Sea level varies as the ocean warms or cools, as water is transferred between the ocean and continents, between the ocean and ice sheets, and as water is redistributed within the ocean due to the tides and changes in the oceanic and atmospheric circulation. Sea level can rise or fall on time scales ranging from hours to centuries, spatial scales from <1 km to global, and with height changes from a few millimeters to a meter or more (due to tides). Sea level integrates and reflects multiple climatic and dynamical signals. Measurements of sea level are the longest-running ocean observation system. This section assesses interannual and longer variations in non-tidal sea level from the instrumented period (late 18th century to the present). Sections 4.3.3 and 4.4.2 assess contributions of glaciers and ice sheets to sea level, Section 5.6 assess reconstructions of sea level from the geological record, Section 10.4.3 assesses detection and attribution of human influences on sea level change, and Chapter 13 synthesizes results and assesses projections of sea level change.
The sea level observing system has evolved over time. There are intermittent records of sea level at four sites in Northern Europe starting in the 1700s. By the late 1800s, there were more tide gauges being operated in Northern Europe, on both North American coasts, and in Australia and New Zealand in the SH (Appendix 3.A). Tide gauges began to be placed on islands far from continental coasts starting in the early 20th century, but a majority of deep-ocean islands did not have an operating tide gauge suitable for climate studies until the early 1970s.
Tide gauge records measure the combined effect of ocean volume change and vertical land motion (VLM). For detecting climate related variability of the ocean volume, the VLM signal must be removed. One component that can be accounted for to a certain extent is the VLM associated with glacial isostatic adjustment (GIA) (Peltier, 2001). In some areas, however, VLM from tectonic activity, groundwater mining, or hydrocarbon extraction is greater than GIA (e.g., Wöppelmann et al., 2009; King et al., 2012); these effects can be reduced by selecting gauges with no known tectonic or subsidence issues (e.g., Douglas, 2001) or by selecting gauges where GIA models have small differences (Spada and Galassi, 2012). More recently, Global Positioning System (GPS) receivers have been installed at tide gauge sites to measure VLM as directly as possible (e.g., Wöppelmann et al., 2009; King et al., 2012). However, these measurements of VLM are only available since the late 1990s at the earliest, and either have to be extrapolated into the past to apply to older records, or used to identify sites without extensive VLM.
Satellite radar altimeters in the 1970s and 1980s made the first nearly global observations of sea level, but these early measurements were highly uncertain and of short duration. The first precise record began with the launch of TOPEX/Poseidon (T/P) in 1992. This satellite and its successors (Jason-1, Jason-2) have provided continuous measurements of sea level variability at 10-day intervals between approximately ±66° latitude. Additional altimeters in different orbits (ERS-1, ERS-2, Envisat, Geosat Follow-on) have allowed for measurements up to ±82° latitude and at different temporal sampling (3 to 35 days), although these measurements are not as accurate as those from the T/P and Jason satellites. Unlike tide gauges, altimetry measures sea level relative to a geodetic reference frame (classically a reference ellipsoid that coincides with the mean shape of the Earth, defined within a globally realized terrestrial reference frame) and thus will not be affected by VLM, although a small correction that depends on the area covered by the satellite (~0.3 mm yr –1 ) must be added to account for the change in location of the ocean bottom due to GIA relative to the reference frame of the satellite (Peltier, 2001; see also Section 13.1.2).
Tide gauges and satellite altimetry measure the combined effect of ocean warming and mass changes on ocean volume. Although variations in the density related to upper-ocean salinity changes cause regional changes in sea level, when globally averaged their effect on sea level rise is an order of magnitude or more smaller than thermal effects (Lowe and Gregory, 2006). The thermal contribution to sea level can be calculated from in situ temperature measurements (Section 3.2). It has only been possible to directly measure the mass component of sea level since the launch of the Gravity Recovery and Climate Experiment (GRACE) in 2002 (Chambers et al., 2004). Before that, estimates were based either on estimates of glacier and ice sheet mass losses or using residuals between sea level measured by altimetry or tide gauges and estimates of the thermosteric component (e.g., Willis et al., 2004; Domingues et al., 2008), which allowed for the estimation of seasonal and interannual variations as well. GIA also causes a gravitational signal in GRACE data that must be removed in order to determine present-day mass changes; this correction is of the same order of magnitude as the expected trend and is still uncertain at the 30% level (Chambers et al., 2010).