AR4 Chapter 5 concluded that observed changes in upper ocean water masses reflect the combination of long-term trends and interannual to decadal variability related to climate modes like ENSO, NAO and SAM. The time series are still generally too short and incomplete to distinguish decadal variability from long-term trends, but understanding of the nature and causes of variability has improved in this assessment. The observed patterns of change in subsurface temperature and salinity (Sections 3.2 and 3.3) are consistent with understanding of how and where water masses form, enhancing the level of confidence in the assessment of the observed changes.
Recent studies showed that the warming of the upper ocean (Section 3.2.2) very likely affects properties of water masses in the interior, in direct and indirect ways. Transport of SST and SSS anomalies caused by changes in surface heat and freshwater fluxes are brought into the ocean's interior by cibtact with the surface ocean (Sections 3.2 and 3.3). Vertical and horizontal displacements of isopycnals due to surface warming could could change salinity and temperature (Section 3.3). Circulation changes (Section 3.6) could also change salinity by shifting the outcrop area of this isopycnal in regions with higher (or lower) E - P. Properties of several deep and bottom water masses are the product of near surface processes and signifcant mixing of etrainment of other ambient water masses (Section 3.5). Changes in the properties entrained or admixed water mass could dominated the observed deep and bottom water mass changes, for instance, in the LNADW and the AABW in the Weddel Sea.
From 1950 to 2000, it is likely that subtropical salinity maximum waters have become more saline, while fresh intermediated waters formed at higher latituted have generally become fresher. In the extratropical North Atlantic, it is very likely that the temperature, salinity, and formation rate of the UNADW is dominated by strong decadal variability related to NAO. It is likely that LNADW has cooled from 1995 to 2005. It is likely that the abyssal layer ventilated by AABW warmed over much of the globe since the 1980s or 1990s respectively, and the volume of cold AABW has been reduced over this time period.