A "zenith angle" effect results in the erroneous inference of IR brightness temperatures at pixels that are at relatively high viewing angles relative to nadir and are inferred to be considerably colder than they are in reality. Failure to correct for this effect results in discontinuities at the boundaries of adjacent geostationary satellites as demonstrated in Figure 1. A relatively crude correction for zenith angle effects on GPI precipitation estimates was determined by comparing retrieved IR temperature information from the GOES satellites which have a relatively large region of overlap and thus allowed the comparison of retrieved IR brightness temperatures from the same earth locations but at varying viewing angles (Joyce and Arkin 1997). More recent work by Joyce et al. (1999) shows that the viewing angle effect is dependent on the temperature of the retrieved pixel temperature, and that a seasonal adjustment is necessary in the extratropical regions (Figure 2). It is the incorporation of the results of these studies that enables the production of a nearly seamless map of IR data.

IR limb correction subroutines are available at the CPC ftp server. The IR correction needs original IR Tb, satellite zenith angle, Julian day, and pixel latitude for inputs and returns a limb corrected Tb.

References:

Joyce, R. J. And P. A. Arkin, 1997: Improved estimates of tropical and subtropical precipitation using the GOES Precipitation Index. J. Atmos. Ocean Tech., vol. 14, 997-1011.

Joyce, R. J., J. E. Janowiak and G. F. Huffman, 2001: Latitudinal and seasonal dependent zenith angle corrections for geostationary satellite IR brightness temperatures. (Journal of Applied Meteorology, in press).