus 0.4oC during 1960-2004 (Fig.3a).The seasonally averaged and SST vary out of phase (Fig. 3b) following the seasonal variation of winds. Although the sun crosses the equator twice a year, in March and September, the diurnal warming haspredominantly annual oscillation that enhances in a warm season from February through May and attenuates in a cold season from June through November following the annual migration of the ITCZ and winds [Cronin and Kessler, 2002]. In contrast, interannual variations of occur coherently with interannual variations of the Southern Oscillation Index (SOI), decreasingin magnitude during the El-Niños when the mixed layer deepens and atmospheric convection enhances in the east, and increasing in magnitude during the La-Niñas when the mixed layer shoals and atmospheric convection shifts westward (Fig.3 a).Theseinterannual variations of occur in accord with interannual variations ofupwelling and diurnal warming that both enhanceduring the La Niña years and weakens during the El Niño yearsin response to interannual variations of convection, solar radiation, and zonal winds [Cronin and Kessler, 2002; Clayson and Weitlich, 2005]. The coherent interannual variations of and SOI are more evident after the 1975 (Fig.3a), while before the 1975 the interannual variations of are statistically insignificant as their magnitude doesn’t exceed the width of the standard deviation corridor. The longterm variations of indicate a tendency of decreasing magnitude during the 1960-2004. If we assume that the longterm variations of reflects the longterm variations of upwelling in the EqEP region (as it occurs at interannual time scales), than the tendency of decreasing magnitude of signals on the tendency of weakening of the upwelling. If that is true, the equatorial Pacific is shifting towards an El Nino –like state.