Global satellite mosaics - infrared, visible and water vapor

The combination of many different weather satellites makes it possible to continually monitor cloud conditions around the world. The Aviation Weather Center (AWC) has global forecast responsibilities so the AWC meteorologists have requirements for continual access to global weather information. In support of these requirements, the AWC has developed methods for generating mosaics of infrared, visible, and water vapor images. These mosaics utilize geostationary satellite data from the two GOES, the two Meteosat, and the Japanese satellites, in addition to using the US NOAA polar satellites.

The first step in creating the mosaics is the remapping of the raw satellite projection data into standard global projections. The AWC uses the Mcidas software from the Space Science and Engineering Center (SSEC) of the University of Wisconsin-Madison for this initial remapping. Where different satellites overlap, one must make the decision of how to merge the overlap regions. Possible methods include maximum or minimum of the overlap value, geographic division, or the most timely. The AWC uses the most timely data method when there is an overlap. This requires checks be ensure that data near the earth edge with considerably lower resolution than the less timely data are not included. Parallax corrections are also made. These adjustment are accomplished by changing the radius of the earth in the navigation calculations. The water vapor and infrared mosaics use a 10 km constant height parallax correction. The infrared mosaic uses the temperature calibration provided by the original satellite data supplier without modification. The water vapor has a limb darkening correction applied that corrects for the apparent colder temperatures nearer the edge of the earth. The visible mosaic is the most difficult mosaic to create because of different characteristics and calibration of the visible sensors on the different satellites, the changing sun angle, the need for brightness normalization for cloud surfaces, and what to do about the areas in the dark. The visible brightness normalization utilizes a cosine correction modified with a linear angular correction term to keep from over brightening clouds at low sun angles. The dark areas are filled with the “fog” image generated from the 11-3.9 micron channel difference when the satellite has the 3.9 channel. The Meteosat has a linear visible digitization while the GOES utilize a square root digitization (for 8 bit data). The Meteosat data is normalized and then converted to the same square root digitization as the GOES data.

The global mosaics are updated every 30 minutes and they are made available to the AWC forecaster’s workstations as ‘roamable’ global mosaics. The data are also available on the web in regional sectors at http://aviationweather.gov/obs/sat/intl/ir.shtml.