In the TAR, quantitative evidence for human influence on climate was based almost exclusively on atmospheric and surface temperature. Since then, anthropogenic influence has also been identified in a range of other climate variables, such as ocean heat content, atmospheric pressure and sea ice extent, thereby contributing further evidence of an anthropogenic influence on climate, and improving confidence in climate models.
Observed changes in ocean heat content have now been shown to be inconsistent with simulated natural climate variability, but consistent with a combination of natural and anthropogenic influences both on a global scale, and in individual ocean basins. Models suggest a substantial anthropogenic contribution to sea level rise, but underestimate the actual rise observed. While some studies suggest that an anthropogenic increase in high-latitude rainfall may have contributed to a freshening of the Arctic Ocean and North Atlantic deep water, these results are still uncertain.
There is no evidence that 20th-century ENSO behaviour is distinguishable from natural variability. By contrast, there has been a detectable human influence on global sea level pressure. Both the NAM and SAM have shown significant trends. Models reproduce the sign but not magnitude of the NAM trend, and models including both greenhouse gas and ozone simulate a realistic trend in the SAM. Anthropogenic influence on either tropical or extratropical cyclones has not been detected, although the apparent increased frequency of intense tropical cyclones, and its relationship to ocean warming, is suggestive of an anthropogenic influence.
Simulations and observations of total atmospheric water vapour averaged over oceans agree closely when the simulations are constrained by observed SSTs, suggesting that anthropogenic influence has contributed to an increase in total atmospheric water vapour. However, global mean precipitation is controlled not by the availability of water vapour, but by a balance between the latent heat of condensation and radiative cooling in the troposphere. This may explain why human influence has not been detected in global precipitation, while the influence of volcanic aerosols has been detected. However, observed changes in the latitudinal distribution of land precipitation are suggestive of a possible human influence as is the observed increased incidence of drought as measured by the Palmer Drought Severity Index. Observational evidence indicates that the frequency of the heaviest rainfall events has likely increased within many land regions in general agreement with model simulations that indicate that rainfall in the heaviest events is likely to increase in line with atmospheric water vapour concentration. Many AGCMs capture the observed decrease in Sahel rainfall when constrained by observed SSTs, although this decrease is not simulated by most AOGCMs. One study found that an observed decrease in Asian monsoon rainfall could only be simulated in response to black carbon aerosol, although conclusions regarding the monsoon response to anthropogenic forcing differ.
Observed decreases in arctic sea ice extent have been shown to be inconsistent with simulated internal variability, and consistent with the simulated response to human influence, but SH sea ice extent has not declined. The decreasing trend in global snow cover and widespread melting of glaciers is consistent with a widespread warming. Anthropogenic forcing has likely contributed substantially to widespread glacier retreat during the 20th century.