Improved Finite Impulse Response Filters for Enhanced Destriping of Geostationary Satellite Data

Abstract

The visible infrared spin scan radiometer (VISSR) common to many geostationary satellites produces visible wavelength data contaminated with stripes. The stripes affect the usefulness of the data in quantitative studies. This paper briefly reviews the cause of the striping and then develops a tapered window finite impulse response (FIR) filter and a constrained least squares FIR filter. Both filters minimize the stripes in the visible data and simultaneously minimize any distortion in the filtered data. These new filters are also designed to minimize ringing near cloud edges—a problem often associated with earlier FIR filter designs used to remove striped noise in satellite data. Finally, the results obtained by using these new filtering methods are quantitatively compared with those produced by other destriping methods. Results from geostationary operational environment satellite (GOES-7) and the geostationary meteorological satellite (GMS-5) scenes show that both filters are superior to the other methods evaluated. Moreover, they are computationally efficient because they are implemented in the spatial domain. Significant differences between the noise structure of GOES-7 and GMS-5 visible data also are discussed; these differences appear not to be in the open literature. The new FIR filter designs can also be adapted for use with data taken with the improved VISSR-like instruments on the GOES-8 and GOES-9 satellites. The importance of proper destriping of geostationary satellite data for geophysical applications is discussed.