An Analysis of Satellite Infrared Soundings at the Mesoscale Using Statistical Structure and Correlation Functions

Abstract

A statistical analysis of satellite infrared sounding data from the Vertical Temperature Profile Radiometer (VTPR) on NOAA 4 was performed in conjunction with the National Severe Storms Laboratory (NSSL) mesoscale sounding period (10 May-12 June 1976). Satellite radiances, retrieved temperatures and moisture information in the form of radiance residuals at a resolution of ?70 km were examined for a 14-day composite period using structure and correlation functions. A structure analysis as a function of data separation distance for a field of measured values can detect the mean nondirectional gradient in the field. Estimates of the relative noise level in the measurements were also obtained by extrapolating the obtained structure to zero separation distance. The rms radiance noise levels for the VTPR channels were found to be close to the design specifications for the VTPR instrument. For retrieved temperatures, the noise level was determined to be ?0.5°C at the three pressure levels examined. The structure functions for all available satellite-derived temperatures in the composite period compared favorably to similar results computed using high-resolution NSSL rawinsonde data. However, moisture correlation results demonstrated that the satellite-derived moisture in the integrated sense is not an equivalent substitute for mesoscale rawinsonde soundings. The structure-function analysis was also applied to each of the 14 individual days of available satellite data. The structure as a function of distance for VTPR channels 6 and 7 reflect mainly lower tropospheric temperature and moisture gradients, respectively. On days with both large temperature and moisture gradients as detected by the satellite these gradients appeared to be associated with severe storms later in the day. A structure-function analysis of satellite-derived 500 mb temperature fields was also found to detect existing upper air patterns on individual days. The information content of individual temperature and moisture fields derived from the satellite soundings was interpreted by comparison with similar fields from conventional rawinsonde soundings. Fields of both temperature and moisture from the two instruments were compared by direct correlation of the two sets of measurements. Discrepancies existed between the compared fields, but they were to a large extent explained by differences between the two measuring systems and time and space scale differences between the measurements. According to the analysis, temperatures were retrieved from the satellite data with reasonable success on most of the days at both 300 and 500 mb, but with much less success at 700 mb. Again, the moisture information extracted from the satellite data was not as promising due to its integrated effect and due to the small observed moisture gradients on some days. Structure-function analysis of satellite temperature soundings was shown to provide a means of interpreting these satellite data. It demonstrated that high-resolution satellite soundings provide information about spatial variations of temperature structure equivalent to that provided by high-density rawinsondes.