Wind speed over water (WindSat)


RESTRICTED only accessable in CEN/MPI network or via CliSAP login What does that mean?

View data via LAS

Access data via OPeNDAP

Data access via file system: /data/icdc/atmosphere/windsat_surfacewind



The WindSat sensor aboard the Coriolis satellite has been the first satellite passive microwave sensor measuring the full stokes vector (at 37 GHz). This allows to retrieve the wind speed AND the wind direction over the ice-free ocean surface - the main products of this satellite sensor. In addition the multi-frequency, multi-polarization capability of this sensor (6.8 GHz, 10.7 GHz, 18.7 GHz, 23.8 GHz, 37.0 GHz) allows to further optimize the wind speed retrieval by contemporary measurements of parameters influencing this retrieval such as the sea surface temperature (SST), columnar values of the water vapor and the cloud liquid water content, and the rain rate (similar to, e.g. SSM/I used for the HOAPS data set).

Remote Sensing Systems (REMSS, http://www.remss.com) used WindSat data to compute the surface wind vector at 10 m height above the ocean surface. To do so various channel combinations are used. Here we only offer the all-weather wind speed product which comes at a somewhat coarser spatial resolution but is basically independent of rain effects. We refer to data set web page at REMSS: http://www.remss.com/missions/windsat and the references section for details.

The product offered here is a modified version of the daily WindSat ocean surface wind vector product v07.0.1 obtained from REMSS in early September 2015. The modification applied by ICDC is the conversion from flat binary into netCDF file format, and the computation of the u- and v-wind vector components from wind speed and direction.

Last data set update at ICDC: May 10, 2019.

To top


Name Units Comment Effective resolution
wind speed m/s all variables separately for ascending and descending overpasses 39km x 71km (under rain); 25km x 38km (else)
wind direction degrees (0 ... 360°) direction into which the wind blows 25km x 38km
u-component m/s -50 ... 50 39km x 71km (under rain); 25km x 38km (else)
v-component m/s -50 ... 50 39km x 71km (under rain); 25km x 38km (else)
rain rate mm/h 0 ... 25 8km x 13km
sea surface temperature (SST) degC -3 ... 34.5 39km x 71km
columnar water vapor content kg/m² 0 ... 75 16km x 27km
columnar cloud liquid water content kg/m² 0 ... 2.45 16km x 27km
decimal time hours since 0 UTC each swath pair has different overpass time --

To top

Coverage, spatial and temporal resolution

Period and temporal resolution:

  • 2003-02-05 to 2019-03-31
  • Daily (2 times, ascending & descending overpasses)

Missing days: Aug. 9-11 2017, Jan. 10-13 2019

Coverage and spatial resolution:

  • Global, over open water
  • Spatial resolution: 0.25° x 0.25°, cartesian grid
  • Geographic longitude: 0°E to 360°E
  • Geographic latitude: -90°N to 90°N
  • Dimension: 1440 columns x 720 rows
  • Altitude: 0.0 m


  • netCDF

To top

Data quality

The data set offered here does not include any explicit uncertainty estimations.

The antenna temperatures measured by the sensor from the fore and aft looking directions are processed into top-of-the-atmosphere brightness temperatures (TBs) and inter-calibrated with other satellite passive microwave instruments (see GMI_ATBD).

The product comes at 0.25° grid resolution. However, the effective resolution of the product depends on the footprint size of the sensors' channels used. All products which are using TBs from the 6.8 GHz channels with the coarsest spatial resolution have the corresponding effective spatial resolution which is given in a separate column in the table under parameters.

The advantage of a multi-frequency, multi-polarization instrument like WindSat is that contributions from atmospheric parameters such as water vapor, cloud liquid water and rain can be directly estimated from the same set of TBs (similar to what is done from SSM/I data for the HOAPS data set) which helps to optimize the wind speed and also wind direction retrieval. High values in any of these three parameters might still deteriorate the wind vector retrieval. In particular the wind direction retrieval using WindSat is hampered by rain rates above 15 mm/hour. Note further that wind direction cannot be retrieved under wind speeds below 3 m/s.

The capability of WindSat to measure TBs at 6.8 GHz and 10.7 GHz also allows to retrieve the SST which also helps to optimize the wind vector retrieval. At the same time, however, RFI and sun glint effect the TBs at these low-frequency channels and the SST retrieval and therefore also the wind vector retrieval; affected grid cells are flagged accordingly.

All above-mentioned parameters retrieved from WindSat data are included in the data set.

We refer to the section references for further information.

To top


Remote Sensing Systems
Santa Rosa, CA, U.S.A.
email: support (at) remss.com

Stefan Kern
ICDC / CEN / University of Hamburg
email: stefan.kern (at) uni-hamburg.de

To top


To top

Data citation

When using the data please cite as follows:

Wentz, F. J., L. Ricciardulli, C. Gentemann, T. Meissner, K. A. Hilburn, and J. Scott, 2013:  Remote Sensing Systems Coriolis WindSat Daily Environmental Suite on 0.25 deg grid, Version 7.0.1. Remote Sensing Systems, Santa Rosa, CA. Available online at www.remss.com/missions/windsat. [Accessed from www.remss.com,  last access date: 2019-05-02]. Downloaded in netCDF file format from the Integrated Climate Data Center (ICDC, icdc.cen.uni-hamburg.de) University of Hamburg, Hamburg, Germany.

In addition please add to the acknowledgements:

WindSat data are produced by Remote Sensing Systems and sponsored by the NASA Earth Science MEaSUREs DISCOVER Project and the NASA Earth Science Physical Oceanography Program. RSS WindSat data are available at www.remss.com.

###BACKLINK###     ###TOPLINK###