Comparison of tropospheric temperatures from radiosondes and satellites: 1979-98

A comprehensive comparison is made between two tropospheric temperature dat asets over the period 1979-98: the most recent and substantially revised (v ersion d) microwave sounding unit (MSU) channel 2 data retrievals, and a gr idded radiosonde analysis provided by the Hadley Centre of the U.K. Meteoro logical office. The latter is vertically weighted to approximate the deep l ayer temperatures measured by the satellite data. At individual grid points , there is good overall agreement among monthly anomalies, especially over the Northern Hemisphere continents where the climate signal is large, altho ugh monthly root-mean-square (rms) differences typically exceed 0.6 degrees C. Over the Tropics, correlations are lower and rms differences can be as large as the standard deviations of monthly anomalies. Differences in the g ridpoint variances are significant at many locations, which presumably refl ects sources of noise in one or both measurement systems. It is often argued for climate purposes that temperature anomalies are larg e in scale so that averaging over larger areas better serves to define the anomalies while reducing sampling error. This is the case for the Tropics ( 20 degrees S-20"N) where the large signal associated with El Nino-Southern Oscillation events is well captured in both datasets. Over the extratropics , however, the results indicate that it is essential to subsample the satel lite data with the radiosonde coverage in both space and time in any evalua tion. For collocated global average monthly anomalies, correlations are sim ilar to 0.9 with rms differences similar to 0.10 degrees C for both lower- (MSU2LT) and mid- (MSU2) tropospheric anomalies. The agreement between the satellite and radiosonde data is slightly better for the latest version of MSU2LT than it is for MSU2, in spite of the highe r noise levels of the former. This is primarily attributable to a strong wa rming trend in the MSU2 data relative to the radiosonde data toward the end of the record. Given the global nature of this discrepancy, it is suspecte d that it primarily reflects problems in the MSU analysis. As radiosonde re cords almost universally contain temporal inhomogeneities as well, caution is required when interpreting trends, which are not known to within 0.1 deg rees C decade(-1) However, the evidence suggests that global surface air te mperatures are indeed warming at a significantly faster rate than troposphe ric temperatures over the past 20 yr, and this is primarily attributable to physical differences in these two quantities.