Atmospheric Structure Deduced from Routine Nimbus 6 Satellite Data

Abstract

The qualities of various data deduced from temperatures and dew points inferred from Nimbus 6 High Resolution Infrared Radiation Sounder (HIRS) and Scanning Microwave Spectrometer (SCAMS) radiances are assessed by comparison with similar data from weighted means of the rawinsonde observations (raobs) that bracketed the time of the satellite orbit across the east central United State for a case study on 25 August 1975. With notable exceptions, the HIRS-SCAMS temperature profiles are fairly good approximations of the weighted-mean raob profiles; those exceptions occur near the surface and near the tropopause. There are large discrepancies between the two sets of dew-point profiles at all levels that lead to questionable determination of precipitable water from satellite information. Thicknesses of layers suffer largest discrepancies at highest levels. Airmass stability may be appraised fairly reliably in regions of marked stability, much less dependably in unstable areas. There was no obvious correlation between the magnitudes of the discrepancies and the patterns of cloudiness observed either from the ground or from images from the GOES-East geostationary satellite. Gridding both the satellite and the mob data objectively on constant-pressure surfaces by the Barnes technique improves the degree of compatibility of the analyses. Standard deviations of the discrepancies between mob and satellite fields generally are larger than the average discrepancies themselves. Both the mean and the standard deviation of the discrepancies in geopotential height and in wind generally increase with altitude. The average discrepancies in temperature and dew point typically decrease with height in the troposphere and increase in the lower stratosphere; and the standard deviations in temperature and dew-point discrepancies are approximately constant with height. The magnitudes of the discrepancies between raob and Nimbus 6 data suggest that further improvement in satellite data is desired if these data are to be used as the principal base for synoptic analysis. However, a fair degree of reliance may be placed on the satellite data over areas from which conventional data are lacking.