Measuring sea level rise
Influences on sea levels
Many meteorological and oceanographic processes, with different time and spatial scales, affect sea levels around Australia. Influences on sea levels can be dynamic and vary from place to place. Many of these processes are considered cyclical – they oscillate back and forth. Although these processes do not have a permanent imprint on the long-term mean (or average) sea level, they can make it difficult to determine underlying trends in sea level.
The timeframes for major influences on sea levels are summarised in the table below (adapted from New Zealand National Institute of Water and Atmospheric Research, Goring & Bell 2001
).
| Driver | Sea-level response | Indicative timescales |
|---|---|---|
| Gravitational attraction of astronomical bodies | Tides | Hours, with some influences extending over multiple years |
| Wind | Set-up | Days |
| Changing atmospheric pressure | Inverted barometer | Days |
| Ocean currents | Variations | Weeks to months |
| Annual temperature cycle | Annual cycle | One year |
| El Niño/Southern Oscillation | Interannnual oscillations | Years |
| Interdecadal Pacific Oscillation | Interdecadal oscillations | Decades |
| Long-term global temperature changes | Long-term sea-level change | Multiple decades to centuries |
Some influences affect sea levels on short timescales (less than one year) including seasonal factors, weather systems and variability of ocean water properties such as temperature and salinity. The physical properties of the ocean water mass change with mixing and ocean currents, and can cause significant variations in sea level lasting from weeks to months.
Larger scale phenomena, like the El Niño Southern Oscillation (ENSO), the nodal tide and the Interdecadal Pacific Oscillation (IPO), can cause fluctuations in sea level spanning years to decades and are more difficult to isolate from sea level trends.
There is a distinct seasonal signature in most sea level records. For example, the Fort Denison (Sydney) record (which has been collected from 1886 until present) has a strong positive (high) bias in autumn and a negative (low) bias in spring, compared with long-term averages.
Longer term, larger scale meteorological influences such as ENSO can influence sea levels on cycles from three to eight years (National Tidal Centre 2010). The inter-annual variability in sea level around Australia correlates with ENSO (Amin 1993; Feng, Li & Meyers 2004; Church, White et al 2004; Church, Hunter et al 2006; Haigh, Eliot et al 2011).
The range of lunar influences on tides occurs over a cycle of about 18.6 years. Throughout this cycle, the moon induces a small amplitude harmonic influence on the position of mean sea level at a fixed location. The influence is at a maximum at the poles with no influence on the equilibrium nodal tide at latitudes of around 35°N and 35°S (Pugh 1987). The tide gauges in NSW are situated between 28°S and 38°S so the amplitude of the nodal tide for the NSW tidal records is therefore considered negligible (less than 4-5mm). Larger values than the theoretical maximum amplitude of the nodal tide have been measured in tide gauge records (Houston & Dean 2011).
At multidecadal time scales sea level also varies with the IPO, which operates over time frames of 20 to 30 years (National Tidal Centre 2010; Holbrook, Goodwin et al 2010). This long period oscillation is a lengthy interdecadal fluctuation in atmospheric pressure and sea surface temperature. When the IPO is low, cooler than average sea surface temperatures occur over the central North Pacific, and vice versa. During the 20th century the IPO exhibited three major phases: it was positive (southeastern tropical Pacific warm) from 1922 to 1946 and 1978 to 1998, and negative between 1947 and 1976. During positive phases of IPO more El Niño like conditions prevail and sea levels in Eastern Australia are lower than average. During negative phases more La Nina like conditions prevail and sea levels are higher than average.
Tide gauge data provides a measure of the water level “relative” to a fixed, land-based reference mark. However, the land upon which the tide gauge is positioned may be moving. Many processes can contribute to the land moving including:
- solid earth tides (relatively small movements of the Earth’s surface caused by the gravity of the moon and sun)
- plate tectonics (movements of the Earth’s crust)
- glacial isostatic adjustment (the rise of ground levels previously depressed by the weight of ice sheets during the last glacial period, also known as post-glacial rebound)
- localised factors including aquifer extraction, reclamation and development loadings.
To accurately estimate sea level change, these processes need to be taken into account. The land itself at a tide gauge might be moving up or down at long term rates comparable with the actual rate of sea level rise. This needs to be considered in isolating the real component attributable to sea level rise.
Sea level rise along the NSW coast can be more accurately determined since the installation in mid-2012 of Continuously Operating Reference Stations at Newcastle and at Fort Denison in Sydney Harbour. These reference stations were funded by OEH to measure any ground movement which could affect the sea level recordings. This will complement the similar sea level monitoring station at Port Kembla operated by the Bureau of Meteorology. These three stations will provide useful data on sea level trends in the medium-long term.
Challenges of isolating long term trends from sea level records
It is difficult to isolate the smaller signal of long-term changes in sea levels from these larger dynamic influences which operate over different time and spatial scales (Zhang & Church 2012). Sea level records which are sufficiently long need to be used so that the more dynamic and cyclical influences do not obscure the long term trend.
Large inter-annual and inter-decadal sea level fluctuations associated with climate variability (such as ENSO and IPO) are difficult to distinguish from underlying long-term trends of sea level rise in shorter data sets. Long record lengths spanning ENSO and IPO cycles are recommended to help distinguish between trends and long-period relative sea level fluctuations in individual records (Douglas, Kearney & Leatherman 2001; You, Lord & Watson 2009) and even then great care needs to taken in the interpretation of measured trends.
Analysis (such as filtering and averaging) can be applied to tidal records of sufficient length to smooth dynamic sea level influences and reveal low amplitude sea level rise.
Measuring sea level in NSW
OEH supports a system of tide gauges which measures ocean water levels continuously at recording sites along the NSW coast, and on Lord Howe and Norfolk Islands. This tide gauge system is managed and maintained by Manly Hydraulics Laboratory. Data from these sites have been presented in a range of reports, including the NSW Ocean and River Entrance Tidal Levels Annual Summary 2011-2012 report.
Analysing historic sea level data
OEH is working with collaborators from CSIRO, University of Tasmania, University of Western Australia, Australian National University and Southampton University (UK) to analyse Australian recorded historical sea levels and determine trends, regional variability and influencing factors.
The team are using publicly available tide-gauge records and satellite-altimeter data around Australia. They will also assess the quality of long term Australian sea level data, review relevant Australian sea level research, investigate spatial and temporal variability of sea level data, and analyse long-term sea level trends around Australia. The results will be interpreted in the context of global sea-level change and will be published in a refereed scientific journal.
Past trends provide valuable evidence in preparing for future environmental change but, by themselves, are insufficient for assessing the risks associated with an uncertain future (Parris, Bromirski et al 2012).