Time Series Comparisons of Satellite and Rocketsonde Temperatures in 1978–79

Title Part Symbols d mean temperature dierence, K, (see eq. (1)) m sample size p pressure, hPa (or 100 Pa) s sample standard deviation, K, (see eq. (2)) T temperature, K t Student's t-test statistic (see eq. (3)) x station temperature value from single sounding, K Z geopotential height, km z geometric altitude, km Subscripts: j pair number (see eq. (1)) m sample size (see eq. (3)) 1 satellite (see eq. (1)) 2 in situ (see eq. (1)) Abbreviations: FSU former Soviet Union IFS in BLOCKINatable falling sphere LAIPAT LIMS inverted prole archive tape LAMAT LIMS map archive tape LIMS Limb Infrared Monitor of the Stratosphere MD mean dierence NMC National Meteorological Center NOAA National Oceanic and Atmospheric Administration RAOB radiosonde observation ROCOB rocketsonde observation SAO semiannual oscillation SD standard deviation SSU stratospheric sounding unit TIROS Television and Infrared Operational Satellite TOVS TIROS Operational Vertical Sounder UTC universal time coordinate VTPR Vertical Temperature Prole Radiometer WMO World Meteorological Organization iii Summary The Limb Infrared Monitor of the Stratosphere (LIMS) experiment made observations from the Nim-bus 7 satellite in 1978 and 1979. Temperature-versus-pressure, T (p), proles were derived from its limb radiance measurements, and those proles were used to register the radiances from other channels and to retrieve species concentration proles from those ra-diances. Therefore, biases in the T (p) results must be known in order to estimate the accuracy of those species proles. LIMS temperatures have been validated in the past with colocated radiosonde and rock-etsonde measurements. The present report describes time series comparisons between satellite and rocket-sonde T (p) values at station locations. This approach to validation retains nearly all the rocketsonde pro-les, increasing sample size signicantly (to 665). As a result, one can know better whether there is a bias that varies as a function of pressure altitude, latitude, or season. The results indicate no clearly signicant bias for LIMS versus Datasonde from 10 hPa to 1 hPa at low and mid latitudes. There is a positive LIMS bias of 2 to 3 K in the upper stratosphere at high latitudes for the Northern Hemisphere in both winter and spring. LIMS is progressively colder than Datasonde from 0.4 hPa (about 03 K) to 0.1 hPa (about 09 K) at all latitudes. A similar comparison between LIMS and the more accurate falling sphere measurements reveals an equivalent mid-latitude LIMS bias at 0.4 hPa but a much smaller bias at …