Tropospheric water vapour is of fundamental importance in determining climate and its response to forcing. Water vapour provides the most important feedback accounting for the large warming predicted by many climate models and about two thirds of the tropical contribution to this is due to the upper half of the troposphere. However the cold, dry conditions in this region have made comprehensive observational studies problematic, and consequently the simulation of Upper Tropospheric Humidity (UTH) in particular remains a considerable source of uncertainty in climate models.

Here we discuss aspects of the interannual variability of UTH as observed by the High Resolution Infrared Sounder (HIRS12) which has flown on all NOAA series satellites since 1979.

An Empirical Orthogonal Function (EOF) analysis has been performed separately on time-series of the 4 seasons DJF, MAM, JJA, and SON. In each season the principal mode of variability is strongly correlated (>0.8) to ENSO indices. Regional aspects of these modes are discussed with reference to low level convergence and Sea Surface Temperature (SST) gradients over the oceanic regions and other dynamical features. In boreal summer (JJA) the first two modes of variability are mixed according to North's rule-of-thumb but there is evidence suggesting that East-West SST gradients in the Indian ocean region are an important aspect of UTH variability in this season. A correlation analysis reinforces the importance of this region for understanding tropic-wide UTH variability.

The realisation of the distribution, variability and 20-year trends of HIRS12 UTH in the Met Office, Hadley Centre Atmospheric models forced with observed SSTs will be presented. Initial analysis of a new radiosonde humidity dataset currently in development will also be presented with particular reference to variations in the vertical profile of humidity through the depth of the free troposphere.