Using NCEP reanalyses for the period 1973–2007, Paltridge et al. (2009) found negative trends in specific humidity above 850 hPa over both the tropics and southern mid-latitudes, and above 600 hPa in the NH mid-latitudes. However, as noted in AR4, reanalysis products suffer from time dependent biases and have been shown to simulate unrealistic trends and variability over the ocean (Mears et al., 2007; John et al., 2009) (Box 2.3). Some reanalysis products do reproduce observed variability in low level humidity over land (Simmons et al., 2010), more complete assesments of multiple reanalysis products yield substantially different and even opposing trends in free tropospheric specific humidity (Chen et al., 2008; Dessler and Davis, 2010). Consequently, reanalysis products are still considered to be unsuitable for the analysis of tropospheric water vapour trends (Sherwood et al., 2010).
In summary, radiosonde, GPS and satellite observations of tropospheric water vapour indicate very likely increases at near global scales since the 1970s occurring at a rate that is generally consistent with the Clausius-Clapeyron relation (about 7% per degree Celsius) and the observed increase in atmospheric temperature. Significant trends in tropospheric relative humidity at large spatial scales have not been observed, with the exception of near-surface air over land where relative humidity has decreased in recent years (Section 2.5.5).