Spatial patterns of climate variability in upper-tropospheric water vapor radiances from satellite data and climate model simulations

The use of multivariate fingerprints and spatial pattern correlation in the detection and attribution of climate change has concentrated on radiosonde temperature fields. However, the large body of radiance data from satellit e-borne instruments includes contiguous datasets of up to 17 yr in length a nd in future years will present the most well-calibrated and large-scale da ta archive available for climate change studies. Here the authors give an e xample of the spatial correlation technique used to analyze satellite radia nce data. They examine yearly mean brightness temperatures from High Resolu tion Infrared Spectrometer (HIRS) channel 17, sensitive to upper-tropospher ic water vapor and temperature. Atmospheric profiles from a climate change run of the Hadley Centre GCM (HADCM2) are used to simulate the pattern of b rightness temperature change for comparison to the satellite data. Investig ation shows that strong regional brightness temperature changes are predict ed in the Tropics and are dominated by changes in relative humidity in the upper troposphere. At midlatitudes only small changes are predicted, partly due to a compensation between the effects of temperature and relative humi dity. The observational data showed some significant regional changes, espe cially at 60 degrees S, where there was a trend toward lower brightness tem peratures. The pattern similarity statistics revealed a small trend between 1979 and 1995 toward the predicted climate change patterns bur this was no t significant. The detection of any trend is complicated by the high natura l variability of HIRS-12 radiances, which is partly associated with the El Nino-Southern Oscillation.