Rahmstorf (2009) range (78–175 cm). The committee’s results differ from the IPCC (2007) results because the committee considered more recent scientific observations and modeling and also used different methods to make projections. For example, although the steric contributions were drawn from the same global climate models used in IPCC (2007), the committee used the global climate model results directly, whereas IPCC (2007) used lower-order models to develop estimates for emission scenarios that were not simulated in global climate models (e.g., A1FI). In addition, the committee used extrapolation methods to project the cryosphere component of sea-level rise, whereas IPCC (2007) used climate models.
The global sea-level projections shown in Figures 5.4 and 5.5 do not include contributions from groundwater depletion and reservoir extraction. Estimates available at the time this report was being written (e.g., Milly et al., 2010) suggested that the sum of these contributions was near zero, within the stated uncertainties. Although some studies have pointed out that the number of new reservoirs has been declining over the past three decades (e.g., Chao et al., 2008), the committee had no firm basis for projecting a growing contribution to sea-level rise from groundwater extraction. A new paper published as this report was nearing release, however, projects that increasing groundwater extraction and decreasing reservoir impoundment will contribute about 1.5 ± 0.8 cm SLE to global sea level in 2030, 3.1 ± 1.1 cm SLE in 2050, and 7.5 ± 2.0 cm SLE in 2100 (Wada et al., 2012b). If confirmed by subsequent analyses, these results indicate that changes in the balance of groundwater depletion and reservoir impoundment could increase the magnitude of future sea-level change.
Only a few studies have attempted to project 21st century sea-level rise along the west coast of the United States. Methods varied, but each study used global climate models forced by the IPCC (2000) low and high greenhouse gas emission scenarios. The results were then downscaled, used in semi-empirical projections, or combined with local information, as discussed below. Each of the studies emphasized that the results represented a range of outcomes, not formal projections of sea-level rise.
The earliest of these studies, Hayhoe et al. (2004), used two global climate models, downscaled to a 150 km2 grid, to simulate climate change in California. Projections of various aspects of climate change were averaged over the 2020–2049 and 2070–2099 periods, relative to the 1961–1990 period, following the approach taken in the IPCC Third Assessment Report. Hayhoe et al. (2004) estimated that sea level along the California coast would rise 8.7 cm to 12.7 cm for the 2020–2049 period, and 19.2 cm to 40.9 cm for the 2070–2099 period, depending on the model and emission scenario used.
Mote et al. (2008) estimated future sea-level rise off Washington for 2050 and 2100, dividing the coastline into three regions according to their vertical land motions. They used global climate models to calculate the thermal expansion and cryosphere contributions to sea-level rise. The rates of global sea-level rise were then adjusted for vertical land motions and for model- predicted seasonal and interannual wind-driven increases in sea level. Mote et al. (2008) projected low, medium, and high sea-level rise for the Puget Sound region of 16 cm, 34 cm, and 128 cm, respectively, by 2100. They also found that some parts of the Olympic Peninsula could experience tectonic uplift that would exceed the low end of projected rates of global sea-level rise, with the medium estimate for sea-level fall between 0 cm and -15 cm by 2050, depending on location, and 0 cm and -30 cm by 2100.
Cayan et al. (2009) projected sea-level rise off California using Rahmstorf’s (2007) semi-empirical method with global average surface air temperature simulated from global models. Assuming that the rate of sea-level rise off the California coast will be the same as the global rate, Cayan et al. (2009) estimated sea-level rise of 30–45 cm by 2050, and 50–140 cm by 2100, relative to 2000.
Tebaldi et al. (2012) projected sea-level rise at 11 tide gage locations along California, Oregon, and Washington using the semi-empirical method of Vermeer and Rahmstorf (2009) to estimate global sea-level rise and 50 years (1959–2008) of tide gage records to estimate local rates and their deviations from global sea-level rise caused by local effects. Based on this information and output from an ensemble of GCM