Swath Data Sets

Data from polar orbiting satellites is observed and archived in swaths, relative to the satellite track, rather than a global grid. The pattern of observations is nearly regular, and forms a grid with axis parallel and perpicular to the satellite track. The spatial resolution of the data sets is usually 25 or 50 km.

Several forms of NSCAT gridded data will be available on-line. Links to descriptions of the data sets are available. The following general table outlines the available gridded data products. Further explanations of winds, pseudostresses, and stresses are available.

Information on Swath Data Sets

The available data sets are described below. The data sets are listed in terms of the short technical name, which will seem like NASA jargon to people unfamiliar with such data sets. Links to the data and detailed descriptions of the data set content and format are included. There is also a general description of swath data.

Background: The Satellite's Orbit

Understanding of scatterometer observation patterns can be useful for utilizing swath observations. Satellite scatterometers are on polar orbiting satellites, which move around the Earth. They are not geostationary, meaning their orbits are not such that they are always above a constant location on the planet. Motion relative to the Earth's surface is essential for current scatterometers: it allows the instrument to view the surface from several angles, which are used to determine the wind direction. However, because of this motion, observations are not in a regular and unchanging grid with respect to the surface. A scatterometer's observational grid has an approximately regular grid with respect to the position of the satellite track (e.g., NSCAT over the Indian Ocean). This track is the line made by tracing the points on the Earth's surface directly below the satellite.

Observational Swaths

For the NASA Scatterometer (NSCAT) there was a 400 km band, centered on the satellite track, in which there were no observations. On each side of this band, there are 600 km wide bands of observations. These bands are referred to as observational swaths. The gap between these observations is called the 'nadir gap'. Note that this instrument design differs from that used in the ERS-1 and ERS-2 scatterometers: they made observations on only one side of the satellite, resulting in roughly half the coverage of NSCAT. In the case of QuikSCAT, there is no nadir gap, and the observations extend an extra 50 km beyond the far edges of NSCAT's observations. QuikSCAT has one very wide swath.

Understanding Swath Data Sets

For NSCAT and QuikSCAT the resolution of the wind data sets is 25 km (a 50 km resolution product is also available for NSCAT winds). For ERS-1 and ERS-2 scatterometers, the resolution is 50 km. The observations are arranged in a series of rows of cells, where the rows are across the swathes (i.e., perpendicular to the satellite track). There are locations (latitude and longitude) specified for each of the cells. For each of the rows, the data sets contain a mean time of the observations from the different viewing angles used to determine the winds. The 25 km resolution NSCAT data set has 48 cells per row (24 for each swath), and QuikSCAT swath data has 76 cells in each row.

The NSCAT and QuikSCAT observations are processed and released by the Jet Propulsion Laboratory. Each file contains the winds (and/or various other information) from one orbit. These files contain the 1 to 4 most likely solutions (called ambiguities) for the wind speed and direction. The file name contains the orbit, and this information can be used to estimate the times and locations of the observations in the file. However, this approach to finding specific temporal subsets is not intuitive to many users of satellite data. Compact data sets are also available in a more intuitive arrangement. These are designed to have the minimum necessary information (thereby reducing size), and an internally documented and easy to use format (netCDF). These files contain only selected vectors (rather than each of the 1 to 4 most likely solutions). The files contain all observations from a single day, with the day and year clearly identifiable in the file name.

Types of Swath Data Sets

There are several types of swath data, commonly refered to as levels. The lower levels contain the data in engineering units, and data has gone through relatively little processing (meaning the data sets are often large). Level zero data, which is rarely made public, has the obsersations expressed in units such as volts or amps. Level one data (e.g., NSCAT L1.7 and QuikSCAT L1B) has the data in instrument specificunits (e.g., backscatter for scatterometer data). Level 2 data (e.g., NSCAT L2.0 and QuikSCAT L2B) is processed into geophysical units. For scatterometers, level 2 data has units for speed (ms) and direction (degrees). There are also hybrid data sets, containing both level 1 and level 2 data (e.g., NSCAT W25 and QuikSCAT L2A). These hybrid data sets tend to be very large. The most likely data sets to be used for oceanography and meteorology applications are not swath data, but the level 3 data sets, which are gridded on longitude-latitude axes. However, for applications requiring a single image from one or two satellite passes, the swath data provide the finest spatial resolution.