You are here: Home Oceanography From Space Our domains of research Air-sea interaction Retrieving turbulent fluxes from satellite

Retrieving turbulent fluxes from satellite

One the main key surface parameters involved in the exchange of energy between the atmosphere and oceans are: wind stress, surface turbulent latent and sensible heat fluxes. These are essential to improve modeling simulations of climate variations and oceanic dynamic process studies. Radars and radiometers onboard satellites provide valuable global measurements used to estimate the turbulent fluxes. The methodology for obtaining the surface turbulent fluxes uses physical properties of active and passive satellite instrument measurements, empirical and inverse models relating satellite observations and surface parameters, and objective analysis merging various satellite estimates. A high-resolution dataset is prepared for the global Ocean during 1992 - 2006, with a spatial resolution of 1 degree, and weekly and monthly temporal resolutions. The satellite data come from the European Remote Sensing satellite scatterometers (ERS-1 and ERS-2), NASA scatterometers (NSCAT and Seawinds onboard ADEOS-1 and QuikScat respectively), and several defense Meteorological Satellite Program (DMSP) radiometers (Special Sensor Microwave/Imager [SSM/I] F10 - F15). The reliability of the derived surface winds and heat fluxes is examined and validated through comprehensive comparisons with available in-situ data. The results are compared to NCEP/NCAR re-analysis and to ECMWF Analysis and re-analysis (ERA-40) wind and heat estimates. Comparisons are also performed with available remotely sensed flux estimates.


The remotely sensed winds and latent heat fluxes are mainly derived from the scatterometers onboard the European Remote Sensing Satellites (ERS-1 and ERS-2), NASA scatterometer (NSCAT) onboard ADEOS-1, Seawinds scatterometer onboard QuikSCAT, and from the radiometers onboard the Defense Meteorological Satellite Program (F10, F11, F13, F14, and F15). The periods of data availability are summarized in the following table:





December 1990 ? November 1997


December 1991 ? May 2000


May 1995 ? Present


May 1997 ? Present


December 1999 ? August 2006


August 1991 ? May 1996


March 1995 ? January 2001


September 1996 ? June 1997


July 1999 ? Present


The methods used to derive the satellite turbulent fluxes are described in Bentamy et al. (2003). They are based on the bulk aerodynamic formulae that parameterize the wind stress and latent heat flux :

(1) stress function
(2) Qlatent function

where ? is the wind stress vector and ? τ ?x, ?y are its zonal and meridional components, respectively. is the latent heat flux; (u10v10 medium), are the surface wind speed (zonal and meridional components) and air specific humidity at 10 m height and neutral stratification; and is the specific humidity at the sea surface equivalent to the saturation value at the sea surface temperature; l is the coefficient for latent heat of evaporation; ?is the air density; CDN and CEN are the bulk drag and the transfer coefficient for water vapor, respectively. The input variables, U10 mediumQa medium and sea surface temperature (SST), from which Qs medium is calculated, can be estimated from satellite observations ( e.g., Bentamy et al., 2002; Bentamy et al, 2003; Liu and Niiler, 1994; Schultz et al., 1993, 1997).

The quality of the derived surface winds and latent heat fluxes was investigated through comprehensive comparisons with buoy and ship estimates (Bentamy et al, 2002 and 2003). The remotely sensed flux observations are then used to estimate regular flux fields in space and time over the global ocean.

The present study employs the weekly- and monthly-averaged fluxes at 1° x 1° resolution available during the study period March 1992 ? March 2006. The accuracy of the resulting weekly fields is determined by comparisons with moored-buoy wind and latent heat flux estimates, which are deployed and maintained by four different institutions in the Atlantic and Pacific oceans. For instance, the agreement between satellite and in-situ data is good enough to suggest that Qlatent medium sources are achieving their accuracy goals. It was found that the satellite weekly ?x, ?y and Qlatent medium exhibited the main known spatial and temporal characteristics at global as well as at local scales. The local variability ofthe three surface parameters is well revealed by the satellite time series in tropical and in North Atlantic areas (with respect to buoy and ship data). The mean and rms difference between buoy and remotely sensed flux estimates are quite low. For instance, at the Tropical area, the bias values for wind stress and latent heat flux are 0.5 10-2N/m2 and 7.0W/m2, respectively. The corresponding rms values are 1.5 10-2N/m2 and 29W/m2.


Using an objective method, the multi-satellite wind vector, wind stress, latent and heat flux estimated over swaths are used to calculate weekly- and monthly averaged flux fields on 1 degree grid between 80°S and 80°N. The surface parameter fields are calculated at global scale, excluding sea ice areas. In order to determine the location of sea ice, the IFREMER/CERSAT weekly sea ice concentration is used. This parameter is estimated for both Arctic and Antarctic from the daily brightness temperature maps from SSM/I. For more details, see Ezraty et al. (2006). Data are freely available at IFREMER/CERSAT:

The turbulent flux fields are calculated during the period March 1992 through March 2006. They are considered as data test that will provide useful insight for production of high space and time resolutions at global and regional scales. Data are available at IFREMER and freely distributed upon request (Contact: This email address is being protected from spambots. You need JavaScript enabled to view it. , This email address is being protected from spambots. You need JavaScript enabled to view it. , and This email address is being protected from spambots. You need JavaScript enabled to view it. ). This study is performed within MERCATOR ( and MERSEA ( projects.

The data files are in NetCDF format supported by several scientific softwares.

The upward (downward) heat fluxes are considered positive (negative).

The file name is as TSATFLUX_yyyymmdd-yyyymmdd, where yyyy, mm, and dd are year, month, and day, respectively. It contains the gridded surface parameters: wind vector; wind stress, latent, and sensible heat fluxes.


Bentamy A., H-L Ayina, P. Queffeulou, D. Croize-Fillon ; 2006 : Improved Near Real Time Surface Wind Resolution over The Mediterranean Sea. Submitted to Ocean Journal.

Ayina H-L, A. Bentamy, G. Madec, Alberto M. MESTAS-NUÑEZ, 2006 : The Impact of Satellite Winds and Latent Heat Fluxes in a Numerical Simulation of the Tropical Pacific Ocean. J. of Climate Vol 19, Issue 22 (November 2006) pp. 5889?5902.

Mestas-Nunez A., A. Bentamy, and K. Katsaros, 2006 : Seasonal and El Ninõ variability in weekly satellite evaporation over the global ocean during 1996-1998. J. of Climate
Volume 19, Issue 10 (May 2006) pp. 2025?2035

J.A. Curry, A. Bentamy, M.A Bourassa, D. Bourras, E.F. Bradley, M. Brunke, S. Castro. S.H. Chou, C.A. Clayson, W.J. Emery, L. Eymard, C.W. Fairall, M. Kubota, B. Lin, W. Perrie, R.R. Reeder, I.A. Renfrew, W.B. Rossow, J. Schulz, S.R Smith, P.J. Webster, G.A. Wick, X. Zeng, 2004 : Seaflux, BAMS Vol. 85, No. 3, pp. 409?424.

Katsaros, K.B., A.M. Mestas-Nuñez, A. Bentamy, E.B. Forde, 2003: Wind bursts and enhanced evaporation in th tropical and subtropical Atlantic Ocean. In Interhemispheric Water Exchange in the Atlantic Ocean, G. Goni and P. Malanotte- Rizzoli (eds.). Elsevier Oceanographic Series.463 ? 474.

Bentamy A., K B. Katsaros, M. Alberto, W. M. Drennan, E. B. Forde, and H. Roquet, 2003 : Satellite Estimates of wind speed and latent heat flux over the global oceans, J. Climate, 16, 637 - 656.

Bentamy A., K B. Katsaros, M. Alberto, W. M. Drennan, E. B. Forde, 2002: Daily surface wind fields produced by merged satellite data. American Geophys. Union, Geophysical Monograph Series Volume 127, 343-349.

Bentamy A., Y. Quilfen, and P. Flament, 2002 : Scatterometer wind fields - a new release over the decade 1991 ? 2001. Can. J. Remote Sensing, 28, 3, 431-449.