On water vapor, surface temperature and the greenhouse e ect|a statistical analysis of tropical-mean data

Water-vapor feedback is one of the important factors that determine the response of the atmosphere to surface warming. To take into account the compensating drying e ects in downdraft regions, averaging over the whole tropics is necessary. However, this operation drastically reduces the number of degrees of freedom and raises question concerning the statistical signi cance of any correlative results obtained using observational data. A more involved statistical analysis is performed here, using multiple data sets, including the global water-vapor datasets of SSM/I (column water), UTH (upper tropospheric relative humidity), the TOVS retrieved upper tropospheric speci c humidity, and the surface temperature data from the NCEP/NCAR Reanalysis dataset. The tropical-mean correlations between speci c humidity and the surface temperature are found to be positive and shown to be statistically signi cant. This conclusion holds even when the averaging is done on the natural logarithm of the upper tropospheric water-vapor content. The e ect on the tropical mean Outgoing Long-Wave Radiation is also discussed.