To appear in the Proceedings of the Final TOGA Conference, Melbourne, Australia, April 2-7, 1995


Roger Lukas
University of Hawaii
Honolulu, Hawaii USA


Peter Webster
University of Colorado
Boulder, Colorado USA


At the time of planning for the TOGA program, the hydrological cycle was viewed as relatively passive in the ENSO phenomenon. The net effect on the thermodynamics of the coupled system was believed to be the transfer of latent heat from the ocean to the atmosphere and its subsequent release to the atmosphere during precipitation. The buoyancy gained by the atmosphere during this part of the hydrological cycle is matched by a buoyancy loss from the ocean, but historically this was not considered in ENSO studies. Neither was the buoyancy gain from precipitation. In particular, the distribution and transport of water in the ocean (the oceanic limb of the hydrological cycle) was neglected, even though salinity may play an important role in the tropical oceans' stratification, circulation, and mixing (Lukas and Lindstrom, 1991; Webster and Lukas, 1992).

Historically, the interannual anomalies associated with ENSO have been considered independently from the mean state and the annual cycle, although it was recognized that there was some relationship between the phase of ENSO and the phase of the annual cycle (cf. Rasmusson and Carpenter, 1982). While it was recognized that latent heat transport is important in the mean and seasonal climatology of the atmosphere over the Pacific, anomalies of diabatic heating were considered to be controlled more-or-less directly by SST anomalies. (The model of Zebiak and Cane [1987] showed that moisture convergence feedback results in a somewhat more complex relationship with maximum heating of the atmosphere possibly displaced from the maximum SST anomaly.) Most important was the assumption that SST anomalies could be simply related to the wind anomalies, through anomalous horizontal advection and through grossly simplified models of vertical mixing.

The purpose of this paper is to show that a complete consideration of the hydrological cycle leads to the conclusion that the mean state of the western Pacific warm pool depends on the existence of variability (transients, intraseasonal oscillations, and ENSO), as suggested by Lukas (1990). The assumption of a simple parametric dependence of interannual variability on the mean and seasonally-varying climate will not work in the modeling of the year-to-year variations of the warm pool.


The hydrological cycle is the transport of water within and between the different reservoirs in the earth's climate system, some transports involving a change of phase, such as evaporation from the ocean to the atmosphere (see Webster, 1994, for a comprehensive review). Processes involving phase changes (such as evaporation and precipitation) are very nonlinear, but even the transports within a reservoir are nonlinear due to the relationships between dynamics and thermodynamics. These nonlinearities support potential feedbacks which can result in unstable modes, oscillatory modes, and damped modes of variability. Figure 1 shows schematically the linkages between the various components of the coupled ocean-atmosphere system, including the hydrological variables q (water vapor) and s (salinity). The numerous potential feedbacks are obvious.

The cycling of water between liquid and vapor phases results in the transfer of heat from the oceans to the atmosphere. These phase changes result in highly nonlinear behavior of the coupled ocean-atmosphere system. This can easily be seen in the form of the bulk parameterization of evaporation:

E = C_e(q_o - q_A)U

where E is evaporation, Ce is a turbulent exchange coefficient, qo is the saturation specific humidity at the temperature of the sea surface, qA is the near-surface atmospheric specific humidity, and U is the magnitude of the near-surface wind. In general, Ce depends on the atmospheric stability, which depends on the flow, the air-sea temperature difference, and the lapse rate. In particular, at low wind speeds, Ce is strongly dependent on U (Liu et al., 1979; Fairall et al., 1995). There are both local and large-scale relationships between Ce, qo, and U, and evaporation is very obviously one of the strong sources of nonlinearity introduced by the hydrological cycle.

Evaporation causes a decrease of buoyancy of the surface waters by cooling and increased salinity, resulting in convection in the upper ocean. Precipitation provides buoyancy to the upper ocean, reducing vertical mixing (Lukas and Lindstrom, 1991) especially nocturnal convection (Anderson et al., 1995). This adds additional potential feedbacks into the coupled system (Figure 1).

Moisture is transported by winds and ocean currents, which are forced by horizontal buoyancy gradients. Diabatic heating is responsible for a major fraction of the buoyancy gradients, but the contributions from the distribution of water vapor in the atmosphere and salinity in the ocean are not negligible. Water vapor is buoyant relative to dry air, but most of the buoyancy flux from the ocean to the atmosphere is potential buoyancy associated with the latent heat flux. Thus the regions of free convection in the ocean and atmosphere can be separated by great distances. This is the situation in the Pacific Ocean, with significant convection in the central subtropical oceanic gyres of both hemispheres, and deep atmospheric convection over the warm pool (Figure 2).


Moisture extracted from the subtropical ocean under the Pacific trade winds converges over eastern Asia and the western Pacific warm pool (Trenberth and Solomon, 1994), providing fuel for the deep convection over both land and ocean. There is a direct counterpart of this transport in the ocean, with the moisture source for the atmosphere overlying a salinity source (moisture sink) for the ocean (Figure 2). The inexact correspondence between the net freshwater flux and the sea surface salinity (Figure 3) can be attributed to ocean currents and vertical mixing.

In order to compensate the moisture convergence over the warm pool, and the net freshwater flux, the ocean circulation supports a salt convergence into the warm pool within the upper thermocline circulation. This is the oceanic counterpart to the Walker and Hadley circulations of the atmosphere. Subduction of the saline waters created by the trade winds occurs as these waters are carried by the wind-driven ocean circulation into the western equatorial Pacific (Figure 3 and Figure 4; see also McCreary and Lu [1994] and Shinoda and Lukas [1995]). The resulting strong mean vertical salinity gradient in the warm pool inhibits vertical mixing, concentrating the heat from insolation in the near-surface layer of the warm pool where it can be most easily extracted by the atmosphere through turbulent heat fluxes on the short time and space scales which were the focus of the TOGA Coupled Ocean-Atmosphere Response Experiment (COARE). An imbalance between the net freshwater flux into the warm pool and the convergence of salt below will result in a modification of vertical mixing which feeds back onto the heat budget of the warm pool.

Of the 3-5 m annual average rainfall over this region, roughly 30­50% is imported, yielding a net freshwater flux of 1­2 myr¯ ¹ (Figure 3; Oberhuber, 1988). Thus, a significant recycling occurs over the warm pool, with the heat being supplied from the ocean. Results from COARE measurements indicate that the latent heat flux was between 90 and 150 W m¯ ² during the COARE IOP, with the lower end of the range during light winds typical of the warm pool (S. Anderson, personal communication). This heat was provided by the 160­240 W m¯ ² of penetrating shortwave radiation. (Other flux terms reduced the net flux into the ocean to 15­20 W m¯ ².) Thus, much of the diurnal heating of the upper ocean by penetrating solar radiation is released back to atmosphere in the form of latent heat. On what time scale does this occur? Because of nonlinearities in the coupling, it happens on a spectrum of time scales, including ENSO.


COARE observations suggest that the largest daily mean evaporation (and smallest daily mean insolation) takes place during the westerly wind burst episodes, which seem to be related to the 30-50 day intraseasonal oscillation, while the largest insolation (and smallest evaporation) occurs during the more typical light winds over the warm pool. Thus, these oscillations seem to play a major role in balancing the heat budget of the warm pool on the long term. They also play a role in balancing the freshwater budget through their impacts on local recycling of moisture, and in vertically entraining salty waters into the mixed layer. ENSO is another transient that may play a similar role in the long term balance of the warm pool.

The largest modulation of incoming solar radiation is obviously the diurnal solar cycle, while the variability of evaporation is strongest on the time scale of the transients and intraseasonal oscillations noted above. This suggests an oceanic mechanism which is sequestering a fraction of the diurnal heating and making it available to the atmosphere on these longer time scales. This mechanism is the penetration of solar radiation below the depth of the nocturnal mixed layer, which in the warm pool is largely controlled by the freshwater flux (Anderson et al., 1995). Siegel et al. (1995) measured typical values of 23 W m¯ ² passing through 30 m depth. Until surface winds are strong enough to mix below this depth, this region of the water column will continue to warm, and observations of frequent temperature inversions near this depth are consistent with this mechanism. Further support for this rectification of the diurnal cycle in the upper ocean comes from the observation that during the earliest stage of the December 1992 westerly wind burst, entrainment of deeper waters into the mixed layer appeared to contribute to a heating of the mixed layer (more than offset by latent heat loss, however.)



The support of NSF and NOAA under TOGA COARE (OCE-9024452) and by NASA under EOS (NAS5-31722) is gratefully acknowledged.


Anderson, S.P., R.A. Weller, and R. Lukas, 1995: Barrier layer formation in the western equatorial Pacific warm pool. Manuscript in revision.

Fairall, C.W., E.F. Bradley, D.P. Roger, J.B. Edson, and G.S. Young, 1995: Bulk parameterization of air-sea fluxes for TOGA COARE. J. Geophys. Res., submitted.

Ji, M., A. Leetmaa and J. Derber, 1995: An ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123, 460-481.

Levitus, S., 1982: Climatological atlas of the world ocean. National Oceanic and Atmospheric Administration Professional Paper 13, Rockville, Md., 173 pp.

Liu, W.T., K.B. Katsaros, and J.A. Businger, 1979: Bulk parameterization of air-sea exchanges of heat and water vapor including the molecular constraints at the interface. J. Atm. Sci., 36, 1722-1735.

Lukas, R., 1990: The role of salinity in the dynamics and thermodynamics of the western Pacific warm pool. International TOGA Scientific Conference Proceedings, WCRP-43, WMO/TD No. 379, 73-81.

Lukas, R. and E. Lindstrom, 1991: The mixed layer of the western equatorial Pacific Ocean. J. Geophys. Res., 96, suppl., 3343-3357.

McCreary, J.P., Jr. and P. Lu, 1994: Interaction between the subtropical and equatorial ocean circulations: The subtropical cell. J. Phys. Oceanogr., 24, 466-497.

Oberhuber, J.M., 1988: An atlas based on the COADS data set: The budgets of heat, buoyancy and turbulent kinetic energy at the surface of the global ocean. Max-Planck-Institute for Meteorology, Report No. 15.

Rasmusson, E.M. and T.H. Carpenter, 1982: Variations in tropical sea surface temperature and surface wind fields associated with the Southern Oscillation/El Nino. Mon. Wea. Rev., 110, 354-384.

Sadler, J., M.A. Lander, A.M. Hori, and L.K. Oda, 1987: Tropical marine climatic atlas, Volume II, Pacific Ocean. UHMET87-02, Department of Meteorology, University of Hawaii, 27 pp.

Siegel, D.A, J.C. Ohlmann, L. Washburn, R. Bidigare, C.T. Nosse, E. Fields, and Y. Zhou, 1995: Solar radiation, phytoplankton pigments and the radiant heating of the equatorial Pacific warm pool. J. Geophys. Res., 100, 4885-4891.

Shinoda, T. and R. Lukas, 1995: Lagrangian mixed layer modeling of the western equatorial Pacific. J. Geophys. Res., 100, 2523-2541.

Trenberth, K.E. and A. Solomon, 1994: The global heat balance: heat transports in the atmosphere and ocean. Clim. Dyn., 10, 107-134.

Webster, P.J., 1994: The role of hydrological processes in ocean-atmosphere interactions. Rev. Geophys., 32, 427-476.

Webster, P.J. and R. Lukas, 1992: TOGA COARE : The Coupled Ocean-Atmosphere Response Experiment. Bull. Am. Met. Soc., 73, 1377-1416.

Zebiak, S.E. and M. Cane, 1987: A model El Niņo-Southern Oscillation. Mon. Wea. Rev., 115, 2262-2278.

Return to UH COARE home page