View all publications

4. Tides, storm surge and the effects of sea level rise

Key points:

  • There is no clear evidence to indicate whether there will be either an increase or decrease in storm surge magnitudes in the next 50 years. The highest storm surge experienced in the past 25 years appears to be about 0.55 m.
  • It is expected that storm tide elevations will rise at the same rate as mean sea level rise. It is recommended to use a sea-level rise of 0.2 m by 2050 and 0.5 m by 2100 for planning purposes.

4.1 Introduction

Coastal hazards, such as inundation and coastal erosion, tend to be caused by a range of inter-relating factors or "drivers" which can be both natural and caused or exacerbated by human actions. Besides earthquakes and underwater landslides (which can cause a tsunami or coastal subsidence) and ocean tides, the main natural causes of coastal hazards arise from extremes in weather such as storms and cycles in ocean-atmosphere response (sea level and currents). It is these weather and climate-related causes that will be altered most by climate change arising from global warming, mostly exacerbating the potential problems for the coast e.g., heightened storm tides, stronger winds and waves, sea-level rise.

This section reviews existing information on tides and sea-level in the Chatham Islands and discusses how climate variability and change may impact on sea levels. It does not consider the wave climate (which is an important factor influencing both coastal inundation and erosion), nor does it provide any significant information on the consequences of inundation or erosion, or how this may change in the future, within the Chatham Islands.

4.2 Tides

Tides are an important factor in determining the potential impact of coastal hazards on the coastline of the Chatham Islands. It is the tide height that governs the likelihood of coastal inundation from storm surge (see next section), with the effect of wave conditions experienced at the coastline highly dependent on water level.

Sea level data for the Chatham Islands is limited. Despite there being three sea level recorders currently installed (2 at Waitangi and the NIWA gauge at Kaingaroa) the length of the data record is relatively short. Tide information for the Chatham Islands can be located at:

The tide range in the Chatham Island is modest, being microtidal with a Spring tide range of around 0.7 m and a Neap tide range of approximately 0.6 m. Based on the tidal constituent data, derived from the NIWA gauge at Kaingaroa, Figure 4.1 shows the probability of exceedence of predicted high water above mean level of the sea. This suggests that the "pragmatical" mean high spring tide (the tide height that is exceeded by about 10-12% of all high tides) is around 0.61 m relative to MLOS [Mean Level Of the Sea, for this period of record. Note thatMean Sea Level(MSL) is a surveyed height datum point, generally measured sometime between about 1930 and 1950 (depending on location). Because sea levels have been rising, it differs from present day mean level of the sea by a few cm. See Glossary of Coastal Guidance Manual for further definitions.] and the Highest Astronomical Tide (HAT) [HAT is the highest possible tide that can occur (excluding meteorological effects).] is 0.74 m relative to MLOS. Figure 4.2 shows the probability of exceedance of high tide levels above MLOS (metres) for the Kaingaroa tide gauge predicted for the 100 year period (2000 to 2099) assuming no sea-level rise. Table 4.1 summarises the high tide levels at key probability values from the analysis of Figure 4.2.

Figure 4.1 Probability of exceedance of predictable high tides based on tide data from the Kaingaroa tide gauge

Thumbnail of image. See figure at its full size (including text description).

Table 4.1 Frequency or probability of exceedance of high tide levels at Kaingaroa over the 100 years from 2000 to 2099, focusing on the higher tidal elevations.

Note: Levels are not tied in to any MSL or survey datum.

% of high tides that will be above the specified level

High tide elevation (m above MLOS)









Figure 4.2 Probability of exceedance of predictable high tides over the next 100 years (2000-2099), excluding the effects of sea-level rise, based on tide data from the Kaingaroa tide-gauge

Thumbnail of image. See figure at its full size (including text description).

4.3 Storm surge and storm tide

Storm surge is the temporary elevation in sea level at the shoreline above the predicted tide height caused by a combination of both low atmospheric pressure (inverted barometer) and set-up by adverse winds, and breaking surf waves during a storm, Figure 4.3. An additional component that needs to be considered for each site is wave run-up, which varies locally with the topography of the coastal margin.

Storm tide is the level of the predicted tide plus storm surge. This demonstrates the important point that high tides play a critical role in determining the elevation of coastal sea levels during a storm.

Figure 4.3 Schematic showing components of a storm tide

Thumbnail of image. See figure at its full size (including text description).

Data from the Kaingaroa tide-gauge were analysed and the storm surge component separated from the predictable part of the tide. Figure 4.4 shows a summary of the storm surge over the three year period that data have been collected at the gauge. Over this period the largest storm surge has been about 0.25 m, with a storm surge of over 0.20 m occurring 15 times over this period.

Estimates of annual exceedance probability (AEP) for storm surge magnitude were derived from the 3 years of storm surge data for Kaingaroa (Figure 4.4) using the r-Largest method. The r-Largest method is based on the idea of using a fixed number, r, of independent extreme values from each year to provide an estimate of the parameters of the extreme value distribution. Practically, the method involves 2 steps: firstly, identification of extreme events; and secondly the selection of a suitable number of independent events from each year of data. The number of events has to be large enough to ensure sufficient data are available to obtain reasonable parameter estimates, but also small enough that the lowest level used still belongs to the extreme tail of the distribution. We used 5 events per year and a Gumbel (EV1) fit to the empirical storm surge distribution. Using this approach, the following storm surge levels were derived:

5 year return period = 0.37 m, and 10 year return period = 0.39 m.

The sea level records for the Chatham Islands are too short to estimate longer return periods of storm surge on the open coast. However, there are longer records (up to 25 years) available of atmospheric pressure and to a lesser extent, winds, which can be used as surrogates for estimating storm surge. Of interest are minimum atmospheric pressures as most storm-surge events have common elements to varying degrees - low atmospheric pressure, high tides, adverse winds and high seas. Typically a 1 hectopascal drop in atmospheric pressure results in about a 1 cm rise in sea level. However, Stanton (1997) observed that in the Chatham Islands, this relationship was less, being around 0.75 cm per hectopascal.

Figure 4.4 Storm surge magnitude over the three year period of sea level recordings from the Kaingaroa tide gauge

Thumbnail of image. See figure at its full size (including text description).

Figure 4.5 shows minimum daily barometric pressure (hPa) and maximum daily storm surge from the Kaingaroa tide gauge from May 2002 to May 2005. This shows an approximate relationship of a 0.60 cm increase in storm surge for every 1 hectopascal reduction in barometric pressure although, as would be expected, there is considerable scatter in the data (i.e. the residual after allowing for inverted barometer, IB, appears to be approximately ±100 mm) since the relationship does not consider wave setup. The r2 of the relationship is 0.67, which indicates that inverted barometer explains 67% of the variance in the maximum daily storm surge. Therefore, an equation to estimate storm surge for the Chatham Islands at Kaingaroa (using barometric pressure) is:

Maximum daily storm surge = -6.04 x Barometric pressure + 6110 (1)

where maximum daily storm surge is in mm, and barometric pressure is in hPa.

Figure 4.5 Minimum daily barometric pressure (hPa) versus maximum daily storm surge (mm) for the Kaingaroa tide-gauge from May 2002 to May 2005

Thumbnail of image. See figure at its full size (including text description).


Site Number

Start of record

Data interval

Minimum bar. pressure (hPa)



1 January 1970


968.3 (28/11/71)

Chatham Island AWS


27 October 1991

3 hourly/hourly

969.0 (17/7/95)

Chatham Island EWS


13 October 1999


977.3 (22/2/04)

Kaingaroa tide gauge


30 April 2002

5 minute

971.6 (22/2/04)

Barometric pressure has also been recorded on the Chatham Islands at the other sites summarised in Table 4.2. For the period of overlap (for all 4 barometric pressure sites), the minimum recorded barometric pressure occurred on 21/22 February 2004 (Figure 4.6, top). From Figure 4.6 (top) it can be seen that the Waitangi and Chatham Island EWS sites do not record as low values for minimum pressure over the February 2004 period. In February 2004 the Waitangi record is no longer recording hourly data for every hour of the day. In the case of the Chatham Island EWS record, however, a 6hPa systematic bias is apparent over the February 2004 period, which may reflect a calibration problem with the pressure transducer.

Figure 4.6 Minimum daily barometric pressure (hPa) for February 2004 (top), and mid November to mid December 1971 (bottom)

Thumbnail of image. See figure at its full size (including text description).

Overall, the lowest recorded barometric pressure is 968.3 hPa (Figure 4.6, bottom), recorded on 28 November 1971 at the Waitangi site (Site 6176). Using Equation 1, this gives a storm surge of approximately 0.26 m due to inverted barometer.

An approximate rule of thumb is that inverted barometer contributes half the set-up in ocean storm surge (above the predicted tide), while the other half comes from wind set-up and other coastal trapped-waves that propagate out from the storm centre (Bell et al., 2000). This rule is only approximate, as the two contributory processes can vary considerably. However, based on the above analysis this suggests that the maximum storm surge experienced over the last 25 years, has been in the order of 0.55 m.

At present there is an insufficient data record to assess the full likelihood of occurrence for a given storm-surge height. Experience from elsewhere in New Zealand suggests that, during the last century the highest storm surge (excluding wave effects) is generally around 1 m above the predicted tide, (Bell et al., 2000), and is limited by the depth of depressions experienced.

4.4 Sea level fluctuations

"Sea-level fluctuations" refers to the fluctuations in the mean level of the sea, after taking out the influence of tides and without the influence of long-term sea-level rise. The main long-term fluctuations (excluding storm-driven fluctuations which are discussed above) are:

  • Annual seasonal heating and cooling cycle by the sun on the ocean surface);
  • interannual (2 to 4 year El Niño-Southern Oscillation [Cycle of alternate El Niño and La Niña episodes that govern climate and sea-level variations around the Pacific and Indian Oceans-commonly called the El Niño-Southern Oscillation or ENSO system.] cycles); and
  • interdecadal (20 to 30 year Interdecadal Pacific Oscillation [Longer "El-Niño-like" 20-30 year cycles of alternate positive and negative phases that effect the wider Pacific Ocean region, abbreviated as IPO. Since 1998 the IPO has been negative.] or IPO cycles).

There is little information on sea level fluctuations in the vicinity of the Chatham Islands but in areas of the North Island seasonal fluctuations are relatively small, averaging around a range of 0.08 m over a year (but can range up to 0.16 m in some years). The highest sea level tends to occur in late summer or early autumn (January to April), when the thermal expansion due to warmer seawater is greatest. Sea level and sea surface temperature variability was assessed by Stanton (1997) but could not verify the hypothesis that there is a direct link between sea surface temperature and sea level associated with the movement of the Subtropical Convergence to the south and north of the Chatham Islands.

The El Niño-Southern Oscillation (ENSO) is a quasi-periodic climate system on cycles of 2 to 4 years. During El Niño episodes (negative SOI), the mean level of the sea is depressed below normal levels typically by up to 0.12 m around New Zealand. The converse is true for strong La Niña episodes, where sea levels are higher than normal by a similar amount. The second climatic feature known to affect long-period sea-level fluctuation is the Interdecadal Pacific Oscillation (IPO) which operates on a 20 to 30 year cycle. The IPO tends to modulate interannual ENSO climate variability over the region, with increased frequency of El Niño events occurring during positive phases of the cycle.

The 20 to 30 year positive phase of IPO cycle probably ended around 1998. An extended period of negative phase IPO would bring more balance between El Niño and La Niña episodes, and has exhibited a quicker rate of sea-level rise than that experienced over the previous positive phase of IPO from 1976 to 1998. This pattern of a quicker sea-level rise during negative phases of the IPO has been demonstrated from the Port of Auckland tide-gauge record. It is possible that a similar trend is occurring regionally, in which case the next 20 to 30 years could see a faster rise in sea level than the mean long-term trend of 1.6 mm/yr.

Based on sea level fluctuation information for the North Island, Figure 4.7 combines all three long-term sea-level fluctuations (annual, ENSO, IPO), implying that the mean level of the sea could vary by up to ±0.25 m above the average mean level of the sea.

Figure 4.7 Summary of the relative magnitudes of long period sea level fluctuations (excluding storm-driven fluctuations and global warming effects)

Thumbnail of image. See figure at its full size (including text description).

4.5 Sea level rise

Sea level is rising around New Zealand, starting around the early to mid part of the 1800's. For New Zealand, the historic rate of rise has been around 1.6 mm/yr or approximately 0.16 m over the past 100 years up to 2000 based on Hannah's (2004) analysis of tide-gauge data from the four main ports. This value also lies mid-way in the range of estimated global sea-level rise of between 1 and 2.5 mm/yr since the early 1800's.

There is no sign yet of any definitive acceleration in the rise of sea level from any New Zealand sea-level gauges. However, the Third Assessment Report of the IPCC (2001) is predicting a slowly increasing acceleration over the next 50 years and beyond (Figure 4.8). The most likely mid-range rates of global mean sea-level rise are between 0.14 and 0.18 m by 2050 and between 0.3 and 0.5 m by 2100, with an upper-limit projection of 0.88 m by 2100. It should also be noted from Figure 4.8 that the acceleration of sea level rise will not be discernable for another 20 to 30 years.

Figure 4.8 IPCC (2001) global mean sea level rise projections (tied back to 1990)

Thumbnail of image. See figure at its full size (including text description).

When it comes to the risks associated with sea-level rise, the quantity of primary interest is the sea-level rise relative to the landmass that the tide gauge sits on. In particular, if the landmass is subsiding, then the relative sea-level rise is higher than the absolute sea-level rise. There is little information on how tectonically stable the Chatham Island land mass is. Also, little is known yet about the regional differences in the rise in ocean levels around the SW Pacific, compared to the global average rates given in IPCC (2001). Taken together, this means that relative sea-level rise within the Chatham Islands can be treated to be similar to the global average rate of rise. So until such time as further information to the contrary becomes available, the projected IPCC (2001) sea-level rise should be used for coastal hazard planning as recommended by the recent Coastal Guidance Manual (MfE, 2004b):

  • 0.2 m by 2050 (relative to 1990)
  • 0.5 m by 2100 (relative to 1990)

Figure 4.9 shows the probability of exceedance (same as Figure 4.2 above) of high tide levels above MLOS (metres) predicted for the 100 year period (2000 to 2099) assuming (1) no sea-level rise (heavy black line); 0.2 m rise in mean sea level by 2050 (solid blue line), and a 0.5 m rise in sea level by 2100 (dashed line).

Figure 4.9 Probability of exceedance of high tide levels above MLOS (metres) at Kaingaroa predicted for the 100 year period (2000 to 2099) assuming (1) no sea-level rise (heavy black line), 0.2 m rise in mean sea level by 2050 (solid blue line), and a 0.5 m rise in sea level by 2100 (light dashed line)

Thumbnail of image. See figure at its full size (including text description).

It is important to note that IPCC expects sea level will continue to rise for several centuries after 2100, even if greenhouse gas emissions are stabilised. This is due to the long lag times needed for the deep oceans to respond to ocean surface heating and the potential contribution from polar ice sheets, particularly from the Greenland ice sheet after 2100.

Ocean tides will not be directly affected by climate change but tidal ranges in shallow harbours and estuaries could be altered by deeper channels (following sea-level rise). Of perhaps greater concern than mean sea level rise, from the viewpoint of flooding, overtopping and in the design and assessment of coast protection works is any increase in the magnitude or frequency of storm-tide water levels. As discussed above, storm-tide levels depend mainly on the magnitude and frequency of storm surges as well as the timing with spring or neap tides. If deeper atmospheric low-pressure systems and stronger winds occur, then surge levels may increase in magnitude. There is no evidence (yet) that low atmospheric pressure systems will become more frequent or more intense with global warming. Furthermore, changes in the pattern of tracking of low-pressure systems and ex-tropical cyclones may also have an effect on extreme water levels due to the complex way that they interact with the continental shelf and coastline.

In summary there is no clear evidence as yet to indicate whether there will be either an increase or decrease in storm surge magnitudes in the next 50 years, and hence how storm tide levels will change. Due to the lack of such information, it is normally assumed that storm tide elevations will rise at the same rate as mean sea level rise.

Coastal hazards are not only dependent on the 'hazard drivers' but also on the geomorphology of the coast. How coastal hazards affect different coastal types, and how these different coastal types will respond to the impacts on climate change, is described in the guidance manual, Coastal Hazards & Climate Change: A Guidance Manual for Local Government in New Zealand, published by the Ministry for the Environment (MfE, 2004b).