Large-scale spatial patterns of sea level change are known to high precision only since 1993, when satellite altimetry became available (Figure 3.10). These data have shown a persistent pattern of change since the early 1990s in the Pacific, with rates of rise in the Warm Pool of the western Pacific up to three times larger than those for GMSL, while rates over much of the eastern Pacific are near zero or negative (Beckley et al., 2010). The increasing sea level in the Warm Pool started shortly before the launch of TOPEX/Poseidon (Merrifield, 2011), and is caused by an intensification of the trade winds (Merrifield and Maltrud, 2011) since the late 1980s that may be related to the Pacific Decadal Oscillation (PDO) (Merrifield et al., 2012; Zhang and Church, 2012). The lower rate of sea level rise since 1993 along the western coast of the United States has also been attributed to changes in the wind stress curl over the North Pacific associated with the PDO (Bromirski et al., 2011). While global maps can be created using EOF analysis (e.g., Church et al., 2004; Llovel et al., 2009), pre-1993 results are still uncertain, as the method assumes that the EOFs since 1993 are capable of representing the patterns in previous decades, and results may be biased in the middle of the ocean where there are no tide gauges to constrain the estimate (Ray and Douglas, 2011). Several studies have examined individual long tide gauge records in the North Atlantic and found coherent decadal-scale fluctuations along both the USA east coast (Sturges and Hong, 1995; Hong et al., 2000; Miller and Douglas, 2007), the European coast (Woodworth et al., 2010; Sturges and Douglas, 2011; Calafat et al., 2012), and the marginal seas in the western North Pacific (Marcos et al., 2012), all related to natural climate variability.
There is still considerable uncertainty on how long large-scale patterns of regional sea level change can persist, especially in the Pacific where the majority of tide gauge records are less than 40 years long. Based on analyses of the longest records in the Atlantic, Indian and Pacific Oceans (including the available gauges in the Southern Ocean) there are significant multi-decadal variations in regional sea level (Holgate, 2007; Woodworth et al., 2009, 2011; Mitchum et al., 2010; Chambers et al., 2012). Hence local rates of sea level rise can be considerably higher or lower than the global mean rate for periods of a decade or more.
The preceding discussion of regional sea level trends has focused on effects that appear to be related to regional ocean volume change, and not those due to vertical land motion. As discussed in Section 3.7.1, vertical land motion can dramatically affect local sea level change. Some extreme examples of vertical land motion are in Neah Bay, Washington, where the signal is +3.8 mm yr –1 (uplift from tectonic activity); Galveston, Texas, where the value is –5.9 mm yr –1 (subsidence from groundwater mining); and Nedre Gavle, Sweden where the value is +7.1 mm yr –1 (uplift from GIA), all computed from nearby GPS receivers (Wöppelmann et al., 2009). These areas will all have long-term rates of sea level rise that are significantly higher or lower than those due to ocean volume change alone, but as these rates are not related to climate change, they are not discussed here.