Gold Coast City Council Storm Tide Study Final Report Addendum

February 2013

Contents

Executive Summary viii

1. Introduction 1 1.1 Background 1 1.2 Aims and Objectives 1 1.3 Definitions 2 1.4 Study Area vs Modelling Domain 5 1.5 Scope and Limitations 5

2. Methodology Overview 8 2.1 Tropical Cyclone Storm Tide Risks 8 2.2 Extra-Tropical and Remote Tropical Cyclone Storm Tide Risks 11

3. Project Data 13 3.1 Bathymetry and Coastline 13 3.2 Hydrographic Data 13 3.3 Tidal Constituents, Predictions and Observations 14 3.4 Wave Climate 16

4. Regional Meteorology 17 4.1 Tropical Cyclone Climatology 17 4.2 Extra-Tropical Climatology 25

5. Numerical Model Development 29 5.1 Tropical Cyclone Wind and Pressure Model 29 5.2 Hydrodynamic Model 29 5.3 Spectral Wave Model 35 5.4 Establishment of the Parametric Tropical Cyclone Models 36 5.5 Simulation Modelling of Tropical Cyclone Impacts 38 5.6 Modelling of Extra-Tropical and Remote Tropical Cyclone Impacts 39

6. Tropical Cyclone Model Calibration and Verification 45 6.1 Deterministic Verification of the Tropical Cyclone Models 45 6.2 Statistical Verification of the Tropical Cyclone Simulation Model 45

7. Results for Present Climate 48

41/22526/412243 Storm Tide Study i Final Report Addendum

7.1 Tropical Cyclone Impacts 48 7.2 Extra-Tropical and Remote Tropical Cyclone Impacts 57 7.3 Combined Climate Storm Tide Impacts 57 7.4 Design Water Level Hydrographs 62

8. Possible Impacts of Climate Change 63 8.1 The Enhanced Greenhouse Effect 63 8.2 Relative Results of the Climate Change Scenarios 66 8.3 Tide plus Surge Climate Change Water Levels 75

9. Conclusions 77

10. References 78

Storm Tide Study 41/22526/412243 Final Report Addendum

Table Index Table 1 Summary of Present Climate Storm Tide Estimates viii Table 2 Summary of Present and Future Climate “Tide plus Surge” Ocean Levels for the Gold Coast Seaway ix Table 3 Monitoring Stations for Tidal Data Collection Operated by Council during the period November 2004 to June 200513 Table 4 Australian tropical cyclone category scale 18 Table 5 Key Statistical TC Climatology Parameters for the Gold Coast Region 24 Table 6 Tidal Calibration - Quantitative Analysis of Model Performance 35 Table 7 Base Storm Parameter Set 36 Table 8 Additional Parameter Sensitivity Testing 37 Table 9 Tidal Planes at Gold Coast Seaway 38 Table 10 Summary of Tropical Cyclone Total Storm Tide Levels for Present Climate 49 Table 11 Tide gauge references for the extra-tropical simulations 57 Table 12 Summary of combined Tropical Cyclone and Extra-Tropical Total Storm Tide Levels for Present Climate 59 Table 13 Enhanced Greenhouse Scenarios for Tropical Cyclones65 Table 14 Estimated Increase in combined Tropical Cyclone and Extra-Tropical Total Storm Tide Levels under Climate Change Scenario in the Year 2060 67 Table 15 Estimated Increase in combined Tropical Cyclone and Extra-Tropical Total Storm Tide Levels under Climate Change Scenario in the Year 2100 71 Table 16 Summary of Present and Future Climate Tide plus Surge Ocean Levels for the Gold Coast Seaway 75 Table 17 Summary of Present and Future Climate Tide plus Surge Ocean Levels for the Coomera River South 76 Table 18 Summary of Present and Future Climate Tide plus Surge Ocean Levels for the Logan River Mouth 76

41/22526/412243 Storm Tide Study iii Final Report Addendum

Figure Index Figure 1-1: Water level components of an extreme storm tide 3 Figure 1-2: Study Area 6 Figure 1-3: Modelling Domain 7 Figure 2-1: Overview of the Tropical Cyclone Methodology 10 Figure 2-2: Overview of the Extra-tropical and Remote Tropical Cyclone Methodology 12 Figure 3-1: Location of the Tweed Offshore Tidal Gauge 15 Figure 4-1: Severe tropical cyclone Hamish at Category 4 intensity paralleling the Queensland coast offshore Fraser Island in March 2009. (US Navy processed image) 17 Figure 4-2: Time history of the frequency of cyclone occurrence within 500 km of the Gold Coast 21 Figure 4-3: Time history of cyclone peak intensity within 500 km of the Gold Coast 21 Figure 4-4: Cyclone tracks capable of affecting the Gold Coast Region classified into top: offshore (21.7%), mid: parallel (47.8%) and bottom: onshore (30.4%). 22 Figure 4-5: Extreme value analysis of cyclone intensity within 500 km of Gold Coast 23 Figure 4-6: Example of an historical East Coast Low and some of the many ECL storm tracks affecting the Gold Coast region (after Harper 2001c) 26 Figure 4-7: Composite Datasets of East Coast Lows Affecting SE Queensland 28 Figure 5-1: Bathymetric Data Used in the Generation of the RHD Model 31 Figure 5-2: One-dimensional Representation of the Coomera River system in the RHD model 32 Figure 5-3: The Gold Coast Seaway tidal residual record. 40 Figure 5-4: The Seven Highest Residual Water Level Events from the Seaway Tidal Record. 41 Figure 5-5: The Variability in Residual Levels for TC Roger Across the Available Tide Gauges. 42 Figure 5-6: Overview of the Tidal Residual Re-sampling Process. 43 Figure 5-7: Sensitivity of the Seaway simulation return periods to the single large TC Roger event. 43 Figure 6-1: Verification of the generated HAT tidal plane at Gold Coast Seaway 46

Storm Tide Study 41/22526/412243 Final Report Addendum

Figure 6-2: Statistical simulation model prediction of regional wind speeds 47 Figure 7-1: Estimated tropical cyclone total storm tide levels for selected sites. 52 Figure 7-2: Estimated tropical cyclone storm tide components for selected sites. 53 Figure 7-3: Estimated tropical cyclone storm tide components for selected sites. 54 Figure 7-4: Estimated tropical cyclone storm components for selected sites. 55 Figure 7-5: Estimated tropical cyclone storm tide components for selected sites. 56 Figure 7-6: Combining tropical and extra-tropical tide plus surge levels for present climate. 58 Figure 8-1: Projection of Global Average Sea Level Rise (after NCCOE 2004) 64 Figure 8-2: Assumed Possible Changes in the Intensity of Tropical Cyclones under Future Climate Change Projections within 500 km of the Gold Coast. 66

41/22526/412243 Storm Tide Study v Final Report Addendum

Appendices A Technical Note on the Interpretation of Statistical Return Periods B Tropical Cyclone Dataset Summary C Calibration of the RHD Model D Calibration to Historical Tropical Cyclone Events E Location of Named Model Sites

Storm Tide Study 41/22526/412243 Final Report Addendum

41/22526/412243 Storm Tide Study vii Final Report Addendum

Executive Summary

With its significant coastal margins, waterways and floodplains the populous Gold Coast region has a significant exposure to the impacts of ocean-related hazards (wind, waves, storm surge, beach erosion and associated rainfall and flooding). Under currently projected future climate scenarios it is likely that slowly rising sea levels and the possibility of increases in the intensity of tropical cyclones will increase this exposure. Accordingly Gold Coast City Council (GCCC) has sought to quantify these threats to enable responsible and sustainable planning decisions.

Previous studies of storm tide risk at the Gold Coast have suggested that tropical cyclones do not presently play a major role due to the combined effects of the narrow continental shelf and the typically decreasing frequency and intensity of tropical cyclones south of Hervey Bay. Instead, other less intense but large scale and more frequent weather systems such as East Coast Lows and even remote tropical cyclones have been thought to dominate the storm tide risk. Through the application of a number of numerical, statistical and data analysis techniques, this study has confirmed these earlier assessments and increased confidence in the water level estimates that are summarised in Table 11, where tropical cyclone influences (bold italics) can be seen to be limited to longer ARI values relative to the other weather systems.

Table 1 Summary of Present Climate Storm Tide Estimates

Estuarine Sites Open Coast

Gold Coast Coomera Logan River Surfers Paradise Seaway River South Mouth

Tide plus Plus Wave ARI Tide plus Surge Only Surge Only Setup

y m AHD m AHD m AHD m AHD m AHD

2 1.16 0.96 1.60 1.16 1.75

5 1.25 1.07 1.69 1.25 2.00

20 1.37 1.22 1.80 1.33 2.30

50 1.47 1.33 1.88 1.38 2.42

100 1.54 1.41 1.94 1.41 2.49

200 1.62 1.50 2.00 1.44 2.55

500 1.69 1.58 2.10 1.48 2.61

2000 1.78 1.68 2.36 1.52 2.71

10000 1.87 1.78 2.78 1.60 2.80

1 This report addendum updates previously determined GHD (2011) storm tide estimates for the Logan River Mouth site and minor changes to some other areas as a result of revised tidal plane estimates.

Storm Tide Study 41/22526/412243 Final Report Addendum

Table 1 presents a selection of results from the analyses that are described in more detail later in the report. The “Tide plus Surge Only” estimates are the ocean levels expected either in the protected estuarine waterways of The Broadwater and adjacent areas, or immediately offshore of the open coast beaches before wave breaking occurs. The “Plus Wave Setup” column shows the effect of the additional ocean level likely to be present at the beach face due to a process associated with wave breaking. The Surfers Paradise site in this case is deemed representative of all open coast beaches in the Gold Coast region, ignoring possible localised wave sheltering in some areas. It is emphasised that the provided Tide plus Surge estimates are considered inherently more reliable than those that include wave setup. This is due to the likely wave interaction with very localised small scale dynamic coastal features. In regard to possible future climate conditions, the study has considered State Government recommendations in regard to expected Sea Level Rise (SLR) and the international consensus in respect of possible tropical cyclone intensity and frequency changes. Although there is no specific advice available on the future climate behaviour of other large scale weather systems, the study has made a nominal allowance that their influences might increase surge magnitudes by 10% by the year 2100. The results of these assumptions are summarized in Table 2 for the Gold Coast Seaway, where the 2060 year SLR allowance is 0.4 m and the 2100 SLR allowance is 0.8 m. It can be seen that the increases in water level attributable to non-SLR effects are estimated to be relatively small, amounting to a maximum of +0.15 m by 2100 at the 10,000 y ARI.

Table 2 Summary of Present and Future Climate “Tide plus Surge” Ocean Levels for the Gold Coast Seaway

Climate Planning Year

ARI 2010 2060 2100

y m AHD m AHD m AHD

2 1.16 1.64 2.04

5 1.25 1.70 2.10

20 1.37 1.82 2.23

50 1.47 1.93 2.33

100 1.54 2.01 2.41

200 1.62 2.10 2.50

500 1.69 2.19 2.58

2000 1.78 2.30 2.68

10000 1.87 2.40 2.80

41/22526/412243 Storm Tide Study ix Final Report Addendum

The adopted methodology for deterministic storm surge and associated wave modelling was built around Council’s existing numerical hydrodynamic modelling capabilities and experience, as well as its extensive and valuable measured tidal database. A large-domain unstructured mesh finite-volume long-wave hydrodynamic model referred to as the Regional HydroDynamic or RHD model was supplied by Council for this purpose. The RHD model is an implementation of the Mike 21 model and was originally developed by DHI Water And Environment Pty Ltd for Council. It is one of several models pertaining to the Gold Coast storm tide decision support system developed in 2008. The RHD model was refined by GHD and its capabilities enhanced to meet key recommendations presented in the Government-sponsored Queensland Climate Change investigations (e.g. Harper 2001, 2004). This included adding a spectral wind-wave model to calculate the growth, decay and transformation of wind-generated waves in the Gold Coast region.

Additional requirements formulated by Council were also accommodated, namely preserving the extent of the original RHD model and ensuring that storm tide influences can penetrate into the estuarine and waterway system of the Broadwater and be accounted for in the statistical aspects of the analysis. In response to the latter requirement, the RHD model was coupled to a network of 1-D estuarine links (representing major rivers and creeks in the Broadwater system). The enhanced modelling system was thus able to:

 reproduce the essential response of the waterway system to storm tide,

 ensure that estuary and waterway responses remain closely correlated event-wise with the open coast response, and

 achieve a reasonable level of calibration of the overall system, notwithstanding the complexity of the region. The stochastic modelling methodology follows SEA (2002) in respect of the use of parametric tropical cyclone modelling and introduces an empirical tidal residual resampling approach for extra-tropical storm tide simulation, based on the approach outlined in Hardy et al. (2004). This addendum to the GHD (2011) report has been prepared to update the previously supplied estimate of storm tide levels for the Logan River Mouth location, which were subsequently found to be affected by the use of modelled rather than measured tidal data in that region. As noted in GHD (2011), this site was located on a 1-D link that, although adjacent to the associated 2-D model, was not calibrated to the same level of accuracy as the 2-D model. Also, in re-issuing the report, the opportunity has been taken to provide more extensive regional reporting of the combined tropical cyclone and extra-tropical storm tide estimates, including wave setup. Additional clarification of some detailed aspects of the analyses has also been provided and there are other minor edits and adjustments that do not affect the original outcomes. The overall methodology remains identical to GHD (2011) in all other respects and the conclusions are unchanged. This Report is subject to, and must be read in conjunction with, the limitations set out in Section 1.5 and the various assumptions and qualifications contained throughout the Report.

Storm Tide Study 41/22526/412243 Final Report Addendum

1. Introduction

This study addresses the understanding, assessment and management of the risk posed by ocean storm tide to population, housing and infrastructure in the region under the jurisdiction of the Gold Coast City Council. It provides essential information that can be used to mitigate the effects of extreme storm tide through the planning process and also delivers design storm tide tail-water levels for fluvial flooding investigations. Both tropical cyclone and other large scale weather systems are considered in the context of present (nominally 2010) and projected future (2060 and 2100) climate and sea level conditions.

1.1 Background As noted by Council in its brief for this study, extreme weather events in the Gold Coast region are associated with tropical cyclones, east coast lows and mid-latitude low pressure systems, each of which exerts an impact in terms of storm surge, winds and waves. Of the above, tropical cyclones are often thought to be the more devastating, as they can typically generate extreme storm surge and associated flooding events. The Gold Coast has experienced more than 45 riverine floods since 1925 (Bureau of Meteorology web site) and passing, often remote, tropical cyclones were associated with many of these events. Historical records indicate that more than 40 tropical cyclones have influenced the Gold Coast region over the last 120 years. In the major flood of 1974, which was associated with the weak Tropical Cyclone Wanda, 1500 people were evacuated and in many places homes were swamped with 1.2 m to 1.5 m of water (Gold Coast Bulletin, Tuesday 29, 1974 p3). As a result of climate change, tropical cyclone impacts could increase in concert with global average air and sea temperatures. Therefore, the potential exists for increasing damage over time to both built and natural environments in the Gold Coast region.

1.2 Aims and Objectives The objective of the study is to quantify the likelihood of coastal and associated waterway areas within the Gold Coast City Council area being inundated by storm tide caused either by severe tropical cyclones or extra-tropical storm systems and influences. Present and future projected climates are considered. The tropical cyclone methodology closely follows the recommendations set out in the Government-sponsored Queensland Climate Change investigations (e.g. Harper 2001, 2004). In particular, the so-called “hybrid” modelling philosophy has been implemented, whereby a range of numerical, analytical and statistical models are constructed to provide a basis for the estimation of storm tide risks and the extrapolation of their impacts to very low probabilities (very high return periods or Average Recurrence Intervals (ARI)). The study analyses are necessarily based on “present climate” but, in conformance with the Scope of Work, commentary on making allowance for possible enhanced-greenhouse effects is included, leading to the identification of additional areas of concern. In regard to extra-tropical storm tide influences, the methodology is one of utilising long term tide gauge records, which are deemed to have captured the significant statistical water level

Storm Tide Study 1 Final Report Addendum

signal of these events. The tide gauge residual levels are then subjected to a re-sampling simulation that effectively extends the water level statistics to facilitate analyses.

With knowledge of the probability of specific water elevations being equalled or exceeded, long term planning can be adopted to mitigate against the more adverse impacts. Emergency response planning can also utilise this information to ensure adequate resources will be allocated to those areas most likely to be affected.

The study outcomes are provided in a number of forms, principally a series of graphs and tables showing storm tide elevation that corresponds to a specific Return Period risk or Average Recurrence Interval (ARI). These provide an essential input to long term planning whilst also allowing a relative ranking of risks for emergency response. The provided storm tide probability levels are for the 2, 5, 10, 20, 50, 100, 200, 500, 2,000 and 10,000 year ARI events.

1.3 Definitions The total seawater level experienced at a coastal, ocean or estuarine site during the passage of a severe large scale ocean storm (e.g. tropical cyclone or extra-tropical low) will be made up of relative contributions from a number of different effects, as depicted in Figure 1-1. The combined or total water level is then termed the storm tide, which is an absolute vertical level, referenced in this report to either Mean Sea Level (MSL) or Australian Height Datum (AHD) where applicable2.

1.3.1 Components of a Storm Tide It is important to understand the different water level components that can comprise the total storm tide at a specific site. These effects can vary throughout any given region in both time and space and depending on the local physical conditions. With reference to Figure 1-1 the following definitions apply:

(a) The Astronomical Tide This is the regular periodic variation in water levels due to the gravitational effects of the moon and sun, which can be predicted with generally very high accuracy at any point in time (past and present) if sufficient measurements are available. The highest expected tide level at any location is termed the Highest Astronomical Tide (HAT) and occurs once each 18.6 y period, although at some sites tide levels similar to HAT may occur several times per year. The tidal variation within the region has been estimated based on the results of numerical hydrodynamic modelling, Council’s measured tidal data and other published information.

2 Adjustments at specific estuarine sites from MSL to AHD have been based on the interpolation of published offsets available from Maritime Safety Queensland (MSQ 2012) and are only approximate.

Storm Tide Study 41/22526/412243 Final Report Addendum

Ocean Waves Extreme Winds MWL Wave Runup Wave Setup SWL

HAT Surge Currents Storm Tide

Expected MSL datum High Tide

Copyright © 2008 Systems Engineering Australia Pty Ltd

Figure 1-1: Water level components of an extreme storm tide

(b) Storm Surge This is the combined result of the severe atmospheric pressure gradients and wind shear stress of the storm acting on the underlying ocean. The storm surge is a long period “wave” capable of sustaining above-normal water levels over a number of hours or even days. The wave travels with and ahead of the storm and may be amplified as it progresses into shallow waters or is confined by coastal features. Typically the length of coastline that is severely affected by a tropical cyclone storm surge is of order 100 km either side of the track although some lesser influences may extend many hundreds of kilometres. The magnitude of the surge is affected by several factors such as storm intensity, size, speed and angle of approach to the coast and the coastal bathymetry. Extra-tropical storm systems such as East Coast Lows may have an extended (time and space) influence but normally at a magnitude lower than that from a severe tropical cyclone.

(c) Breaking Wave Setup Severe wind fields also create abnormally high sea conditions and extreme waves may propagate large distances from the centre of a storm as ocean swell. These waves experience little or no attenuation in deepwater regions and an offshore storm can impact several hundred kilometres of coastline. As the waves enter shallower waters they refract and steepen under the action of shoaling until their stored energy is dissipated by wave breaking either offshore or at a beach or reef. After breaking, a portion of the wave kinetic energy is converted into potential energy which, through the continuous action of many waves, is capable of sustaining shoreward water levels that are above the still-water level (SWL) further offshore. This increase in still- water level immediately after wave breaking occurs on a beach face is known as breaking wave setup and applies to most natural beaches and reefs. Wave setup is only associated with the rapid energy losses occurring during breaking and does not necessarily occur in river mouths,

Storm Tide Study 3 Final Report Addendum

swampy lands or areas that suffer inundation to the extent that waves do not immediately break but rather are degraded more gradually through frictional or diffractive effects. Also, the presence of deep channels behind wave shoals is expected to limit the propagation of wave setup by providing return flows and the like. Accordingly the statistical water level modelling here assumes that vertical wave setup contributions cease when the tide plus surge level exceeds a nominal open coast “dune crest” elevation. (d) Breaking Wave Runup

While much of the wave energy at the open coast prior to inundation occurring can be converted into wave setup, there remains some residual energy in the form of individual waves that will generate vertical runup and may cause localised intermittent impacts and erosion at elevations above that of the nominated storm tide level. These effects are best estimated with specific information about the land-sea interface, which may be changing in time as the storm tide increases in height. This includes the slope and porosity of the shoreline, vegetation and the incident wave height and period. In the present study the additional effects of wave runup on the open coast are approximated by empirical formulae.

(e) Still water level (SWL) and mean water level (MWL) The storm surge, mainly caused by the interaction of the extreme wind-driven currents and the coastline, raises coastal water levels above the normally expected tide over a large area, producing the so-called still-water level or SWL. This is the highest water level at a point on the shoreline if all short period wind wave action is smoothed out.

Meanwhile, the extreme-wind generated ocean waves, combinations of swell and local seas, are driven before the strong winds and ride upon the SWL. As part of the process of wave breaking, a portion of their kinetic energy (momentum) can then be transferred into potential energy as vertical wave setup, yielding a higher localised mean water level (MWL). As previously mentioned, this effect is not always active and not always effective as it depends upon local beach and dune geometry.

(f) Overland inundation and wave penetration When normally dry land becomes inundated during a severe storm tide episode, the sea begins to quickly flood inland as an intermittent “wave front”, driven by the initial momentum of the surge, products of wave setup and runup and the local surface wind stress. This flow then reacts to the local ground contours and the encountered hydraulic roughness due to either natural vegetation or housing and other infrastructure. It will continue inland until a dynamic balance is reached between the applied hydraulic gradients, wind stress and the land surface resistance or until it becomes constrained by elevation and creates ponding etc. As the storm surge abates or the tide reduces, a significant ebb flow can be created which is commonly responsible for much of the observed coastline scouring after such extreme inundation events.

(g) Specific effects not considered in this study The present study is focussed on the estimation of storm tide elevation and does not address the velocity of the encroaching storm tide flow. Further, as the new “stillwater” surface gradually reforms behind the propagating front, the exact extent to which individual unbroken or partially reformed ocean waves might further penetrate into a coastal region will be very site-specific. No over land wave modelling has been undertaken as part of this study.

Storm Tide Study 41/22526/412243 Final Report Addendum

There remain other related phenomena that are not addressed here but which can also have an effect on the local water level. These may include unsteady surf beat in specific high energy wave environments, and stormwater and/or river runoff3. It is recommended that suitably qualified practitioners consider these effects on a case by case basis when designing specifically exposed facilities.

1.3.2 Return Period Concepts The present study reports its findings in terms of statistical Return Periods (or Average Recurrence Intervals or ARI). It is important to understand that a Return Period is simply the expected average elapsed time in years between equalling or exceeding a specified event level. This concept does not guarantee that the nominated event’s return period number of years will have elapsed before such an event occurs again. In fact, the probability of experiencing the “n” year return period event within any consecutive period of “n” years is approximately 64%, i.e. more likely than not. For example, the 100 year and 1,000 year event could both occur in the same year or one might occur twice in the same year, etc. Appendix A provides more explicit advice on the choice of return periods in the context of encounter probability, which is a concept better suited to decision making leading to good planning.

1.4 Study Area vs Modelling Domain Figure 1-2 shows the study area along with a network of hydrographic stations where surface water elevation, flow discharge and current magnitude were measured by Council’s staff in the period 11 November 2001 to 01 June 2005. The data collected during this period has been used for the tidal calibration of RHD and some of these place names appear in the study summary tables. The area extends from Russell Island to the north to the mouth of the Tweed River to the South and covers the Broadwater system inclusive of the Logan, Pimpama, Coomera and Nerang Rivers and two creeks - Tallebudgera and Currumbin. The extent of the modelling domain shown in Figure 1-3 is significantly larger. It comprises approximately 2.27 million km2 of the Coral Sea offshore from the Queensland Coast with its eastern boundary reaching New Caledonia or approximately 164ºE. Along the coast, the modelling domain extends from 16ºS to 31ºS.

1.5 Scope and Limitations This document (the “Report”): 1. has been prepared by GHD Pty Ltd (“GHD”) for Gold Coast City Council (“Client”); 2. may only be used and relied on by the Client; 3. must not be copied to, used by, or relied on by any person other than the Client without the prior written consent of GHD; 4. may only be used for the purpose of addressing the associated Client Scope of Work. GHD and its servants, employees and officers otherwise expressly disclaim responsibility to any person other than the Client arising from or in connection with this Report. To the maximum extent permitted by law, all implied warranties and conditions in relation to the services provided by GHD and the Report are excluded unless they are expressly stated to apply in this Report.

3 Council is separately addressing the issue of riverine flooding using the ocean levels derived in this study.

Storm Tide Study 5 Final Report Addendum

The opinions, conclusions and any recommendations in this Report are based on assumptions made by GHD when undertaking the services and preparing the Report (“Assumptions”) and are detailed throughout the Report. GHD expressly disclaims responsibility for any error in, or omission from, this Report arising from or in connection with any of the Assumptions being incorrect.

Subject to the paragraphs in this section of the Report, the opinions, conclusions and any recommendations in this Report are based on conditions encountered and information reviewed at the time of preparation and may be relied on until circumstances indicate otherwise, after which time, GHD expressly disclaims responsibility for any error in, or omission from, this Report arising from or in connection with those opinions, conclusions and any recommendations.

Figure 1-2: Study Area

Storm Tide Study 41/22526/412243 Final Report Addendum

Figure 1-3: Modelling Domain

Storm Tide Study 7 Final Report Addendum

2. Methodology Overview

The Gold Coast region lies at the southern extremity of severe tropical cyclone influence and so is less susceptible to the types of potentially catastrophic storm tide inundation events that are possible in the northern regions of Queensland, especially the Gulf of Carpentaria (e.g. Harper 1999). Also, the region is protected against significant wind-setup by the presence of a narrow continental shelf with deep water relatively close to the coast. Accordingly, when the overall risk of tide plus surge effects from tropical cyclones has been previously assessed (e.g. Harper 2004b, CSIRO 2000) it would seem that the Gold Coast is not at high risk. However, the region is greatly exposed to extreme waves emanating from a vast area of the Coral Sea and the breaking wave setup (and runup) effects are considerably greater in this area than those behind the protection of the Great Barrier Reef. When wave setup is added to the tide plus surge effects, the risk to the open coast margins is much enhanced and so the study methodology seeks to emphasise the likely effects of extreme breaking wave setup.

Also, some recent studies that include this region (e.g. Harper 2001) have been limited in their scope to considering tropical cyclones only, whereas it is well known that lesser energetic but more frequent large scale storm events can have significant impacts. For example, sub-tropical cyclones (e.g. so-called TC Wanda of 1974) can generate a modest surge, yet high wave setup and often with accompanying rain and fluvial flooding. Also, continentally-linked East Coast Lows are a specific feature of this area, and produce persistent coastally-trapped wind fields that can also generate large waves. There are also influences from strong SE wind events generated by Tasman Sea high pressure systems that can create coastally trapped long waves and associated high seas. These extra-tropical systems typically produce elevated water levels over periods of many days and, although limited in magnitude, can dominate shorter term return period water level statistics because of their likely interaction with several tidal cycles.

2.1 Tropical Cyclone Storm Tide Risks Extreme storm tide levels caused by tropical cyclones cannot be estimated solely on the basis of historically measured water levels (Harper 2001). This is because the available record of tropical cyclones affecting any single location on the coast is quite short, the resulting storm surge response is often complex and very site specific, and the final storm tide is dependent on the relative phasing with the astronomical tide. Hence, measured storm tide data alone is typically inadequate for extrapolation to very low probabilities of occurrence. To overcome this problem, it is necessary to formulate a statistical model of the coastal region that will attempt to re-create the observed region-wide tropical cyclone climatology and numerically generate long sequences of potential storm tide scenarios. The statistical model must be supported by a series of deterministic hydrodynamic models that will describe the effect that an individual cyclone has on the coastal region, i.e. the relationship between the wind speed and atmospheric pressure patterns and the resulting storm surge and wave set-up for a given cyclone scenario. This is then combined with a tidal description of the region that recreates the known tidal characteristics. When the effect of a single cyclone can be adequately described, the statistical model is used to generate many thousands of possible situations and the resulting statistics are used to determine the probability of storm tide levels throughout the

Storm Tide Study 41/22526/412243 Final Report Addendum

study area. Importantly, the accuracy of the model predictions is checked against historical data where possible, or compared with long term measurements of wind speed at airports in the region, which are typically less subject to localised effects.

The methodology applied here closely follows the recommendations set out in the Government- sponsored Queensland Climate Change (QCC) investigations (e.g. Harper 2001, 2004). In particular, the so-called “hybrid” modelling philosophy has been implemented, whereby a range of numerical, analytical and statistical models are constructed to provide a basis for the estimation of storm tide risks and the extrapolation of their impacts to very low probabilities (very high return periods).

Figure 2-1 provides an overall conceptual view of the tropical cyclone methodology, which is based firstly on the availability of data to describe the tropical cyclone threat to the region, data to describe the coastal geography, historical storm tide data for calibration and for defining the regional tide characteristics. Data on regional winds is also used for model validation and finally, the coastal infrastructure assets must be identified. Chapter 3 discusses the study data in more detail.

A climatological risk assessment of the threat from tropical cyclones in the region is then undertaken to obtain statistical descriptions that can be extrapolated to return periods of interest. This includes statistics describing the expected variation in storm frequency, intensity, path and size within the region. Chapter 4 discusses the detailed climate analyses that have been required.

In parallel with the development of the climatology, numerical models that can estimate the impacts of tropical cyclones on the underlying ocean are established. A numerical hydrodynamic model is used to estimate the strength of the wind driven currents and resulting storm surge, while a spectral wave model is used to estimate wave heights and periods, which contribute the breaking wave set-up water level component and wave runup. The models are constructed based on regional bathymetry data, comprising flexible mesh numerical grids to resolve the near-shore islands, capes and bays. Details are given in Chapter 5.

The numerical storm surge and wave models are driven by a tropical cyclone wind and pressure field model that generates the complex winds representative of a moving tropical cyclone, according to a set of parameters supplied to it. For example, the set of parameters that approximate tropical cyclone Dinah, which impacted the region on 1967, was used as part of the verification of the storm surge and wave models in Chapter 6.

A much wider set of parameters was then used to simulate the effects of many hundreds of possible cyclones in the region. These parameters were chosen based on the identified range of values from the known long-term climatology of the region. When the results of simulating the wide range of possible cyclones is obtained, the resulting storm surge and wave heights are parameterised (simplified) into a form that is amenable to statistical modelling. This enables the otherwise very computationally intensive numerical surge and wave model results to be re- generated and interpolated very efficiently to enable a simulation of many thousands of years of possible cyclone events. The accuracy of this parameterisation is checked to ensure it is consistent with the other analysis assumptions.

Storm Tide Study 9 Final Report Addendum

Figure 2-1: Overview of the Tropical Cyclone Methodology

After the parametric surge and wave models are established and tested, the statistical model is built by combining them with the climatology description. At this point, the local astronomical tide is included and also the wave height and period is converted to breaking wave setup so that the overall height of the combined storm tide (tide + surge + setup) can be determined at any open coast location in the study area during the passage of a synthetic cyclone. The probability of water level exceedance can then be obtained by simulating an extended period of possible tropical cyclones affecting the region (50,000 years has been used) and accumulating the resulting time history of the tide, the surge and the wave setup at each coastal location. The added impact of intermittent breaking wave runup is also estimated in an analogous manner.

In this context, the model is not used to predict the future, but rather to estimate what the past experience up until this date might have been if 50,000 years of measurements had been available in the context of an unchanging underlying climate. A very long period is simulated simply to enable very low probabilities to be reliably estimated. For example, simulating 50,000

Storm Tide Study 41/22526/412243 Final Report Addendum

years provides 50 estimates of the 1,000 year return period water level and 5 estimates of the 10,000 year water level, upon which the average levels for those return periods will be based. While there will be a single highest value produced during the simulation it’s nominal return period of 50,000 years will have a very high variability associated with it. Accordingly, confidence in the accuracy of the prediction diminishes as the return period increases.

The statistical model is then verified by comparing its probability predictions against other data wherever possible. Clearly this is not possible in the case of the storm tide itself, but the tide statistics can be checked against their known probability of exceedance and also the predicted wind speeds (which are separately accumulated by the model) are compared with the available long-term regional wind records. Other checks are also done to ensure that the linear superposition of tide, surge and setup is a reasonable approximation to the real situation where there may be some interaction between these events. Chapter 6 details these checks.

Next in Chapter 7, the predicted exceedance of coastal water levels generated from the statistical modelling process for each point of interest is used to select the 20, 50, 100, 200, 500, 2,000 and 10,000 year ARI storm tide elevations of interest.

The possible effects of enhanced-greenhouse induced climate change are considered in a subsequent step, whereby possible future climate scenarios are simulated and those results compared with the estimates for “present climate” (refer Chapter 8).

2.2 Extra-Tropical and Remote Tropical Cyclone Storm Tide Risks The methodology so far described is targeted towards close-approach high energy tropical cyclone storm events that might be expected to dominate the storm tide risk well beyond the 100 y ARI. For shorter ARI events (e.g. 2, 5) there are many more frequent yet possibly locally benign weather events that, while difficult to parameterise within a stochastic model, will influence the predicted levels and produce water levels that, for example, will likely exceed Highest Astronomical Tide (HAT) at ARIs below 18.6 y. Indeed, many of these less energetic long wave events are triggered by very complex baroclinic ocean interactions of large scale and/or propagation distances, such that it would prove very difficult or impossible to numerically model them in any accurate or comprehensive manner. Likewise, remote tropical cyclones produce coastal impacts that are similar to extra-tropical storms.

Accordingly, the present study methodology addresses the effect of these important extra- tropical and remote tropical disturbances based on an analysis of measured tidal residuals. An empirical approach (refer Figure 2-2) has been adopted that stochastically re-samples tidal residual signals against a series of shifted tidal start dates in order to obtain the long period water level variations out to approximately 1,000 years ARI. While the Gold Coast Seaway record is not as long as some nearby sites like Brisbane and Mooloolaba, it is the most representative for the region and forms the basis of the analysis here for the Broadwater. Other gauges are used for the open coast and the southern Moreton Bay, and the details are given in Chapter 3.

The resulting statistics of the extra-tropical and remote tropical cyclone events are then probabilistically combined with the results obtained from the close approach tropical cyclone modelling to more accurately represent the shorter 2, 5, 20, 50, 100 years ARI water levels.

Storm Tide Study 11 Final Report Addendum

Figure 2-2: Overview of the Extra-tropical and Remote Tropical Cyclone Methodology

Storm Tide Study 41/22526/412243 Final Report Addendum

3. Project Data

Project data supplied by Council comprised a number of hydrodynamic models of varying resolution, their corresponding bathymetric datasets, wave records, multiple water surface elevation and instantaneous flow discharge records. These datasets were further complemented with tidal constituents, tidal predictions and tidal observations at tidal stations Brisbane Bar, Gold Coast Seaway and Offshore Tweed obtained from Maritime Safety Queensland (MSQ). The hydrodynamic models are described in Chapter 5 together with additional tidal data (information on tidal planes). Climate data is reviewed and analysed separately in Chapter 4.

3.1 Bathymetry and Coastline The bathymetry (geometry of the sea bed) and coastline information used in the models are as provided to GCCC by DHI Water And Environment Pty Ltd. It is understood that these have been originally extracted from C-MAP (Jeppesen Commercial Marine (Norway)) and further augmented with the so described “best available” (circa 2007) locally generated datasets.

It is noted that the quality of the data (bathymetry, water elevations, currents, winds, waves etc) determines the level of model validation that can be achieved.

3.2 Hydrographic Data

3.2.1 Water Surface Elevation Water surface elevation data in the Broadwater system was provided by Council at 29 monitoring stations. The stations are listed in Table 3 and shown in Figure 1-2. These are continuous records at 15-minute intervals of varying duration depending on location – two weeks at Brygon Creek and up to two months at the Logan River mouth.

Table 3 Monitoring Stations for Tidal Data Collection Operated by Council during the period November 2004 to June 2005

# Monitoring Station

1 B1 Russel Island West

2 B2 Russel Island East

3 B3 Logan River Mouth

4 B4 Jumpinpin

5 B5 Steiglitz Jetty

6 B6 Pimpama

7 B7 Coomera River Mouth (South)

8 B8 Runaway Bay (Allisee Harbour )

Storm Tide Study 13 Final Report Addendum

9 B9 Grand Hotel Labrador

10 B10 Southport Seaway

11 B11 Nerang Mouth

12 L1 Logan River Boat Hire Jetty

13 L2 Logan River 289 Rotary Park Road Pylon

14 L3 Logan River Boat Ramp (Pacific Highway)

15 L4 Logan River (Station Road Bethania)

16 L5 Albert River(Martens Street Yatala)

17 C1 Coomera River (North Arm)

18 C2 Marine Precinct (Coomera)

19 C3 Crab Farm (Coombabah Creek)

20 C4 Monterey Keys (Saltwater Creek)

21 C5 Coomera Shores

22 C6 Brygon Creek

23 N1 Evandale

24 N2 Campbel St Sorrento

25 N3 Royal Pines Harbour

26 N4 ARC Resort (Eady Ave)

27 N5 Sunshine Boulevarde

28 N6 Beechcomber Crt (Burleigh Waters)

29 N7 Weedons Crossing

3.2.2 Instantaneous Flow Discharge Records of instantaneous flow discharge (14 transects in total) were provided by Council at the Coomera mouth (four transects), Nerang River (one transect), Logan-Albert (one transect), Gold Coast Seaway (three transects), Broadwater (four transects) and Jumpinpin. The Broadwater transects were undertaken to the east and west of Sovereign Island, near Russel Island and at Victoria Point. Each record captured peak flood and ebb conditions.

3.3 Tidal Constituents, Predictions and Observations The data was provided by Maritime Safety Queensland for stations Brisbane Bar and Gold Coast Seaway and by Manly Hydraulics Laboratory (NSW Public Works) for the Offshore Tweed tidal station. The Tweed Offshore tidal gauge, operated since 1982, is positioned in 28 m of depth, 3.5 km offshore, between Tweed Heads and Fingal Head (Figure 3-1).

Storm Tide Study 41/22526/412243 Final Report Addendum

Figure 3-1: Location of the Tweed Offshore Tidal Gauge

Storm Tide Study 15 Final Report Addendum

3.4 Wave Climate Since 21/02/1987, The former Beach Protection Authority, now part of the Queensland Department of Environment and Resource Management (DERM) and Council operate jointly a directional Datawell waverider buoy system located offshore from the Gold Coast seaway (latitude 27° 57.930'S and longitude 153° 26.530'E). The wave data records made available for the analysis included half-hourly time series of significant wave height Hs(m), maximum wave height Hmax(m), zero-crossing wave period Tz(s) and peak wave period Tp(s). The last record on file was 28/02/2010.

Longer term wave data from the so-called Brisbane waverider buoy located offshore Cape Lookout to the north of the Gold Coast has also been used to estimate regional wave characteristics during extra-tropical events.

Storm Tide Study 41/22526/412243 Final Report Addendum

4. Regional Meteorology

4.1 Tropical Cyclone Climatology

4.1.1 Tropical Cyclone Characteristics The tropical cyclone is a large scale and potentially very intense tropical low pressure weather system that affects the Queensland region typically between November and April (Harper 2001a). In Australia, such systems are upgraded to severe tropical cyclone status (referred to as hurricanes or typhoons in some countries) when average, or sustained, surface wind speeds exceed 120 km h-1. The accompanying shorter-period destructive wind gusts are often 50 per cent higher than the sustained winds. In the southern hemisphere, tropical cyclone winds circulate clockwise around the centre, as seen in the spiral cloud patterns of the satellite image in Figure 4-1 for severe tropical cyclone Hamish in 2009.

Figure 4-1: Severe tropical cyclone Hamish at Category 4 intensity paralleling the Queensland coast offshore Fraser Island in March 2009. (US Navy processed image)

Storm Tide Study 17 Final Report Addendum

There are three components of a tropical cyclone that combine to make up the total cyclone hazard - strong winds, intense rainfall and induced ocean effects, including extreme waves, currents, storm surge and resulting storm tide. The destructive force of cyclones is usually expressed in terms of the strongest wind gusts likely to be experienced. Maximum wind gust is related to the central pressure and structure of the system, whilst extreme waves and storm surge, are linked more closely to the combination of the mean surface winds, central pressure and regional bathymetry.

The Commonwealth Bureau of Meteorology (BoM) uses the five-category system shown in Table 4 for classifying tropical cyclone intensity in Australia. “Severe” cyclones are those of Category 3 and above.

Table 4 Australian tropical cyclone category scale

Category Maximum 3 s Wind Gust (km h-1) Potential Damage

1 <125 minor

2 125-170 moderate

3 170-225 major

4 225-280 devastating

5 >280 extreme

The main structural features of a severe tropical cyclone at the earth’s surface are the eye, the eye wall and the spiral rainbands. The eye is the area at the centre of the cyclone at which the surface atmospheric pressure is lowest. It is typically 20 to 50 km in diameter, skies are often clear and winds are light. The eye wall is an area of cumulonimbus clouds, which swirls around the eye. Tornado-like vortices of even more extreme winds may also occur associated with the eye wall and outer rain bands but are more likely at landfall. The rain bands spiral inwards towards the eye and can extend over 1000 km or more in diameter. The heaviest rainfall and the strongest winds, however, are usually associated with the eye wall.

For any given central pressure, the spatial size of individual tropical cyclones can vary enormously. Generally, smaller cyclones occur at lower latitudes and larger cyclones at higher latitudes but there are many exceptions. Large cyclones can have impacts far from their track, especially on waves and low levels of storm surge. For example, David crossed the coast near Yeppoon in 1976 and caused significant coastal impacts in south eastern Queensland; Roger in 1993 remained 300 km offshore of Sandy Cape but produced the highest recorded water levels in the Gold Coast Seaway in over 20 years and the highest recorded waves in over 30 years at the Brisbane waverider buoy offshore Point Lookout; Justin in 1997 offshore Cairns caused increased water levels along the entire east coast; Yali in 1998 passed 500 km east of Brisbane and caused increased water levels and beach erosion from the sunshine coast to Northern NSW.

Cyclonic winds circulate clockwise in the Southern Hemisphere and the wind field within a moving cyclone is generally asymmetric so that winds are typically stronger to the left of the

Storm Tide Study 41/22526/412243 Final Report Addendum

direction of motion of the system (the “track”). This is because on the left-hand side the direction of cyclone movement and circulation tends to act together; on the right-hand side, they are opposed. During a coast crossing in the Southern Hemisphere, the cyclonic wind direction is onshore to the left of the eye (seen from the cyclone) and offshore to the right. Given specifically favourable conditions, tropical cyclones can continue to intensify until they are efficiently utilising all of the available energy from the immediate atmospheric and oceanic sources. This maximum potential intensity (MPI) is a function of the climatology of regional sea surface temperature (SST) and atmospheric temperature and humidity profiles. When applying a thermodynamic MPI model for the Queensland coast (Holland 1997a, 1997b), indicative values for the MPI increase northwards from about 940 hPa near Brisbane to 880 hPa for regions north of Townsville. Thankfully, it is rare for any cyclone to reach its MPI because environmental conditions often act to limit intensities in the Queensland region. The present study however, makes allowance for this extreme condition.

4.1.2 Dataset Description The study has considered all available records of tropical cyclones from official BoM records (National Climate Centre) as well as local Queensland Regional Office records in Brisbane. However, only those cyclones which entered within a 500 km radius of Gold Coast have been included in the statistical analyses. The choice of a 500 km radius is based on capturing all events that would have been capable of directly affecting the Gold Coast area within a 24 hour period and provides a sufficient sample of the statistical population to enable reasonably reliable estimates to be made of intensity, frequency of occurrence and track. The complete cyclone data set since the early 1900s shows a fluctuation in recorded occurrences of tropical cyclones that is due not just to the natural variability of these large scale storms, but also the often poor detection rate prior to the introduction of satellites in the late 1950’s and early 1960’s (Holland 1981). In order to ensure a stable and reliable statistical series for model extrapolation purposes, only data since 1959/60 onwards is used in the present study. This provides a total of 50 cyclone seasons up until 2008/2009, the latest year available at the time of commencement.

The BoM tropical cyclone data set consists of a series of estimated positions of the centre of each cyclone, together with the estimated central pressure (hPa), at an interval of typically 6 hours. Little or no information about the size of the cyclone is normally available (except in recent years), so that the radius to maximum winds is a parameter which has to be further estimated. Some editing of the official data sets has been undertaken to remove duplicate storm records, correct known errors and make other adjustments based on advice from the Severe Weather Section at the Queensland Regional Office in Brisbane (Jeff Callaghan, personal communication). Appendix B provides a summary listing of all historical tropical cyclones considered in this study.

4.1.3 Analysis and Interpretation A total of 46 cyclones have occurred within the 50 season record and within the 500 km study region, averaging 0.92 cyclones per season. The time history of the frequency of cyclone occurrence is shown in Figure 4-2, showing a fluctuation about a 5 year average value of

Storm Tide Study 19 Final Report Addendum

between 0 to 3 storms per year. Some years indicate zero storms within the 500 km radius (1961/62, 1965/66, 1974/75, 1982/84, 1985/89, 1990/91, 1996/97, 1998/2004, 2005/08, 2009/10) while the maximum number during this time has been 4 storms in one season (1966/67, 1971/72, 1973/74 and 1975/76). Clustering of storm events in the past has resulted in significant coastal erosion episodes along the Gold Coast region.

The variability in cyclone occurrences over a 3 to 5 year span is now known to be strongly associated with the so-called El Niño - Southern Oscillation (ENSO) phenomenon (e.g. Nicholls 1992, Basher and Zheng 2000). ENSO refers to a quasi-biennial oscillation of the sea surface temperatures (SST) in the eastern tropical Pacific Ocean. During a so-called El Niño period, the SST is warmer than normal in the east and rainfall and tropical cyclone activity in northern Australia tends to decrease. In the reverse situation, called La Niña, the SST in the eastern Pacific is cooler than normal and rainfall and tropical cyclone activity increases along the east coast of Australia. The Southern Oscillation Index (SOI) is a measure of the strength of the ENSO episodes, derived from surface pressure data at Darwin and Tahiti. The SOI is also plotted on Figure 4-2, where it can be seen that a generally persistently negative SOI (El Niño) has been associated with a decrease in cyclone occurrences over the past 20 years in the Gold Coast region. Since 1959 though, the number of El Niño - La Niña cycles is approximately equal, although the strengths have varied (Pielke and Landsea 1999). This suggests that the long-term average frequency of occurrence of 0.901 storms per season for the statistical region is reasonably reliable. However, it should be noted that ENSO fluctuations specifically alter the true likelihood of storm tide risk in any particular year of exposure. Some researchers (e.g. Power et al. 1999) suggested that the trends of the 1980s and 90s may have started reversing and that the western Pacific could enter a period of prolonged La Niña activity in the new millennia, but following years had seen only mild La Niña or near neutral conditions persisting. Even 2008/09, with a persistently high SOI, was not classed as a strong La Niña due to mixed SST signals. However 2010/11 has established itself as one of the strongest La Niña events on record, ranking amongst the top 5 since 1900, and facilitating extensive and persistent flooding across much of Queensland, the February event that severely impacted Brisbane, and the occurrence of TC Yasi in Far North Queensland.

The corresponding time history of minimum storm central pressures is shown in Figure 4-3, illustrating the great variety possible in intensities. The 5 year average line in this case has been significantly shifted downwards around 2009 due to TC Hamish.

The tracks of tropical cyclones often appear random and chaotic but a more cohesive structure can be seen when the storms are grouped into what are believed to be common statistical populations that relate to areas of genesis and broad-scale movement. The present study assumes three basic track classes exist in this region, being offshore moving, parallel to coast and onshore moving. The 46 storm sample is split into these classes as shown in Figure 4-4. The few over-land examples of the offshore class in this region are predominantly exiting decayed previously landfalling storms moving eastwards while the over-sea examples are relatively weak near-coast developing systems. The parallel class are concentrated about 200 to 500 km offshore but also contain examples of oblique coast-crossing events and some over- land storms.

Storm Tide Study 41/22526/412243 Final Report Addendum

Tropical Cyclone Statistics Within 500km of Gold Coast

10 25 SOI Annual 9 20 SOI 5 yr 8 15 7 10 6 5 5 0 4 -5

3 -10 Annual SOIAnnual Value 2 5 yr Av -15

Number of Cyclones Number Per Year 1 -20 0 -25 1959 1964 1969 1974 1979 1984 1989 1994 1999 2004 2009

Season

Figure 4-2: Time history of the frequency of cyclone occurrence within 500 km of the Gold Coast

Tropical Cyclone Statistics Within 500km of Gold Coast

1020

1000

980 Average 960

940

920 Minimum Central Pressure Minimum hPa Central 900 1959 1964 1969 1974 1979 1984 1989 1994 1999 2004 2009 2014 Season

Figure 4-3: Time history of cyclone peak intensity within 500 km of the Gold Coast

Storm Tide Study 21 Final Report Addendum

Figure 4-4: Cyclone tracks capable of affecting the Gold Coast Region classified into top: offshore (21.7%), mid: parallel (47.8%) and bottom: onshore (30.4%).

Storm Tide Study 41/22526/412243 Final Report Addendum

The most significant parameter affecting regional storm tide is the intensity of the tropical cyclone winds. This is typically indirectly represented by the central pressure of the cyclone but also depends in part on other scale parameters. The estimated minimum central pressure for each of the 46 storms is then statistically analysed using Extreme Value Theory (Benjamin and Cornell, 1970) to obtain the likelihood of particularly intense storms occurring anywhere within the 500 km radius region. The statistical analyses are undertaken firstly for each separate track class and then combined into a single regional prediction, summarised graphically in Figure 4-5 in terms of return period and the approximate cyclone category based on Table 4. It can be shown that the most intense cyclones are contributed mainly by the onshore class, which typically represent fully mature storms in favourable steering currents.

1000

Data 980 Category Combined Tracks 2 960 3

940

4 920

900 5 Minimum Central Pressure Minimum pcCentral hPa

880 1 10 100 1000 10000 Return Period y

Figure 4-5: Extreme value analysis of cyclone intensity within 500 km of Gold Coast

Coupled with this theoretical (normally unbounded) analysis there needs to be a consideration of the maximum potential intensity (MPI) that might be sustained in any region. This is a function of a number of physical parameters but principally the sea surface temperature and the upper atmosphere profile (Holland 1997b). For the South East Queensland region the MPI is assessed as 940 hPa (Holland 1997a) – similar to the measured central pressure of Dinah when crossing Sandy Cape in 1967. Based on the present analysis, this MPI has a return period of approximately 70 years anywhere within 500 km of the Gold Coast, which indicates that a Category 4 TC is not expected within the 500 km radius under present climate conditions. Many other storm parameters are also extracted during the analysis phase. For example, the variation in forward speed, which adds to the strength of the cyclonic winds, the duration of storms, track bearing and the tendency for a proportion of storms to weaken (fill) as they move closer to the coast are based directly on the recorded data set. All of the above statistical estimates of tropical cyclone behaviour and strength have been assembled for use by the statistical storm tide model and used as a “template” to allow the generation of many thousands of synthetic storm events. The radius to maximum winds and Holland wind peakedness values

Storm Tide Study 23 Final Report Addendum

are based on recommendations in Harper (2001b) for latitude 28°S and overland decay considerations have been neglected. Table 5 summarises the key model parameters for the 500 km radius statistical region used for this study.

Table 5 Key Statistical TC Climatology Parameters for the Gold Coast Region

Track Statistical Model Parameters

Population Name Variable Units Value

Ambient Pressure pn hPa 1007

% This Track 21.7

Average Number Per Year 0.20

Gumbel Intensity U hPa 996.3

Offshore Parameters α 0.1438

Moving Max Potential Intensity MPI hPa 990

Overland Decay Fd km 1e5

- std dev 60

Radius to Max Wind mean km 76

+ std dev km 84

- std dev 0.8

Wind Peakedness mean - 0.9

+ std dev - 0.9

% This Track 47.8

Average Number Per Year 0.44

Gumbel Intensity U hPa 990.5

Parallel Parameters α 0.0715

Moving Max Potential Intensity MPI hPa 940

Overland Decay Fd km 1e5

- std dev 27

Radius to Max Wind mean km 48

+ std dev 80

- std dev 0.9

Wind Peakedness mean - 1.0

+ std dev 1.3

Storm Tide Study 41/22526/412243 Final Report Addendum

% This Track 30.4

Average Number Per Year 0.28

Gumbel Intensity U hPa 987.2

Onshore Parameters α 0.0990

Moving Max Potential Intensity MPI hPa 940

Overland Decay Fd km 1e5

- std dev 27

Radius to Max Wind mean km 48

+ std dev 80

- std dev 0.9

Wind Peakedness mean - 1.0

+ std dev 1.3

4.2 Extra-Tropical Climatology In the Gold Coast region, the extra-tropical climatology is dominated by East Coast Lows, although occasionally sub-tropical systems are present, together with extra-tropical transitioning cyclones. However, the tropical cyclone database has traditionally included many of these sub- or transitioning cyclonic events.

Much of the following material has been extracted from Harper (2001c).

4.2.1 Genesis of East Cost Lows East Coast Lows (ECLs), also known as east coast cyclones, winter cyclones or easterly trough lows, are one of a family of low pressure systems that most often develop during the winter months along the east coast of Australia between 25ºS and 40ºS (Holland et al. 1987, Hopkins and Holland 1997). These large scale storm systems often develop rapidly and can become quite intense, with storm force winds extending over wide areas. These events contribute significantly to flooding and wind damage along the coastal margins as well as marine accidents, storm surge and beach erosion in south east Queensland.

East coast lows typically form after a low or deep trough intensifies in the upper atmosphere over eastern Australia. A low pressure system then develops at sea level near the coast to the east of the upper level system, often intensifying rapidly. These cells of low pressure are typically quite small relative to the broad synoptic features but can interact with developing high pressure systems to the south to produce severe gale conditions over periods of up to several days (Allen and Callaghan 2000, Callaghan 1986). These storm systems draw their energy from a combination of strong ocean temperature gradients, coastal convergence, uplift and a supply of moist sub-tropical air at the surface. The East Australian Current and the Great Dividing Range are principal players in the development of these storms, the circulation centres of which

Storm Tide Study 25 Final Report Addendum

often track very close to the coast over considerable distances. An example of the tracks of several prominent systems is shown in Figure 4-6.

Figure 4-6: Example of an historical East Coast Low and some of the many ECL storm tracks affecting the Gold Coast region (after Harper 2001c)

Although the nominal storm centres may be close to the coast, their impacts extend over considerable distances, as can be seen in the example, where the steep gradients in the surface pressure fields and regions of strong onshore winds are indicated. The onshore flow is responsible for the heavy rains and, combined with the extended fetch regions over the ocean, the generation of high waves. Generally low but persistent storm surge impacts are also possible, whereby the strong clockwise winds create a net onshore flow at the surface causing a rise in water levels along the coast. The “inverted barometer” pressure effect can also be significant, with some east coast lows having central pressures below 990 hPa. Wave setup caused by breaking wave processes at the coast also contributes to the total storm tide impact.

Prior to the introduction of satellite imagery in the early 1960s, many east coast lows were classified as tropical cyclones. While their impacts may be similar or even possibly greater in some cases, the east coast low has a different physical mechanism and a highly asymmetrical poleward cloud pattern where the heaviest rainfall frequently occurs. Another feature of east coast low development is the tendency for clustering of events when conditions remain favourable. For example, near Brisbane, almost one third of events occur within 20 days of a preceding event (Allen and Callaghan 2000).

Storm Tide Study 41/22526/412243 Final Report Addendum

4.2.2 Climatology of East Coast Lows There have been a number of studies into the frequency of occurrence and relative intensity of east coast lows. PWD (1985) addressed the coastal impacts of these systems on the NSW coastline, especially from a storm surge and wave setup perspective. Callaghan (1986) and Holland et al. (1987) considered the synoptic precursors to storm development as an aid to forecasting. Hopkins and Holland (1997) looked at the association between east coast lows and heavy-rain days. Allen and Callaghan (2000) considered the impacts of east coast lows on extreme wave heights in the SE Queensland coastal region. Unfortunately, east coast lows have not been systematically recorded in the manner that tropical cyclones have been since the turn of the century. They are typically more complex systems which are often difficult to categorise. Accordingly, many of the studies have concentrated on detailed investigations of historical weather charts and station observations to reconstruct a time history of occurrences. The longest assembled record available (1880 to 1980) is from PWD (1985), which considered the region from Tweed Heads south to Gabo Island, near Bass Strait. This study classified the various storm systems into six categories, depending on the synoptic situation, as summarised in Kemp and Douglas (1981). Holland et al. (1987) considered the period 1970-1985 and used three broad classifications. Hopkins and Holland (1997) broadened this to 1958-1992 and Allen and Callaghan (2000) focussed on 1976- 1997 when wave data was available.

In Harper (2001c) a composite data set was created based essentially on PWD (1985), using their categories E, S, I and C for the northern sector, and Allen and Callaghan (2000) using their type 1 and 2 events. Two additional heavy rain events from Hopkins and Holland (1997) were also included. This composite set covered the 118 year period 1880 – 1997 and considers only those east coast low events which had some impact on SE Queensland. On this basis the areal extent of the data set was within about a 500 km radius of Brisbane.

More recently, the Bureau of Meteorology initiated an attempt at a comprehensive study of maritime cyclones and related weather events affecting the NSW coastal areas - the NSW Maritime Low Database Project (NMLDP). This is of relevance to South East Queensland also because the study area extends northwards to Fraser Island and the data record ranges from 1959 to 2006.

For the purposes of this study, a further composite dataset from the NSW Maritime Low Database Project was created based on two factors;

 The Longitude and Latitude of the storms (within 500 km radius of Gold Coast);

 The “Eastern Troughs” as classified in the original project. These two factors have been named as NML 1 and NML 2 respectively.

This brief overview provides a comparison of the current study with the previous studies mentioned in Harper (2001c). This new composite dataset covers the 48 year period 1959 – 2006 and considers only those east coast low events which had the potential for some impact on SE Queensland. On this basis the areal extent of the data set is within about a 500 km radius of Gold Coast. During the creation of the composite dataset it was confirmed that the NMLDP database was checked against data from Hopkins and Holland (1997), Figure 4-7 presents this data set comparison as the 10 year averaged number of storms, overlaid by a 10

Storm Tide Study 27 Final Report Addendum

year averaged Southern Oscillation Index Value (SOI), where “East Coast Lows” refers to Harper (2001c).

East Coast Lows NML 1 NML 2 SOI

8.0 5 7.0 4 3 6.0 2 5.0 1

Storms 4.0 0 3.0 -1 -2 2.0

-3 10 yr Averaged Number of yr Averaged 10

1.0 -4 10yr Averaged SOI ValueAveraged 10yr 0.0 -5 1885 1905 1925 1945 1965 1985 2005

Figure 4-7: Composite Datasets of East Coast Lows Affecting SE Queensland

It is important to remember that, like tropical cyclones, the availability of regular satellite imaging revolutionised the monitoring of these types of weather events. In 1960, experimental satellite images became available. However, it was 1966 when two images per day could be obtained from the polar orbiting satellites. Prior to the availability of satellite imagery, significant under- sampling of east coast lows is likely. On the contrary, it is interesting to note that regardless of NML 1 and NML 2 the east coast data displays a sharp increase in the number of storms after the late 1960s. This gives an indication that the advent of satellite imagery resulted in a number of additional storms (with no significant impact) being added to the dataset resulting in over- sampling of east coast storms.

Notwithstanding the above information it remains problematical to model these events with any accuracy and thus the present study has adopted an empirical approach that bypasses the need for the meteorological analysis and deals directly with the recorded water level impacts, which are sufficiently numerous to argue that they can be used for reliable statistical analysis.

Storm Tide Study 41/22526/412243 Final Report Addendum

5. Numerical Model Development

This section describes the necessary tropical cyclone wind and pressure model, storm surge, wave and statistical models adopted for the project.

5.1 Tropical Cyclone Wind and Pressure Model The model is based on Holland (1980) and Harper and Holland (1999) and detailed in Harper (2001a). It has been used extensively throughout Australia and internationally to represent the broad scale wind and pressure fields of a mature tropical cyclone. It relies on a series of parameters to describe a tropical cyclone when it is over an open ocean environment, namely:

 the central Mean Sea Level pressure p0

 the surrounding, or ambient, pressure pn

 the radius to maximum winds R

 the wind field peakedness factor B; and

 the storm track (speed Vfm and direction θfm) The model generates estimates of the 10 minute average wind speed and direction at a height of 10 m above the ocean surface for supply to the hydrodynamic models for storm surge and waves. It also estimates the 3 sec wind gust for comparison with long term wind records at regional sites such as Brisbane Airport. The MSL pressure is also supplied to the hydrodynamic model as it has an influence on the generation of the storm surge.

Examples of model-generated wind fields are shown in Chapter 6 in respect of the demonstration of TC Dinah and the “1954” cyclone.

5.2 Hydrodynamic Model With respect to hydrodynamic modelling, GHD’s first task was to retrieve and select, from a number of existing models under Council’s custody, the most adequate model for the project, and to redevelop and enhance the selected model for optimised performance while delivering to Council a calibrated engineering tool for future hydrodynamic implementations.

Three models, developed using the finite-volume method as implemented within the DHI Mike suite of models, were considered. These are:

 the Regional HydroDynamic (RHD) model (see details of the model in sections 5.2.1 to 5.2.3)

 an intermediate scale, high-resolution model incorporating the river systems discharging into the Broadwater and referred to as the Local HydroDynamic (LHD) model and

 a high resolution, local-scale Gold Coast Environmental Modelling System or GEMS representing the river systems and adjacent flood plains. After reviewing the extent of the models and undertaking a series of performance tests, the conclusion was reached that the LHD and GEMS models did not fully satisfy the requirements of the project.

Storm Tide Study 29 Final Report Addendum

Accordingly, all deterministic storm surge and associated wave modelling was built around Council’s existing large scale Regional HydroDynamic (RHD) model.

5.2.1 Computational Grid The original RHD model is a two-dimensional, depth-averaged implementation of Mike Flexible Mesh (Mike FM) developed in 2008 by DHI Water And Environment Pty Ltd. The model covers approximately 2.27 million square kilometres of the Coral Sea offshore from the Queensland Coast with its eastern boundary reaching 164ºE as shown in Figure 1-3. Along the coast, the model extends from 16ºS to 31ºS on a relatively coarse grid. The model has a total of 41,917 elements and 21,603 nodes. Mesh size and mesh quality together with quality coastal bathymetry and offshore boundary conditions are key to successful model calibration and operation. To ensure the later, the original RHD model was redeveloped to include several major modifications which are described in the following sections.

5.2.2 Mesh Size and Extent of Model Preserving the extent of the original RHD model and implementing the recommendations presented in the Government-sponsored Queensland Climate Change investigations (e.g. Harper 2001, 2004) were key requirements of the project. With respect to mesh size and model extent, it is required to ensure adequate resolution of the wind field and provide proper global- scale forcing of the storm tide.

Taking advantage of the flexible (or unstructured) character of the mesh, the mesh size, of the order of 35 km offshore, was gradually decreased down to 800 m in the north section of Moreton Bay, 600 m around Jumpinpin and in the range of 70 to 150 m in the Broadwater. A substantial amount of effort was spent on representing the network of channels and islands in the Broadwater while keeping the total number of mesh elements to a minimum. The newly generated triangular mesh (smallest allowable angle of 29 degrees) has a total of 36,809 elements and 19,391 nodes. While the total number of elements has been decreased in comparison with the original RHD model, there are substantially more elements in the Broadwater area under the new configuration.

The new mesh was adopted for tidal calibration of the model and all storm surge and wave operational runs undertaken for the study. The generation and nesting of additional local, high resolution models near the coast was therefore not necessary thus generally improving modelling efficiency. The results from the tidal calibration of the model presented in section 5.2 confirm that the quality of the newly generated mesh is satisfactory.

5.2.3 Bathymetry and Coastal Features The original RHD model (as provided by Council) was mainly conceived as a global-scale, parent model with the intent to generate storm surge type of boundary conditions for local, smaller-scale, high-resolution coastal models. The representation of the Broadwater area in the model was therefore rather coarse and key features such as islands, tidal flats, channels and river entrances were excluded from the schematization. Figure 5-1 illustrates the complexity of the bathymetry in the approaches from north to the Broadwater (Moreton Bay Islands) and

Storm Tide Study 41/22526/412243 Final Report Addendum

within the Broadwater. As evident from the figure, the conveyance capacity of the channels may vary substantially along their length with the data showing evidence of shallow reaches that act as a weir with respect to the tidal streams and truncate the tidal signal. A coarse or inaccurate representation of the sea-bed has the potential to misrepresent the conveyance of the waterways and to reduce significantly the reliability of the final results.

As described in chapter 3, all bathymetry and coastline information used in the model have been extracted from C-MAP and further augmented with the best available (circa 2007) locally generated bathymetric datasets. It is beyond the scope of the present study to document how the bathymetry of the individual channels may have been modified in the last 10 years and what impact, if any, a potential change in bathymetry may have on the calibration of the RHD model presented in Chapter 5. All hydrographic data used in the tidal calibration of the RHD model was collected by Council in 2004 and 2005.

Moreton Bay Islands Coomera and Sovereign Islands

Gold Coast Seaway Southport

Figure 5-1: Bathymetric Data Used in the Generation of the RHD Model

5.2.4 One-Dimensional Links A series of 1-D (Mike-11) models, representing four river systems (The Logan, Pimpama, Coomera and Nerang) and two creeks (Tallebudgera and Currumbin) that fall under GCCC

Storm Tide Study 31 Final Report Addendum

jurisdiction, were integrated with the RHD model with the aim of (1) reproducing the essential response of the area to storm tide, (2) achieving a reasonable level of calibration and (3) ensuring that “river and lake” locations of interest to GCCC are accounted for in the statistical aspects of the analysis thus allowing local design storm tide levels for planning and development purposes to be established. Figure 5-2 serves as an example illustrating the integration or linkage of the Coomera River system with the RHD model. Black dots along the alignment of the river denote river cross-sections, large yellow dots indicate field monitoring stations and large green dots are used to indicate the location of SATSIM points, i.e. where statistical results are reported by the SATSIM model. As seen from the figure, care has been taken to position all SATSIM points within the Gold Coast Broadwater in deep water where dry land (which could temporarily disable the relevant SATSIM point) is not likely to occur during simulation. The consideration applies to the river locations as well.

Figure 5-2: One-dimensional Representation of the Coomera River system in the RHD model The main function of the one-dimensional links is to schematically represent the storage potential of the river systems connected to the Broadwater and thus to contribute to the accurate representation of the tidal prism in the area. In order to represent the links, two sets of one-dimensional Mike-11 models were tried. The first one was created by GHD specifically for the project while the second one comprised a set of calibrated models created by Council. It was found that Council models, while presenting the advantage of being carefully calibrated needed a substantial amount of re-work in order to be integrated in the RHD. As a result, the set of Mike-11 models developed by GHD was adopted for the project. The option of using Council

Storm Tide Study 41/22526/412243 Final Report Addendum

models in future implementations of the RHD could improve further the quality of the predictions.

5.2.5 Offshore Boundary Conditions The original RHD model (as provided to GHD) did not have open sea boundary conditions. Instead, all boundary polygons were set to land. To enable the modelling of tides, the north, east and south boundaries of the model were set as open and a tidal signal for the duration of the calibration period (see section 5.2.6) generated along these boundaries using two global tidal models.

The first global tidal model is the one integrated into the DHI suite of models. The second one is operated by the National Tidal Centre or NTC. The first model provided directly time histories of water-surface elevation along the open boundaries of the RHD model whereas the second global tidal model delivered 8 major tidal constituents per each of the nominated boundary locations which were then used to generate the time histories of water-surface elevation necessary to drive the model.

A series of tests were undertaken with both sets of boundary conditions and the results of the tests compared to tidal records at various coastal locations(e.g. Gold Coast Seaway, Brisbane Bar, Tangalooma, Mooloolaba, etc). The comparison led to the conclusion that the tidal signal provided by the NTC was of superior quality and the NTC dataset was adopted for the tidal calibration of the RHD model.

5.2.6 Model Calibration to Tides Verification of the model by comparison to both tidal elevation data and storm surge data is provided to demonstrate that the model is capable of simulating realistic surface elevations for historical storm events for which measured surge data are available.

With respect to tides, the RHD model was calibrated against water surface elevation and instantaneous flow discharge measurements at as many as 20 of the 29 locations listed in Table 3. The results of the calibration (to water surface elevation) are presented in Table 6 quantitatively analysed in terms of the root-mean-square-error or RMSE. Provided a time-series of observed Oi and simulated Si values, RMSE are calculated as follows:

2 Oi  Si  RMSE   N where N denotes the number of samples included in the statistical measure. Numbers in bold in the table indicate better performance. Names in italic pertain to stations deployed in river and lake environment. Two major observations noted from the quantitative analysis summarized in Table 6. First, RMSE has been estimated for long periods – up to 7 months for analysis involving tidal predictions and up to 4 months for analysis involving tidal measurements. The shorter period used to estimate RMSE was that for Brygon Creek – two weeks, subject to data availability. The second observation is that model performance is sensitive to channel conveyance – a parameter that could be partially controlled by the so-called depth cut off parameter in the DHI

Storm Tide Study 33 Final Report Addendum

modelling environment. Results obtained with a -3.0 m depth cut off are superior to those obtained with a -4.0 m depth cut off (e.g. better RMSE for a larger number of stations). Under - 3.0 m depth cut off conditions, the best performance of the RHD model was observed at the Gold Coast Seaway Station – RMSE of 0.06 m, whereas the Steiglitz Jetty produced the worst RMSE estimate – 0.20 m. The results are in accordance with expectations, that is, the predictions corresponding to the Gold Coast Seaway station benefit from the high resolution and high quality bathymetric data used in the modelling of the Broadwater area. On the contrary, at remote locations where grid resolution may not be sufficient thus a coarse rendering of the bathymetry is expected to occur (Steiglitz Jetty), the match between model results and measurements is less satisfactory. It is also speculated that in some instances datum issues may have affected the data provided by Council and model performance will further improve if calibrated one-dimensional fluvial links are adopted.

The results from the tidal calibration of the RHD model (-3.0 m depth cut off conditions) are good considering the global extent of the model and the dynamic nature of the sea bed reflected in potential differences between the existing bathymetry in 2004, at the time when field measurements were made by Council, versus bathymetry as obtained from C-Map database. The performance of the enhanced RHD model both in terms of water surface elevation (Figures C-1 to C-23) and instantaneous flow discharge (Figures C-24 to C-38) has been illustrated in a series of graphs provided in Appendix C. The following notation has been adopted in the graphs:

 Blue solid line – predicted tidal signal based on 8 major constituents as presented in the tidal tables

 Red solid line – tidal signal as simulated by the RHD model

 Purple solid line – tidal signal obtained using the one-dimensional links integrated into the RHD model

 Black solid line – measurements. As indicated in section 3.2, the water surface elevation measurements consist of continuous records at 15-minute intervals of varying duration depending on location – two weeks at Brygon Creek and up to two months at the Logan River mouth. Instantaneous flow discharge records are shorter with each record capturing one peak flood and one ebb condition. Irrespective of the short duration of the flow discharge records, a period of several tidal cycles is included in the plots in order to inform on the range and variability of the measured entity during this period.

Comparisons of instantaneous flow discharge (modelled versus measured) confirm the findings from the calibration to water surface elevation with the best matches between simulation results and measurements found at the Gold Coast Seaway entrance but also at the mouth of the Nerang River and the southern arm of the Coomera River. Based on these results and the set of results obtained from the calibration of the RHD to historical events (refer next section), a series of production simulation runs using -3.0 m depth cut off conditions were undertaken with the results of these runs presented in Chapter 7.

Storm Tide Study 41/22526/412243 Final Report Addendum

Table 6 Tidal Calibration - Quantitative Analysis of Model Performance

Monitoring Station -3.0 m depth -4.0 m depth Period of RMSE estimation cut off cut off

Jumpinpin 0.11 0.09 4 months

Gold Coast Seaway 0.06 0.10 7 months

Brisbane Bar 0.10 0.08 7 months

Tangalooma Point 0.10 0.11 7 months

Bongaree Jetty 0.09 0.11 7 months

Mooloolaba 0.07 0.07 7 months

Steiglitz Jetty 0.20 0.21 1 month

B9 Grand Hotel Labrador 0.15 0.15 1 month

B10 Southport Seaway 0.13 0.13 4 months

L1 Logan River Boat Jetty 0.15 0.20 3 months

B3 Logan River Mouth 0.14 0.17 3 months

B7 Coomera River Mouth 0.09 0.09 3.5 months

C2 Marine Precinct Coomera 0.10 0.12 4 months

Monterey Keys 0.07 0.08 5 weeks

Coomera Shores 0.09 0.10 5 weeks

Brygon Creek 0.10 0.11 2 weeks

B11 Nerang Mouth 0.13 0.13 4 months

Ni Evandale 0.09 0.09 5 weeks

N2 Seven Oak (Nerang) 0.13 0.12 5 weeks

N3 Royal Pines (Nerang) 0.11 0.10 4 months

5.2.7 Model Calibration to Historical Storm Events The process of model calibration and its outcomes are presented in Appendix D.

5.3 Spectral Wave Model Sharing the unstructured mesh of the RHD, a fully spectral wind-wave model based on the wave action conservation equation has been developed to calculate the growth, decay and transformation of wind-generated waves in the Gold Coast region. The wave model is an implementation of Mike 21 SW with the discretisation of the governing equation in spectral

Storm Tide Study 35 Final Report Addendum

space performed using a cell-centred finite volume method while the time integration is performed using an explicit fractional step approach (Mike 21 SW User’s manual).

5.4 Establishment of the Parametric Tropical Cyclone Models

5.4.1 Tropical Cyclone Parameter Selection for Full Scale Modelling The climatology assessment from Section 4.1.3 identified the principal cyclone parameter values likely to apply to the Gold Coast region. These form the basis of a series of conceptual straight-line and constant speed synthetic cyclone tracks, which when modelled systematically by the fully numerical models, provide a response function for surge and waves that can be readily interpolated to provide output at any coastal location. Each of the conceptual cyclones is described by the same set of parameters as presented in Section 4.1.3, except that the pressure difference p is introduced to specify the storm intensity:

p = pn – po A total of 294 individual simulations were then used to form the “base” storm surge response and 189 simulations to form the wave response, as summarised in Table 7. These comprised three values each for intensity, radii, wind field peakedness factor B, forward speeds and angles of approach.

Table 7 Base Storm Parameter Set

Δp R B Vfm θfm Coastal Crossing Distance X

hPa km ms-1 bearing˚ km

220˚ 260˚ 160˚

30 25 0.9 2 220 150 150 0 (0)

60 50 1.0 6 260 100 100 -50 (-100)

80 80 1.3 8 160 50 (200) 50 (200) -100 (-200)

0 (0) 0 (0) -200

-50 (200) -50 (200)

Totals

3 3 3 3 3 5 (3) 5 4 (3)

The coastal crossing distance shown here is measured from Surfers Paradise, this being a convenient reference location. The numbers in brackets refer to the wave model track selections, which were distributed more widely than the surge. Each simulation included an elapsed real time of 30 h, with the start of the cyclone being 18 h before “landfall” and continuing until 12 h afterwards. In the case of the parallel-moving storms, “landfall” is the time

Storm Tide Study 41/22526/412243 Final Report Addendum

of closest approach to the reference location. Each model cyclone also underwent an additional initial 12 h build-up period, with the storm held stationary, to reduce numerical transient effects. In addition to the base set of storms, a series of 6 special sensitivity tests (Table 8) were also undertaken to explore the surge and wave response at the upper and lower limits of the storm intensity ranges and to check linearity and scaling assumptions. These were done with a selected range of fixed values for the other parameters.

Table 8 Additional Parameter Sensitivity Testing

Type p R B Vfm θfm X Water Level

hPa km - ms-1 ° km

Intensity 10 50 1.0 6 160 0 MSL

105 50 1.0 6 220 50 MSL

260 50 MSL

Water 60 50 1.0 6 160 0 -1.0m Level 220 50 +1.0m

260 50

Finally, since all the base simulations were conducted at mean sea level (MSL), a further set of 6 sensitivity tests was undertaken at +1.0 m and -1.0 m, representative of the approximate tidal range in the region. These results were used to devise a surge-tide interaction function for the model (refer Section 5.5.2). Each simulation provided a time history output of water elevation, wave height, period and direction each 10 minutes at 171 B-grid locations from Cooktown to Townsville and 201 C-grid coastal locations from Palmer Point (near Fishery Falls) south to Dunk Island and Tully Heads.

5.4.2 Processing of the Numerical Model Results Each of the above full scale numerical model simulations were processed according to a method developed by SEA (2002), which combines the output in such a way as to extract the underlying regional and local storm surge and wave responses. The method is an enhanced form of earlier analyses undertaken for the Beach Protection Authority (Harper and McMonagle 1985). All of the model output for each track direction is condensed into a series of characteristic alongshore and offshore spatial profiles and a time history profile, all of which are scaled according to the intensity of the cyclone, its size and speed. Multiple track directions can be added as necessary to complete the description of the regional response. Additionally, each specific location is allocated a local response function that describes any localised changes in surge or wave height behaviour (including time differences) peculiar to that location. The method allows the rapid recreation of a storm surge or wave height response at any of the coastal locations based on a set of supplied storm parameters.

Storm Tide Study 37 Final Report Addendum

The parametric model is optimised for highest accuracy at the time of the predicted peak condition (surge or wave height) and typically reproduces the numerical model results to within about 5% for surge and within 0.5 m for wave height and within 2 s for peak spectral wave period.

5.5 Simulation Modelling of Tropical Cyclone Impacts The statistical simulation model SATSIM (Surge and Tide Simulation) has been developed over many years. Originally based on Harper and McMonagle (1982), the present model includes an enhanced form of the SEA (2002) parametric surge and wave model. In summary, the model generates an artificial history of tropical cyclones based on the assumed parameter climatology. The model maintains a clock that calculates the occurrence of the next event based on random number sequences and then allocates the necessary parameters, randomly sampled from the climatology distributions. Each cyclone’s predicted wind, surge and wave response at each of the sites of interest is then generated by the parametric models, interpolating as necessary between the available modelled scenarios. The wave height and period estimate is converted into a breaking wave set-up height before being added to the surge and both are superimposed on the background astronomical tide for that date in time. Wave run-up height above the mean water level (excluding set-up) is also predicted at the 1% exceedance level. This is repeated for 50,000 years of synthetic cyclones and the exceedance statistics of the combined total water level at each site then forms the basis of the probabilistic storm tide level predictions.

5.5.1 Astronomical Tide Effects Astronomical tides in the region are semi-diurnal with a marked diurnal inequality (a significant difference between heights of consecutive high or low tides). The Standard Port for the region is the Gold Coast Seaway, where values of the tidal constituents have been obtained from MSQ and the tidal planes are given in the Tide Tables (QDOT 2010) as:

Table 9 Tidal Planes at Gold Coast Seaway

Tidal Plane Abbreviation m AHD

Highest Astronomical Tide HAT 1.15

Mean High Water Springs MWHS 0.66

Mean Sea Level MSL 0.00

Mean Low Water Springs MLWS -0.65

Lowest Astronomical Tide LAT -0.76

In order to provide for this variation in tide amplitude along the coast and within the complex waterways, a simple linear interpolation of tidal planes has been undertaken between the available data, leading to a set of "range ratios" relative to the standard port above. The tidal planes were based on published MSQ values augmented by some results obtained from the 2- D hydrodynamic modelling. At many of the modelled sites, especially those north of Coomera and in the 1-D linked node locations, the range ratios are only approximate. Tidal phase

Storm Tide Study 41/22526/412243 Final Report Addendum

differences are generally small and not included as they simply represent a further random variation within the model.

5.5.2 Surge – Tide Interactions As discussed earlier, special model tests were undertaken to determine the extent to which there might be non-linear interaction between the astronomical tide and the storm surge in this region. If there is, the test conducted at a consistently low tide level (-1.0 m AHD) would normally be expected to produce a larger storm surge component than that conducted at a consistently high tide level (+1.0 m AHD). While this does not represent the actual dynamics of surge-tide interaction, it is likely that it will detect any significant level of dependency. The results of these tests indicated a very low level of non-linear interaction (mean < 2%) and so no interaction was assumed in the subsequent statistical modelling.

5.5.3 Wave Setup and Wave Runup The total water level is calculated on the basis of the modelled tide plus surge result, a breaking wave set-up component after Hanslow and Nielsen (1993) and a 1% wave run-up estimate after Nielsen and Hanslow (1991). Both the wave set-up and run-up magnitudes are relative to the local still-water tide plus surge level.

For partly-barred entrances such as Tallebudgera and Currumbin Creeks, the degree of wave setup that might be present will likely depend on the entrance geometry at the time and very likely would change during extreme events. Without undertaking more detailed assessments outside of the present scope it is recommended that a nominal 50% of the open coast wave setup component be assumed at these entrances when assigning tailwater levels for fluvial flooding as a minimum precautionary allowance.

It is emphasised that the provided tide plus surge estimates are considered inherently more reliable than those that include wave setup or wave runup. This is due to the likely wave interaction with very localised small scale dynamic coastal features.

5.6 Modelling of Extra-Tropical and Remote Tropical Cyclone Impacts This aspect of the analysis was undertaken by the Australian Maritime College Marine Modelling Unit (MMU) and follows the method described in Hardy et al. (2004) used for estimating extra-tropical cyclone storm surge contributions in the Townsville region. It is based on the re-sampling of the tidal residual record from a suitably long and reliable tide gauge record in the region of interest (e.g. Figure 5-3). Implicitly it is assumed that the available record of ocean water levels from the tide gauge has fully captured the inherent range of variability of extra-tropical and remote tropical cyclone storm surges in the region, which have combined with the tidal variation to produce the total storm tide level. While this cannot be guaranteed, the incidence of the storms of interest is relatively frequent (averaging two or three per year of some significance) and a record of the order of 20 years is highly likely to have sampled close to the maximum energy level possible from these events. Importantly though, this empirical technique does not allow for any extrapolation of storm surge magnitudes beyond those already measured.

Storm Tide Study 39 Final Report Addendum

Figure 5-3: The Gold Coast Seaway tidal residual record.

5.6.1 Regional Variability in Tidal Residuals Tidal records were obtained from MSQ for Brisbane Bar (36 y), Mooloolaba, Gold Coast Seaway (23 y) and from the NSW MHL for Tweed River and Offshore Tweed (28 y). The data was decimated to represent hourly values to overcome the typically changing data interval over time and harmonically analysed to obtain the tidal constituents. This also required some correction and repair of the data in some cases. The residual water level time history was then obtained by subtracting the harmonically-predicted tide levels over the period of the measurements. This analysis showed that some regional variability existed in the residual magnitudes, with the Gold Coast Seaway being the highest and the Tweed Offshore being the lowest, with The Tweed River, Mooloolaba and Brisbane residual magnitudes being between these limits.

The identified variability was greater than expected and the “exposed” coastal locations of the Seaway and Tweed River were significantly greater than the other locations, leading to some speculation as to the possible influence of wave-related effects (either wave setup or gauge characteristics). Nevertheless, after some investigation, it was not possible to objectively reject the higher residuals evident in the Seaway tide record, which were mainly associated with TC Roger in 1996, and these were accepted as applicable to the Broadwater. The Offshore Tweed residuals have been retained for consideration of the “open coast” water levels.

Figure 5-4 presents the 7 highest residual events from the Seaway tide gauge that illustrate the scale and persistence of these events. As noted, some of these events were associated with significant rainfall and this may have affected the recorded water level, but detailed examination of this was beyond the present scope. Importantly though, the persistence of the water level residual (storm surge) is such that it will effectively penetrate the complex waterways and

Storm Tide Study 41/22526/412243 Final Report Addendum

appear as a temporary increase in the mean water level for several days. This contrasts with the shorter period tidal signal, which is significantly modified through friction and attenuated throughout the waterways such that the MHWS and HAT tidal planes, for example, are typically reduced. Similarly, tropical cyclone induced storm surge is of a similar period to the tide and is also likely to be attenuated.

Figure 5-5 illustrates the identified variability in the residual magnitudes across the various tide gauges for the specific event of TC Roger in 1996. These differences remain unresolved at this time but it should be noted that the Seaway tide gauge at the time was located quite close to the entrance (opposite Wavebreak Island) and that wave effects may have influenced this result. Subsequently in July 1999 the gauge was moved much further south to the Marine Operations Base on The Spit. It is assumed in this analysis, consistent with the expressed view of MSQ, that these two locations can be considered identical for tide purposes. However, the fact that 6 of the top 7 events at these sites has occurred prior to 1999 castes some suspicion on this assumption, which could only be resolved by a statistical analysis of the difference in “storminess” over the two periods and some critical examination of the earlier instrumentation. Without this analysis the more conservative of the two tide gauge options (Seaway and Offshore Tweed) has been adopted.

TC Roger ECL/rain ECL/rain ECL ECL/rain ECL ECL

Figure 5-4: The Seven Highest Residual Water Level Events from the Seaway Tidal Record.

Storm Tide Study 41 Final Report Addendum

Figure 5-5: The Variability in Residual Levels for TC Roger Across the Available Tide Gauges.

5.6.2 Simulation of Residuals After selection of the Seaway residual record as representative of The Broadwater the tidal re- sampling simulation was conducted as outlined below in Figure 5-6. This was repeated for the Offshore Tweed gauge for application to the open coastline from Jumpinpin to Point Danger. Brisbane Bar residuals were re-sampled against tide predictions generated at the Logan River mouth in southern Moreton Bay using constituents4 derived from the harmonic analysis of Council’s measured data. For the Coomera River Mouth (South) site, the published MSQ (2012) range ratio of 0.86 for Runaway Bay5 was applied relative to the Seaway values. The sensitivity to the remote TC Roger event at the Seaway was also investigated and this is illustrated in Figure 5-7 in comparison with the Offshore Tweed gauge.

4 The published MSQ offset from MSL to AHD of 0.11 m has later been applied to the Logan River site. 5 Although Paradise Point is closer to the Coomera site, the Runaway Bay site has been deemed to be more representative because of its more open situation. In any case the differences are likely small.

Storm Tide Study 41/22526/412243 Final Report Addendum

Step 1 – Analyse tidal signal at point of interest to obtain harmonics and residual

Step 2 – Generate 1000 years of predicted tides and overlay the residuals

40 x 23 years = 920 years + + + + + Predicted tide over approx 1000 years

1800 1850 1900 2750 2800

Combined Tide + Residual for 920 years

Step 3 – Re-sample the 1000 years of predicted tides and residuals

Combined Tide + Residual for 920 years

1800

- + Start time shifted + - 1 week x 50 times (equiv. approx 50,000 years)

Figure 5-6: Overview of the Tidal Residual Re-sampling Process.

Figure 5-7: Sensitivity of the Seaway simulation return periods to the single large TC Roger event.

Storm Tide Study 43 Final Report Addendum

5.6.3 Extra-Tropical and Remote Tropical Cyclone Wave Setup The influence of waves and wave setup has been neglected for areas that are not on the exposed open coast, but for the other areas the extra-tropical component has been estimated as follows. Firstly, the nearby 23 y wave record from the “Seaway waverider” was inspected and subjected to Extreme Value Analysis. However it became clear that non-homogeneity in the data, likely due to the repositioning of the buoy at various depths over time, had contaminated this dataset.

Next, the 21 y analysis of regional wave heights by Allan and Callaghan (1999) based on the “Brisbane Waverider” located near Point Lookout on North Stradbroke Island was used. This had the advantage that the analysis was presented for “Total”, “Cyclonic” and “Non-Cyclonic” wave events. Using this, a regression between the various components was developed and these relationships were applied to the results of the nearshore spectral wave modelling incorporated into the SATSIM tropical cyclone simulation model.

This permitted estimation of the likely non-cyclonic and “total” wave climate at the open coast sites based on the cyclonic estimates alone, but with an allowance for wave breaking in the shallower waters. Wave setup elevations were then estimated analogously to the tropical cyclone method but assuming a fixed Tp value of 10 s, which is representative of the area. Although there is likely not a clear stratification between the event sets of Allen and Callaghan and the present study, it can be noted that the non-cyclonic set tends to dominate the wave climate in this record and especially at lower return periods. The “Surfers Paradise” model site is used as the nominal open coast site for estimating the combined tide plus surge plus breaking wave setup elevations as were listed in Table 1.

Storm Tide Study 41/22526/412243 Final Report Addendum

6. Tropical Cyclone Model Calibration and Verification

6.1 Deterministic Verification of the Tropical Cyclone Models The deterministic accuracy of the numerical wind, wave and storm tide models has been tested by demonstration hindcasts of the effects of tropical cyclones TC 1954 and TC Dinah 1967, although a lack of quantitative data restricts the extent to which comparisons can be made. Appendix D details the results of the hindcast analyses for reference.

In summary, the modelled winds and pressures compare favourably with the available quantitative measures across the region. The modelled wave and storm tide time series are then consistent with the qualitative impacts reported from each event, but the measured peak tidal residual levels (storm surge) from the Brisbane Bar are of the order of 0.3 to 0.4 higher than modelled. Unfortunately there is no additional objective water level data available outside of Moreton Bay, which has not been comprehensively modelled as part of this investigation. However, it remains likely that broadscale effects are also responsible for this slight underprediction. An alternative view could be that local wave setup in Moreton Bay is also partly responsible for the model underpredictions. In any case the degree of mismatch is not regarded as significant enough to caste doubt on the ability of the modelling system to represent conditions in the Gold Coast region.

6.2 Statistical Verification of the Tropical Cyclone Simulation Model There is no clear method by which the statistical aspects of the model can be verified, other than ensuring that the various component parts of the model are performing correctly. The only statistical checks that can be done relate to the model’s re-creation of the astronomical tide statistics and a comparison of its wind speed predictions with long term regional values.

Figure 6-1 shows the modelled statistics of high tides at Gold Coast Seaway, compared with the specified HAT tidal plane of 1.15 m AHD. Normally, HAT is associated with an 18.6 y tidal cycle; hence it should fall at around the 20 y return period value if fully sampled. However, the statistical model only samples 6 months of each year (the cyclone season from November to April) and so the apparent return period of HAT here has been essentially doubled. The remaining differences are due to the use of a half-hour tidal sample, a 0.1 m discretisation level in the model and a reduced set of tidal constituents (the principal 37 only) being used. On this basis, the model is deemed to be correctly sampling the astronomical tide.

The next test considers the model’s prediction of mean and gust wind speeds when compared with an analysis of up to 48 years of gust wind speed data from Brisbane Airport, which is the longest record available in the region, and also 29 years of data from Cape Moreton. Gust wind data has been chosen in preference to the available synoptic (3 hourly 10 min means) because of the more reliable daily peak sampling, which leads to a more homogeneous record. The raw wind data was obtained from the Bureau of Meteorology and analysed to extract the peak winds occurring only during periods when a TC was within a 300 km radius of the site so that the effects of other severe weather such as isolated local thunderstorms were likely excluded or

Storm Tide Study 45 Final Report Addendum

reduced. While the Brisbane data is likely of a higher precision than the Cape Moreton data6 due to its longer instrumented period, The Cape Moreton site is more representative of the open ocean winds that the model seeks to represent. Importantly though, the high elevation of Cape Moreton necessitates an adjustment for topographic effects, which has been based on the wind tunnel testing of the site reported in Ginger and Harper (2004). The data were also windowed over a 7 day period to ensure independent samples were obtained and are ranked in Figure 6-2 using traditional quantile plotting techniques (Harper 1999). The SATSIM predictions for mean and gust wind speeds based on a 50,000 y simulation are then overlaid, with the model assuming a nominal gust factor of 1.41. The comparison between the modelled and measured gusts is of interest, whereby the model tends to follow the more exposed Cape Moreton data. This is the desired result, given that the Brisbane Airport site is known to suffer attenuation due to the effects of Moreton Bay (J. Callaghan, personal communication).

Overall, this shows a very favourable comparison; the model generally following the trend of the better exposed data, and is a good verification of the model’s capabilities.

2.0 1.8 1.6 1.4 1.2 HAT 1.0 0.8 0.6 0.4 0.2

0.0 Seaway AstronomicalSeaway Level Tide mAHD 1 10 100 1000 Return Period y

Figure 6-1: Verification of the generated HAT tidal plane at Gold Coast Seaway

6 It should be noted that this record does not include the peak estimated winds during TC 1954 and Dinah that are reported in Appendix D but discounted as being unreliable.

Storm Tide Study 41/22526/412243 Final Report Addendum

70 SATSIM 10min Mean 60 SATSIM 3sec Gust Brisbane_Aero 3sec 50 Cape Moreton 3sec 40

30

20 +10m Wind Speed +10m Speed m/sWind 10

0 1 10 100 1000 Return Period y

Figure 6-2: Statistical simulation model prediction of regional wind speeds

Storm Tide Study 47 Final Report Addendum

7. Results for Present Climate

7.1 Tropical Cyclone Impacts The results are provided in a series of tables and illustrated in graphical form for selected sites. Table 10 lists the Total Storm Tide results (i.e. including wave setup) for a number of named locations within the SATSIM model domain, which have been broadly grouped into open coast and waterway locations. The spatial location of these sites is separately summarised in Appendix E. Graphical illustrations of selected sites are presented in Figure 7-1 that show comparisons between sites and Figure 7-1 to Figure 7-5 present details of the storm tide components, showing the presence or absence of wave setup influences. Also shown in Figure 7-1 to Figure 7-5 are comparisons with previous studies where available. For example, the early BPA study results (Harper 1985) are overplotted at Jumpinpin, Surfers Paradise and Point Danger. These are reasonably similar to the present estimates, although they included an allowance for extra-tropical impacts. Also shown are the McInnes et al. (2000) total storm tide results for the Coomera River Mouth and the Seaway. These values are similar to the present estimates at Surfers Paradise but significantly higher than the Coomera River values, likely due to that study’s assumption that ocean wave setup levels would penetrate the Broadwater. Finally, the Hardy et al. (2004) tide plus surge results are compared at Surfers Paradise and can be seen to be an almost identical match. Later, Figure 7-6 also summarises the tropical cyclone storm tide contributions at selected regional sites relative to the other extra-tropical contributions that are detailed in the following section.

Storm Tide Study 41/22526/412243 Final Report Addendum

Table 10 Summary of Tropical Cyclone Total Storm Tide Levels for Present Climate

Estimated Return Period of Tropical Cyclone Total Storm Tide Level (including wave setup)

Site 2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

m m m m m m m m AHD AHD AHD AHD AHD AHD AHD AHD m AHD

Swan_Bay 1.1 1.2 1.4 1.5 1.6 1.7 1.7 2.0 2.1

Jumpinpin_Entrance 1.1 1.2 1.4 1.5 1.6 1.7 1.7 2.1 2.1

Couran_Cove_Ocean 1.4 1.5 1.9 2.1 2.2 2.4 2.5 2.9 3.0

Sheraton_Mirage 1.4 1.5 1.8 1.9 2.1 2.2 2.3 2.7 2.8

Main_Beach 1.4 1.5 1.8 2.0 2.1 2.3 2.3 2.7 2.8

Narrow_Neck 1.4 1.5 1.8 2.0 2.1 2.3 2.4 2.8 2.9

Surfers_Paradise 1.4 1.5 1.8 1.9 2.0 2.2 2.3 2.7 2.7

Broadbeach 1.4 1.5 1.9 2.1 2.2 2.4 2.5 3.0 3.1

Mermaid_Beach 1.4 1.5 1.9 2.1 2.3 2.5 2.6 3.1 3.2

Nobby_Beach 1.4 1.5 1.9 2.1 2.2 2.4 2.5 2.9 3.1

Miami_Beach 1.3 1.4 1.7 1.9 2.0 2.2 2.2 2.5 2.6

Burleigh_Heads 1.4 1.4 1.7 1.9 2.0 2.2 2.3 2.7 2.8

Tallebudgera_Ck 1.3 1.4 1.7 1.8 1.9 2.1 2.1 2.5 2.5

Palm_Beach 1.3 1.4 1.7 1.9 2.0 2.1 2.2 2.5 2.6

Currumbin_Point 1.4 1.5 1.8 2.0 2.1 2.4 2.4 2.8 2.9

Tugun 1.4 1.5 1.8 1.9 2.1 2.2 2.3 2.7 2.8

Billinga 1.4 1.5 1.8 2.0 2.1 2.3 2.4 2.7 2.8

Kirra_Beach 1.4 1.5 1.9 2.0 2.2 2.3 2.4 2.8 2.8

Coolangatta 1.4 1.5 1.9 2.1 2.2 2.4 2.4 2.8 2.9

Greenmount 1.5 1.6 1.9 2.1 2.2 2.4 2.5 2.9 3.0

Rainbow_Bay 1.5 1.6 2.0 2.2 2.3 2.5 2.6 3.1 3.2

Point_Danger 1.5 1.6 1.9 2.1 2.3 2.5 2.6 3.0 3.1

Leticia_Spit 1.5 1.6 2.0 2.2 2.4 2.6 2.7 3.1 3.2

Fingal 1.5 1.6 2.0 2.2 2.3 2.5 2.6 3.0 3.1

Rocky_Point 1.3 1.4 1.6 1.7 1.8 2.0 2.1 2.6 2.7

Storm Tide Study 49 Final Report Addendum

Estimated Return Period of Tropical Cyclone Total Storm Tide Level (including wave setup)

Site 2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

m m m m m m m m AHD AHD AHD AHD AHD AHD AHD AHD m AHD

Little_Rocky_Point 1.3 1.3 1.6 1.6 1.7 1.9 1.9 2.4 2.5

Cabbage_Tree_Point 1.3 1.3 1.6 1.6 1.7 1.9 1.9 2.4 2.5

Pimpama_Island 1.3 1.3 1.6 1.6 1.7 1.9 1.9 2.4 2.5

Steiglitz 1.3 1.3 1.6 1.6 1.7 1.9 1.9 2.4 2.5

Cabbage_Tree_Point 1.3 1.3 1.6 1.6 1.7 1.9 1.9 2.3 2.4

Steiglitz 1.3 1.4 1.6 1.6 1.7 1.9 1.9 2.3 2.4

Sandy_Beach 1.3 1.3 1.6 1.6 1.7 1.8 1.9 2.3 2.4

Jacobs_Well 0.9 1.0 1.1 1.2 1.3 1.5 1.5 1.9 1.9

Couran 0.8 0.9 1.0 1.1 1.2 1.4 1.5 1.9 1.9

Couran_Cove 0.8 0.9 1.0 1.1 1.2 1.4 1.5 1.9 1.9

Sovereign_Islands 0.9 0.9 1.0 1.1 1.1 1.3 1.3 1.5 1.6

Paradise_Point 0.9 0.9 1.0 1.1 1.1 1.3 1.3 1.5 1.6

Currigee 0.9 0.9 1.0 1.1 1.2 1.3 1.3 1.6 1.6

Runaway_Bay 0.9 0.9 1.0 1.1 1.2 1.3 1.4 1.6 1.7

Lands_End 0.9 0.9 1.1 1.1 1.2 1.3 1.4 1.6 1.7

Porpoise_Point 0.9 0.9 1.1 1.2 1.2 1.3 1.4 1.6 1.7

Nerang_Head 1.0 1.0 1.1 1.2 1.3 1.4 1.5 1.7 1.7

Gold_Coast_Seaway 0.8 0.9 1.0 1.1 1.1 1.2 1.3 1.5 1.5

Wave_Break_Island 0.9 1.0 1.1 1.2 1.2 1.4 1.4 1.7 1.7

The_Broadwater 0.9 0.9 1.1 1.1 1.2 1.4 1.4 1.7 1.8 (The_Spit)

Seaworld 0.9 0.9 1.1 1.1 1.2 1.4 1.5 1.8 1.8

Southport 0.9 0.9 1.1 1.1 1.2 1.4 1.5 1.8 1.9

MOB_Tide_Gauge 0.9 0.9 1.0 1.1 1.2 1.4 1.5 1.8 1.8

Koureyabba 1.3 1.3 1.6 1.6 1.7 1.8 1.9 2.3 2.3

Jumpinpin 0.7 0.7 0.8 0.9 0.9 1.0 1.1 1.3 1.3

Labrador 0.9 0.9 1.1 1.1 1.2 1.4 1.4 1.7 1.8

Storm Tide Study 41/22526/412243 Final Report Addendum

Estimated Return Period of Tropical Cyclone Total Storm Tide Level (including wave setup)

Site 2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

m m m m m m m m AHD AHD AHD AHD AHD AHD AHD AHD m AHD

Nerang_R._Bundall 0.6 0.6 0.8 0.8 0.9 1.1 1.1 1.4 1.5 (tidal_station)

N3_Royal_Pines 0.6 0.6 0.8 0.8 0.9 1.1 1.1 1.4 1.5 Harbour

N2_Campbell 0.6 0.6 0.8 0.8 0.9 1.1 1.1 1.4 1.4 St_Sorrento

N1_Evandale 0.6 0.6 0.8 0.8 0.9 1.1 1.1 1.4 1.4

B11_Nerang_Mouth 0.8 0.9 1.0 1.1 1.1 1.3 1.3 1.6 1.6

C4_Monterey_Keys 0.7 0.7 0.9 0.9 1.0 1.1 1.1 1.3 1.4 (Saltwater_Creek)

C1_Coomera_River 0.7 0.8 0.9 1.0 1.1 1.3 1.3 1.7 1.8 (North_Arm)

C6_Brygon_Creek 0.7 0.7 0.9 0.9 1.0 1.1 1.1 1.4 1.4

C2_Marine_Precinct 0.7 0.7 0.9 0.9 1.0 1.1 1.1 1.3 1.4 (Coomera)

C5_Coomera_Shores 0.7 0.7 0.9 0.9 1.0 1.1 1.1 1.3 1.4

Coomera_River_Mouth_ 0.7 0.8 0.9 0.9 1.0 1.1 1.1 1.3 1.4 (south)

Coombabah_Lake 0.4 0.5 0.6 0.6 0.7 0.8 0.8 1.1 1.1

L2_Logan_River_289 1.1 1.2 1.4 1.5 1.7 1.9 2.0 2.7 2.9 Rotary_Park

L1_Logan_River_Boat_ 1.1 1.2 1.4 1.5 1.7 1.9 2.0 2.7 2.9 Hire_Jetty

B3_Logan_River_Mouth 1.1 1.2 1.4 1.5 1.7 1.9 2.0 2.6 2.8

B6_Pimpama 0.6 0.6 0.7 0.9 1.0 1.2 1.2 1.6 1.7

TALLEBUDGERA 0.8 0.8 0.9 1.0 1.0 1.1 1.2 1.4 1.4 Schuster Park

TALLEBUDGERA Elanora WQ Control 0.8 0.8 1.0 1.0 1.0 1.1 1.2 1.4 1.4 Centre

CURRUMBIN Elanora 0.8 0.9 1.0 1.1 1.1 1.2 1.2 1.4 1.5 Community Center

Storm Tide Study 51 Final Report Addendum

4.0

Jumpinpin_Entrance 3.5 Couran_Cove_Ocean Surfers_Paradise 3.0 Point_Danger

2.5

2.0

1.5 HAT Snapper Rocks

1.0 TotalStorm Tide Level AHD m 0.5

0.0 10 100 1000 10000 Return Period y

4.0

3.5 B3_Logan_River_Mouth

Coomera_River_Mouth_(south) 3.0 Gold_Coast_Seaway

2.5 CURRUMBIN

2.0

1.5 HAT Seaway

1.0 TotalStorm Tide Level AHD m 0.5

0.0 10 100 1000 10000 Return Period y

Figure 7-1: Estimated tropical cyclone total storm tide levels for selected sites.

Storm Tide Study 41/22526/412243 Final Report Addendum

Total Storm Tide m AHD 4.0 8.0 Tide+Surge m AHD Jumpinpin_Entrance Surge m MSL 3.5 7.0 Setup m

3.0 Total Wave Runup m AHD 6.0 BPA 1985 Tide+Surge m AHD 2.5 Hs m 5.0

2.0 4.0

1.5 3.0 WaterLevel m HAT

1.0 2.0 SIgnificantWave Height m Hs

0.5 1.0

0.0 0.0 10 100 1000 10000 Return Period y

4.0 2.0

Total Storm Tide m AHD B3_Logan_River_Mouth 1.8 3.5 Tide+Surge m AHD

Surge m MSL 1.6 3.0 Setup m 1.4 Total Wave Runup m AHD 2.5 Hs m 1.2 2.0 1.0

HAT 0.8

1.5 WaterLevel m 0.6 1.0

0.4 SIgnificantWave Height m Hs 0.5 0.2

0.0 0.0 10 100 1000 10000 Return Period y

Figure 7-2: Estimated tropical cyclone storm tide components for selected sites.

Storm Tide Study 53 Final Report Addendum

Total Storm Tide m AHD

4.0 Tide+Surge m AHD 8.0

Surge m MSL Couran_Cove_Ocean 3.5 7.0 Setup m

3.0 Total Wave Runup m AHD 6.0 Hs m 2.5 5.0

2.0 4.0

1.5 3.0 Water Level WaterLevel m HAT

1.0 2.0 SIgnificant SIgnificant Wave Height Hsm 0.5 1.0

0.0 0.0 10 100 1000 10000 Return Period y

4.0 Total Storm Tide m AHD 2.0 Tide+Surge m AHD Coomera_River_Mouth_(south) 1.8 3.5 Surge m MSL Setup m 1.6 3.0 Total Wave Runup m AHD McInnes et al. 2000 (Total) 1.4 2.5 Hs m 1.2

2.0 1.0

0.8

1.5 Water Level WaterLevel m HAT 0.6 1.0

0.4 SIgnificant SIgnificant Wave Height Hsm 0.5 0.2

0.0 0.0 10 100 1000 10000 Return Period y

Figure 7-3: Estimated tropical cyclone storm tide components for selected sites.

Storm Tide Study 41/22526/412243 Final Report Addendum

Total Storm Tide m AHD 4.0 Tide+Surge m AHD 8.0 Surge m MSL Surfers_Paradise Setup m 3.5 7.0 QCC Tide + Surge m AHD Total Wave Runup m AHD 3.0 BPA 1985 Tide+Surge m AHD 6.0 McInnes et al. 2000 (Seaw ay Total) 2.5 Hs m 5.0

2.0 4.0

1.5 3.0 Water Level Water m HAT

1.0 2.0 SIgnificant HsWave Height m 0.5 1.0

0.0 0.0 10 100 1000 10000

Return Period y

4.0 Total Storm Tide m AHD 2.0 Tide+Surge m AHD Gold_Coast_Seaway 1.8 3.5 Surge m MSL Setup m 1.6 3.0 Total Wave Runup m AHD McInnes et al. 2000 (Seaw ay Total) 1.4 2.5 Hs m 1.2

2.0 1.0

0.8

1.5 Water Level Water m HAT 0.6 1.0

0.4 SIgnificant HsWave Height m 0.5 0.2

0.0 0.0 10 100 1000 10000

Return Period y

Figure 7-4: Estimated tropical cyclone storm components for selected sites.

Storm Tide Study 55 Final Report Addendum

Total Storm Tide m AHD 4.0 Tide+Surge m AHD 8.0 Surge m MSL Point_Danger 3.5 Setup m 7.0 BPA 1985 Tide + Surge m AHD 3.0 Total Wave Runup m AHD 6.0 Hs m 2.5 5.0

2.0 4.0

1.5 3.0 Water Level Water m HAT

1.0 2.0 SIgnificant HsWave Height m

0.5 1.0

0.0 0.0 10 100 1000 10000 Return Period y

4.0 2.0 Total Storm Tide m AHD CURRUMBIN 1.8 3.5 Tide+Surge m AHD Surge m MSL 1.6 3.0 Setup m 1.4 2.5 Total Wave Runup Level m AHD 1.2 Hs m 2.0 1.0

0.8

1.5 Water Level Water m HAT 0.6 1.0

0.4 SIgnificant HsWave Height m 0.5 0.2

0.0 0.0 10 100 1000 10000 Return Period y

Figure 7-5: Estimated tropical cyclone storm tide components for selected sites.

Storm Tide Study 41/22526/412243 Final Report Addendum

7.2 Extra-Tropical and Remote Tropical Cyclone Impacts The extra-tropical storm tide results derive from the tidal residual re-sampling using the tide gauge references summarised in Table 11 below. The estuarine site estimates are applicable to locations in their general vicinity, while the open coast site is deemed representative of the whole Gold Coast region. The extra-tropical storm tide response for each of these four exposures is summarised in the middle panel of Figure 7-6.

Table 11 Tide gauge references for the extra-tropical simulations

The Broadwater Coomera River Open Coast Logan River Mouth Mouth

Tide Gauge Seaway Seaway Offshore Tweed Brisbane Bar Residuals

Constituents Seaway Seaway Offshore Tweed Data-derived for Logan River Mouth

Range Ratios n/a Runaway Bay n/a n/a

7.3 Combined Climate Storm Tide Impacts The tropical cyclone and extra-tropical Tide plus Surge elevations are simply linearly added in probability space to obtain the combined impacts. This is possible because the events are considered independent in the analysis context. For example, if the 1.2 m AHD elevation has a return period of 100 y (0.01 p.a.) at the site due to tropical cyclone effects and a 100 y return period also for extra-tropical effects, then the combined probability is approximated by 0.01+0.01=0.02 p.a.7, which results in the 1.2 m actually being equivalent to a 50 y return period.

Figure 7-6 illustrates this process by summarising the tropical cyclone (top), extra-tropical (middle) and the resulting combined (bottom) Surge plus Tide contributions at the selected regional sites. It can be seen that the extra-tropical impacts dominate the Seaway and Coomera Mouth sites for all return periods, and the tropical cyclone impacts only begin to affect the open coast Surfers Paradise site at the 10,000 ARI. The Logan River Mouth site is tropical cyclone affected beyond the 500 y ARI due to the influence of Moreton Bay, which is capable of generating more significant wind setup than the open coast environment and this influence is present in the southern reaches of the bay.

Table 12 provides the estimated combined Total Storm Tide levels (including wave setup) for all the modelled sites. This has been formed by applying the applicable extra-tropical return period curve derived for the reference sites in Table 11 with the tropical return period curves from Table 10. It can be noted that all open coast sites have been assigned the same wave setup and extra-tropical Tide plus Surge levels. No wave setup has been applied to any of the estuarine sites, except for 50% of the open coast setup at Tallebudgera and Currumbin Creeks.

7 The formula is 1/Rtot = 1/R1 + 1/R2 – 1/(R1R2)

Storm Tide Study 57 Final Report Addendum

2.8 2.6 Gold_Coast_Seaway 2.4 Logan_River_Mouth 2.2 Surfers_Paradise 2.0 Coomera_River_South 1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 Tropical Cyclone TropicalCyclone + m Surge TideAHD 0.2 0.0 1 10 100 1000 10000 Return Period y

2.8 2.6 Gold_Coast_Seaway 2.4 Logan_River_Mouth 2.2 Surfers_Paradise 2 Coomera_River_South 1.8 1.6 1.4 1.2 1

Tropical Tide plus m Surge AHD 0.8 - 0.6

Extra 0.4 0.2 0 1 10 100 1000 10000

Return Period y

3.2

3.0 Gold_Coast_Seaway 2.8 Logan_River_Mouth 2.6 Surfers_Paradise 2.4 2.2 Coomera_River_South 2.0 1.8 1.6 1.4 1.2 1.0 0.8

Combined Tide+ Surge AHD m 0.6 0.4 0.2 0.0 1 10 100 1000 10000

Return Period y

Figure 7-6: Combining tropical and extra-tropical tide plus surge levels for present climate.

Storm Tide Study 41/22526/412243 Final Report Addendum

Table 12 Summary of combined Tropical Cyclone and Extra-Tropical Total Storm Tide Levels for Present Climate

Estimated Return Period of Combined Total Storm Tide Level (including wave setup)

Site 2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

m m m m m m m m AHD AHD AHD AHD AHD AHD AHD AHD m AHD

Swan_Bay 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Jumpinpin_Entrance 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Couran_Cove_Ocean 1.8 2.0 2.3 2.4 2.5 2.6 2.7 2.8 3.0

Sheraton_Mirage 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Main_Beach 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Narrow_Neck 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.9

Surfers_Paradise 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Broadbeach 1.8 2.0 2.3 2.4 2.5 2.6 2.7 2.8 3.1

Mermaid_Beach 1.8 2.0 2.3 2.4 2.5 2.6 2.7 2.8 3.2

Nobby_Beach 1.8 2.0 2.3 2.4 2.5 2.6 2.7 2.8 3.1

Miami_Beach 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Burleigh_Heads 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Tallebudgera_Ck 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Palm_Beach 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Currumbin_Point 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.8 2.9

Tugun 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Billinga 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Kirra_Beach 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.7 2.8

Coolangatta 1.8 2.0 2.3 2.4 2.5 2.6 2.6 2.8 3.0

Greenmount 1.8 2.0 2.3 2.4 2.5 2.6 2.7 2.8 3.0

Rainbow_Bay 1.8 2.0 2.3 2.5 2.5 2.6 2.7 2.9 3.2

Point_Danger 1.8 2.0 2.3 2.4 2.5 2.6 2.7 2.9 3.1

Leticia_Spit 1.8 2.0 2.3 2.5 2.5 2.6 2.7 2.9 3.3

Storm Tide Study 59 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level (including wave setup)

Site 2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

m m m m m m m m AHD AHD AHD AHD AHD AHD AHD AHD m AHD

Fingal 1.8 2.0 2.3 2.5 2.5 2.6 2.7 2.9 3.1

Rocky_Point 1.6 1.7 1.8 1.9 2.0 2.0 2.1 2.4 2.7

Little_Rocky_Point 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.2 2.5

Cabbage_Tree_Point 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.2 2.5

Pimpama_Island 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.2 2.5

Steiglitz 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.2 2.5

Cabbage_Tree_Point 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.2 2.4

Steiglitz 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.2 2.4

Sandy_Beach 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.2 2.4

Jacobs_Well 1.6 1.7 1.8 1.9 1.9 2.0 2.0 2.1 2.3

Couran 1.0 1.1 1.3 1.4 1.4 1.5 1.6 1.7 2.0

Couran_Cove 1.0 1.1 1.3 1.4 1.4 1.5 1.6 1.7 2.0

Sovereign_Islands 1.0 1.1 1.2 1.4 1.4 1.5 1.6 1.7 1.8

Paradise_Point 1.0 1.1 1.2 1.4 1.4 1.5 1.6 1.7 1.8

Currigee 1.0 1.1 1.2 1.4 1.4 1.5 1.6 1.7 1.8

Runaway_Bay 1.0 1.1 1.2 1.4 1.4 1.5 1.6 1.7 1.8

Lands_End 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9

Porpoise_Point 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9

Nerang_Head 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9

Gold_Coast_Seaway 1.2 1.3 1.4 1.5 1.5 1.6 1.7 1.8 1.9

Wave_Break_Island 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9

The_Broadwater 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9 (The_Spit)

Seaworld 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9

Southport 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9

MOB_Tide_Gauge 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9

Storm Tide Study 41/22526/412243 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level (including wave setup)

Site 2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

m m m m m m m m AHD AHD AHD AHD AHD AHD AHD AHD m AHD

Koureyabba 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.2 2.4

Jumpinpin 1.2 1.3 1.3 1.4 1.4 1.4 1.5 1.5 1.5

Labrador 1.2 1.3 1.4 1.5 1.6 1.6 1.7 1.8 1.9

Nerang_R._Bundall 1.2 1.3 1.4 1.5 1.5 1.6 1.7 1.8 1.9 (tidal_station)

N3_Royal_Pines 1.2 1.3 1.4 1.5 1.5 1.6 1.7 1.8 1.9 Harbour

N2_Campbell 1.2 1.3 1.4 1.5 1.5 1.6 1.7 1.8 1.9 St_Sorrento

N1_Evandale 1.2 1.3 1.4 1.5 1.5 1.6 1.7 1.8 1.9

B11_Nerang_Mouth 1.2 1.3 1.4 1.5 1.5 1.6 1.7 1.8 1.9

C4_Monterey_Keys 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 (Saltwater_Creek)

C1_Coomera_River 1.0 1.1 1.2 1.4 1.4 1.5 1.6 1.7 1.8 (North_Arm)

C6_Brygon_Creek 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8

C2_Marine_Precinct 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 (Coomera)

C5_Coomera_Shores 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8

Coomera_River_Mouth_ 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 (south)

Coombabah_Lake 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8

L2_Logan_River_289 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.4 2.9 Rotary_Park

L1_Logan_River_Boat_ 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.5 2.9 Hire_Jetty

B3_Logan_River_Mouth 1.6 1.7 1.8 1.9 1.9 2.0 2.1 2.4 2.8

B6_Pimpama 1.6 1.7 1.8 1.9 1.9 2.0 2.0 2.1 2.3

TALLEBUDGERA 1.5 1.6 1.8 1.9 2.0 2.0 2.0 2.1 2.2 Schuster Park

Storm Tide Study 61 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level (including wave setup)

Site 2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

m m m m m m m m AHD AHD AHD AHD AHD AHD AHD AHD m AHD

TALLEBUDGERA 1.5 1.6 1.8 1.9 2.0 2.0 2.0 2.1 2.2 Elanora WQ Control Centre

CURRUMBIN Elanora 1.5 1.6 1.8 1.9 2.0 2.0 2.0 2.1 2.2 Community Center

7.4 Design Water Level Hydrographs For the purpose of more detailed GCCC inhouse modelling of storm tide propagation throughout the waterways and river systems, design hydrographs for tailwater levels may be adapted from the GHD (2011) study results to match the updated levels presented here or simply factoring of the tidal signal may be appropriate.

Storm Tide Study 41/22526/412243 Final Report Addendum

8. Possible Impacts of Climate Change

8.1 The Enhanced Greenhouse Effect Over the past two decades there has been a growing awareness of the potential impacts that human-induced global climate change may have, and especially its possible effects on the coastal environment (NCCOE 2004). The estimated return periods for storm tide levels in the present study rely on the assumption that the natural environment, although highly variable, remains statistically static and that probability distributions for tropical cyclones and sea level are unchanging with the passage of time. However, the proven rise in atmospheric carbon dioxide levels and an increasing trend in mean air temperatures and rising sea levels points to the likelihood of the Earth being subject to an enhanced "greenhouse" effect, which means that these static assumptions will be in error to some extent. Consequently, some consideration of the possible impacts of future climate change on modifying the present storm tide estimates is addressed in this section. The effect that these possible climate changes might have already had on the past historical data is not able to be quantified and is therefore neglected at this time.

8.1.1 Potential Sea Level Rise Global sea levels are expected to rise as a consequence of enhanced greenhouse warming of the earth (IPCC 2007). The observed rate of global average sea level rise measured by TOPEX/Poseidon satellite altimetry during the decade 1993 to 2003 was 3.1 ± 0.7 mm p.a., although there are large regional differences. This is close to the estimated total of 2.8 ± 0.7 mm p.a. for the following climate-related contributions, in order of decreasing contribution:

 An accelerating thermal expansion throughout the 21st century;

 The melting of glaciers;

 Retreat of the Greenland ice shelf; and

 Antarctic ice losses. The official projections of global average sea level rise by 2100 are in the range 0.18 to 0.59 m (IPCC 2007, CSIRO 2007; 5% to 95% confidence levels for six greenhouse gas emission scenarios). These represent increases in the lower limit of about 0.1 m over the previous IPCC assessment reductions at the higher limit due to a separate consideration of ice flow uncertainties. However, making allowance for the ice flow uncertainties increases the upper limit by 2100 to levels only slightly lower than those previously adopted (e.g. NCCOE 2004). Accordingly the trends displayed in Figure 8-1 based on earlier IPCC assessments are consistent with the recommendations of DERM (2009). This also allows comparison with other recently published QCC levels (Hardy et al. 2004, Harper 2004), whereby the upper level estimates at 2050 are considered and then extended to 2060 and 2100 for completeness. Although the year 2100 is normally quoted, it is important to note that if greenhouse gas concentrations were stabilised (even at present levels), sea level is nonetheless predicted to continue to rise for hundreds of years.

Storm Tide Study 63 Final Report Addendum

1.0

0.9

0.8

0.7 Central Estimate

0.6 Max and Min

0.5

0.4

0.3

0.2

0.1 Projected Global Average Sea Average Projected Level Rise Global (m) 0.0 1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100 Year

Figure 8-1: Projection of Global Average Sea Level Rise (after NCCOE 2004)

8.1.2 Possible Changes to Extreme Weather Systems Driving Storm Tide (a) Potential changes in Tropical Cyclone Intensity

Although IPCC (2007) does address aspects of future tropical cyclone climatology, this area of research is advancing rapidly and the preferred reference is Knutson et al. (2010), which summarises the status of current research in this area and concludes that there is an agreed likely increase in the Maximum Potential Intensity (MPI) of tropical cyclones as the mean global temperature rises of between +3 to +21% by the year 2100 (between +2 and +11% if expressed as maximum wind speed rather than central pressure deficit). This has been nominally assumed to be reasonably represented by a 10% increase in MPI pressure deficit by 2050 and a 20% increase by 2100. (b) Potential changes in Tropical Cyclone Frequency and Track

Likewise, a Knutson et al. (2010) report that the consensus from many advanced modelling studies is that the global frequency of tropical cyclones will either decrease or remain essentially unchanged. There is an expressed low confidence in some modelling studies that project changes ranging from -6 to -34% globally, and up to ±50% or more in individual basins by 2100. Regarding tracks, there is low confidence in estimates of changed areas of genesis or tracks. Accordingly no changes are adopted here for the year 2060, but a nominal precautionary allowance for a +10% change has been assumed by the year 2100.

(c) Potential Increase in Extra-Tropical Storm System Intensity

Storm Tide Study 41/22526/412243 Final Report Addendum

There is no specific advice available in this regard, although modelling studies have provided conflicting evidence. McInnes et al. (2007) is the most comprehensive climate change assessment available for the NSW coast and utilises outputs from a number of CSIRO climate models, focusing on Wooli and Batemans Bay as indicative coastal environments. Although the study attempted to provide indications of future trends for 2030 and 2070, the results are highly variable across a range of parameters. Taking the higher estimates of change in each case, the study suggested that the 100 y ARI (Average storm surge might increase by as much as 4% by 2030 or 2070 at Wooli, but actually decrease at Batemans Bay by as much as 3% by 2070. Significant wave height estimates were also highly variable between models and sites, with a range of up to 9% at Wooli by 2100 and 32% at Batemans Bay for storms from the SSE direction. Taken as a whole, these analyses are considered to be too variable to be regarded as reliable indicators for the Gold Coast region.

As a more practical alternative, the present analysis considers a nominal increase in extra- tropical generated storm surge magnitude (i.e. residuals) and wave heights of 10% by the year 2100, with a 5% increase applied for 2060.

8.1.3 Tropical Cyclone Climate Change Scenarios Considered for this Study In light of the above scientific projections and assumptions, the climate change scenarios considered in this study are summarised below in Table 13. In regard to these assumptions: (a) A rise in mean sea level (MSL) will also lead to a rise in HAT and the tidal characteristics may also change slightly as a result, but this effect is ignored. Also, although AHD is based on MSL, it is assumed here that the AHD datum will remain where it is now.

(b) An increase in tropical cyclone MPI is not a straightforward concept to apply to the statistical description of individual cyclone central pressure values. The interpretation made here is that the most intense of cyclones may increase their intensity but that not all cyclones will be more intense. The way that this is applied is shown in Figure 8-2, whereby the potential % increase (relative to p) is blended into the present climate description used by the statistical model.

Table 13 Enhanced Greenhouse Scenarios for Tropical Cyclones

Scenario Increase in Increase in Increase in TC Increase in Mean Year Extra-Tropical Frequency of TC Maximum Sea Level Storm Surge Occurrence Potential (DERM 2009) Magnitude Intensity (MPI) Pressure Deficit

% % % m

2060 5 0 10 0.40

2100 10 10 20 0.80

Storm Tide Study 65 Final Report Addendum

1000

990 Present Climate 980 2060 970 2100 960 950 940 930

Central Pressure pcCentral (hPa) 920 910 900 1 10 100 1000 Return Period (y)

Figure 8-2: Assumed Possible Changes in the Intensity of Tropical Cyclones under Future Climate Change Projections within 500 km of the Gold Coast.

8.2 Relative Results of the Climate Change Scenarios To simplify the presentation of the potential impact of climate change, only the increases in estimated storm tide levels are shown for 2060 and 2100. The values below in Table 14 and Table 15 may then be added to those in Table 12 to obtain the absolute estimate of the potential storm tide probability levels relative to present day AHD by the year 2060 or 2100 respectively for combined tropical cyclone and extra-tropical events. The results for the year 2060 show an average increase in total storm tide levels across all sites of interest of about 0.45 m for the 100 y return period and 0.50 m for the 10,000 year. The corresponding averaged results for the year 2100 are increases of about 0.85 m and 0.96 m respectively. The increases for open coast sites vary partly as a result of the dune crest elevation at each location, which can have the effect of limiting the breaking wave set-up component differently under different scenarios. The effect of the static increase in MSL has been subtracted at the base of each table to reveal an estimate of the separate combined effects of the increase in frequency of occurrence of tropical cyclones and their increased MPI, plus the nominal increases in extra-tropical surge and wave magnitudes. These can then be added to, for example, the more “central” estimates of projected MSL rise if required. These show that the storm-related climate changes do not begin to become significant over and above the MSL rise until well beyond the 100 y return period risk level.

Storm Tide Study 41/22526/412243 Final Report Addendum

Table 14 Estimated Increase in combined Tropical Cyclone and Extra-Tropical Total Storm Tide Levels under Climate Change Scenario in the Year 2060

Estimated Return Period of Combined Total Storm Tide Level Difference 2060 (including wave setup)

2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

Site m m m m m m m m m

Swan_Bay 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4

Jumpinpin_Entrance 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4

Couran_Cove_Ocean 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5

Sheraton_Mirage 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5

Main_Beach 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5

Narrow_Neck 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.6

Surfers_Paradise 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5

Broadbeach 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5

Mermaid_Beach 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5

Nobby_Beach 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5

Miami_Beach 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4

Burleigh_Heads 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5

Tallebudgera_Creek 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4

Palm_Beach 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4

Currumbin_Point 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5

Tugun 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5

Billinga 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5

Kirra_Beach 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.6

Coolangatta 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5

Greenmount 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5

Rainbow_Bay 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5 0.5

Point_Danger 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5

Leticia_Spit 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5 0.6

Fingal 0.4 0.4 0.4 0.4 0.4 0.4 0.5 0.5 0.5

Rocky_Point 0.5 0.5 0.5 0.5 0.5 0.6 0.6 0.7 0.7

Storm Tide Study 67 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level Difference 2060 (including wave setup)

2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

Site m m m m m m m m m

Little_Rocky_Point 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.7 0.7

Cabbage_Tree_Point 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.7 0.7

Pimpama_Island 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.7 0.7

Steiglitz 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.7 0.7

Cabbage_Tree_Point 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.6 0.7

Steiglitz 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.6 0.7

Sandy_Beach 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.6 0.7

Jacobs_Well 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Couran 0.5 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5

Couran_Cove 0.5 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5

Sovereign_Islands 0.5 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.6

Paradise_Point 0.4 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.6

Currigee 0.4 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5

Runaway_Bay 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5

Lands_End 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Porpoise_Point 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Nerang_Head 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Gold_Coast_Seaway 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Wave_Break_Island 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

The_Broadwater 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6 (The_Spit)

Seaworld 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Southport 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5

MOB_Tide_Gauge 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Koureyabba 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.7

Jumpinpin 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.5

Labrador 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Storm Tide Study 41/22526/412243 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level Difference 2060 (including wave setup)

2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

Site m m m m m m m m m

Nerang_R._Bundall 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6 (tidal_station)

N3_Royal_Pines 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6 Harbour

N2_Campbell 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6 St_Sorrento

N1_Evandale 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

B11_Nerang_Mouth 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

C4_Monterey_Keys 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6 (Saltwater_Creek)

C1_Coomera_River 0.5 0.4 0.4 0.5 0.5 0.5 0.5 0.5 0.5 (North_Arm)

C6_Brygon_Creek 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

C2_Marine_Precinct 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6 (Coomera)

C5_Coomera_Shores 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

Coomera_River_Mouth 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6 (south)

Coombabah_Lake 0.5 0.4 0.5 0.5 0.5 0.5 0.5 0.5 0.6

L2_Logan_River_289 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.7 0.7 Rotary_Park

L1_Logan_River_Boat_ 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.6 0.7 Hire_Jetty

B3_Logan_River_Mouth 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.7 0.7

B6_Pimpama 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6

TALLEBUDGERA 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 Schuster Park

TALLEBUDGERA Elanora WQ Control 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 Centre

CURRUMBIN Elanora 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.4 0.5 Community Center

Storm Tide Study 69 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level Difference 2060 (including wave setup)

2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

Site m m m m m m m m m

Average = 0.45 0.43 0.44 0.45 0.45 0.46 0.47 0.50 0.55

Less static rise = 0.05 0.03 0.04 0.05 0.05 0.06 0.07 0.10 0.15

Storm Tide Study 41/22526/412243 Final Report Addendum

Table 15 Estimated Increase in combined Tropical Cyclone and Extra-Tropical Total Storm Tide Levels under Climate Change Scenario in the Year 2100

Estimated Return Period of Combined Total Storm Tide Level Difference 2100 (including wave setup)

2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

Site m m m m m m m m m

Swan_Bay 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8

Jumpinpin_Entrance 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8

Couran_Cove_Ocean 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9 0.9

Sheraton_Mirage 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8

Main_Beach 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8

Narrow_Neck 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9

Surfers_Paradise 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9

Broadbeach 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9 0.9

Mermaid_Beach 0.8 0.8 0.8 0.8 0.8 0.9 0.9 1.0 1.0

Nobby_Beach 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9 1.0

Miami_Beach 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8

Burleigh_Heads 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8

Tallebudgera_Creek 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8

Palm_Beach 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8

Currumbin_Point 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9 0.9

Tugun 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9

Billinga 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9

Kirra_Beach 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9

Coolangatta 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9 1.0

Greenmount 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.9 1.0

Rainbow_Bay 0.8 0.8 0.8 0.8 0.8 0.9 0.9 1.0 1.0

Point_Danger 0.8 0.8 0.8 0.9 0.9 0.9 0.9 1.0 1.0

Leticia_Spit 0.8 0.8 0.9 0.9 0.9 0.9 1.0 1.0 1.0

Fingal 0.8 0.8 0.8 0.9 0.9 0.9 0.9 1.0 1.0

Storm Tide Study 71 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level Difference 2100 (including wave setup)

2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

Site m m m m m m m m m

Rocky_Point 0.9 1.0 1.0 1.1 1.1 1.1 1.2 1.3 1.4

Little_Rocky_Point 0.9 1.0 1.0 1.0 1.0 1.1 1.1 1.2 1.3

Cabbage_Tree_Point 0.9 1.0 1.0 1.0 1.0 1.1 1.1 1.2 1.3

Pimpama_Island 0.9 1.0 1.0 1.0 1.0 1.1 1.1 1.2 1.3

Steiglitz 0.9 1.0 1.0 1.0 1.0 1.1 1.1 1.2 1.3

Cabbage_Tree_Point 0.9 1.0 1.0 1.0 1.0 1.1 1.1 1.2 1.3

Steiglitz 0.9 1.0 1.0 1.0 1.0 1.1 1.1 1.2 1.3

Sandy_Beach 0.9 1.0 1.0 1.0 1.0 1.1 1.1 1.2 1.3

Jacobs_Well 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Couran 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 1.0

Couran_Cove 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Sovereign_Islands 0.9 0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9

Paradise_Point 0.8 0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9

Currigee 0.8 0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9

Runaway_Bay 0.8 0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9

Lands_End 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Porpoise_Point 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Nerang_Head 0.9 0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9

Gold_Coast_Seaway 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Wave_Break_Island 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

The_Broadwater 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9 (The_Spit)

Seaworld 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Southport 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

MOB_Tide_Gauge 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Koureyabba 0.9 1.0 1.0 1.0 1.0 1.0 1.1 1.2 1.2

Jumpinpin 0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Labrador 0.9 0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9

Storm Tide Study 41/22526/412243 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level Difference 2100 (including wave setup)

2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

Site m m m m m m m m m

Nerang_R._Bundall 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9 (tidal_station)

N3_Royal_Pines 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9 Harbour

N2_Campbell 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9 St_Sorrento

N1_Evandale 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

B11_Nerang_Mouth 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

C4_Monterey_Keys 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9 (Saltwater_Creek)

C1_Coomera_River 0.9 0.8 0.8 0.9 0.9 0.9 0.9 0.9 0.9 (North_Arm)

C6_Brygon_Creek 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

C2_Marine_Precinct_(C 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9 oomera)

C5_Coomera_Shores 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

Coomera_River_Mouth 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9 (south)

Coombabah_Lake 0.9 0.8 0.9 0.9 0.9 0.9 0.9 0.9 0.9

L2_Logan_River_289 0.9 0.9 0.9 0.9 0.9 1.0 1.1 1.2 1.2 Rotary_Park

L1_Logan_River_Boat 0.9 0.9 0.9 0.9 0.9 1.0 1.1 1.2 1.2 Hire_Jetty

B3_Logan_River_Mouth 0.9 0.9 0.9 0.9 0.9 1.0 1.1 1.2 1.3

B6_Pimpama 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9 0.9

TALLEBUDGERA 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 Schuster Park

TALLEBUDGERA 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 Elanora WQ Control Centre

CURRUMBIN Elanora 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 0.8 Community Center

Storm Tide Study 73 Final Report Addendum

Estimated Return Period of Combined Total Storm Tide Level Difference 2100 (including wave setup)

2 y 5 y 20 y 50 y 100 y 200 y 500 y 2000 y 10000 y

Site m m m m m m m m m

Average = 0.85 0.84 0.86 0.87 0.87 0.88 0.90 0.93 0.96

Less static rise = 0.05 0.04 0.06 0.07 0.07 0.08 0.10 0.13 0.16

Storm Tide Study 41/22526/412243 Final Report Addendum

8.3 Tide plus Surge Climate Change Water Levels As Council’s principal interest in this study is the derivation of tailwater levels to assist fluvial flooding studies, the combined tropical and extra-tropical Tide plus Surge estimates (excluding wave setup) are separately provided here in terms of their absolute levels. These are tabulated in Table 16 through Table 18 for the principal estuarine sites of interest.

It can be seen that the increases in water level attributable to non-SLR effects are relatively small in The Broadwater, amounting to a maximum of 0.13 m by 2100 at the 10,000 y ARI. However in southern Moreton Bay at the Logan River mouth they reach 0.34 m by 2100 at the 10,000 y ARI.

Table 16 Summary of Present and Future Climate Tide plus Surge Ocean Levels for the Gold Coast Seaway

Climate Planning Year

ARI 2010 2060 2100

y m AHD m AHD m AHD

2 1.16 1.64 2.04

5 1.25 1.70 2.10

20 1.37 1.82 2.23

50 1.47 1.93 2.33

100 1.54 2.01 2.41

200 1.62 2.10 2.50

500 1.69 2.19 2.58

2000 1.78 2.30 2.68

10000 1.87 2.40 2.80

Storm Tide Study 75 Final Report Addendum

Table 17 Summary of Present and Future Climate Tide plus Surge Ocean Levels for the Coomera River South

Climate Planning Year

ARI 2010 2060 2100

y m AHD m AHD m AHD

2 0.96 1.40 1.80

5 1.07 1.50 1.90

20 1.22 1.67 2.07

50 1.33 1.80 2.19

100 1.41 1.86 2.26

200 1.50 1.95 2.34

500 1.58 2.05 2.44

2000 1.68 2.16 2.56

10000 1.78 2.29 2.68

Table 18 Summary of Present and Future Climate Tide plus Surge Ocean Levels for the Logan River Mouth

Climate Planning Year

ARI 2010 2060 2100

y m AHD m AHD m AHD

2 1.60 2.10 2.50

5 1.69 2.16 2.58

20 1.80 2.28 2.70

50 1.88 2.37 2.80

100 1.94 2.44 2.90

200 2.00 2.52 3.05

500 2.10 2.67 3.30

2000 2.36 3.02 3.60

10000 2.78 3.45 3.92

Storm Tide Study 41/22526/412243 Final Report Addendum

9. Conclusions

A comprehensive assessment has been made of the storm tide risk in the Gold Coast coastal region due to the possible effects of tropical cyclones and other large scale weather systems. Historical data analysis, detailed numerical hydrodynamic modelling and state-of-the-art statistical modelling have been combined to provide enhanced understanding, assessment and management of the risk posed by storm tide to population, housing and infrastructure in the region. The analyses are based on a number of assumptions and the possible impacts of these are discussed in the relevant sections of the report.

The study provides essential information that can be used to help mitigate the effects of extreme storm tide through the planning process and also delivers design storm tide tailwater levels for use in fluvial flooding studies.

Within the limits of the available data, resources and timescale, the study has considered:

 The long-term historical record of tropical cyclones in the region, including preferred tracks, speeds, directions and intensities;

 The spatial and temporal characteristics of storm surges generated by tropical cyclones interacting with the complex coastal features;

 The broad-scale ocean response of extra-tropical and remote tropical cyclone influences as captured by the regional tide gauge records;

 Associated extreme waves and estimated breaking wave set-up levels at the coastline;

 The astronomical tide, which varies considerably throughout the study region. The acceptable accuracy of the various models has been confirmed by comparison with available historical wind and storm surge data and also the published tide tables. Demonstration hindcasts of TC Dinah and the “1954” cyclone have illustrated the ability of the various model components applied in this study to reasonably reproduce the qualitative and some quantitative impacts of these historical storm events, although there are significant data limitations. While the principal predictions of extreme storm tide levels has been undertaken within the concept of “present” climate, additional guidance on the possible influence of an “enhanced greenhouse” climate by the year 2100 has also been included based on current scientific opinion.

Guidance has also been provided on the interpretation of storm tide return periods and how such information might be used for decision making. It is emphasised that the provided Tide plus Surge estimates are considered inherently more reliable than those that include wave setup or wave runup. This is due to the likely wave interaction with very localised small scale dynamic coastal features.

It is concluded that storm tide levels in the Gold Coast region, including the waterways and the open coast, are currently dominated by extra-tropical and remote tropical cyclone influences rather than the threat of an intense land-falling tropical cyclone. The projected influences of future climate change do not change this situation, although rising sea level itself remains a significant threat to the coastal margins.

Storm Tide Study 77 Final Report Addendum

10. References

Allen M.A. and Callaghan J. (2000) Extreme wave conditions for the south east Queensland coastal region. Environment Technical Report 32, Environmental Protection Agency, Qld. Basher, R. E. and Zheng, X. (1995) Tropical cyclones in the Southwest Pacific: spatial patterns and relationships to southern oscillation and sea surface temperature, Journal of Climate, Vol 8, May, 1249-1260. Benjamin and Cornell (1970) Probability, statistics and decision for civil engineers, McGraw-Hill.

Callaghan J. (1986) Subtropical cyclogenesis off Australia’s east coast. Proc. 2nd Intl Conf Southern Hemisphere Meteorology, American Meteorological Society, Wellington, NZ, Dec, 38- 41.

CSIRO (2007) Climate change in Australia – technical report. In association with the Bureau of Meteorology and the Australian Greenhouse Office. 148pp. DERM (2009) Draft Queensland Coastal Plan, Draft State Planning Policy Coastal Protection. Department of Environment and Resource Management, Aug, 56pp.

Dvorak V.F. (1984) Tropical cyclone intensity analysis using satellite data. NOAA Technical Report NESDIS 11, 45pp.

GHD Pty Ltd (2011) Storm tide study: Report. Prepared for Gold Coast City Council, March, 237 pp. GHD/SEA (2007) Townsville-Thuringowa Storm Tide Study. Prepared for Townsville and Thuringowa City Councils, April, 210pp

Ginger J. and Harper B. (2004) Wind velocity field at Cape Moreton. Proc 11th AWES Workshop, Australian Wind Engineering Society, Darwin, Jun.

Hanslow D.J. and Nielsen P. (1993) Shoreline setup on natural beaches. J Coastal Res, Special Issue 15, 1-10. Hardy T.A., Mason L.B. and Astorquia A. (2004) Queensland climate change and community vulnerability to tropical cyclones – ocean hazards assessment: Stage 3 - the frequency of surge plus tide during tropical cyclones for selected open coast locations along the Queensland East coast, Queensland Government, Jul, 61pp.

Harper B.A. (1985) Storm tide statistics – Surfers Paradise, Blain Bremner and Williams Pty Ltd, Beach Protection Authority of Queensland, January, 50 pp. Harper, B. A. (1999) Storm tide threat in Queensland: history, prediction and relative risks. Conservation Tech Rep No. 10, RE 208, Dept of Environment and Heritage, Jan.

Harper B A (Ed.) (2001a) Queensland climate change and community vulnerability to tropical cyclones - ocean hazards assessment - Stage 1, Report prep by Systems Engineering Australia Pty Ltd in association with James Cook University Marine Modelling Unit, Queensland Government, March, 375pp.

Storm Tide Study 41/22526/412243 Final Report Addendum

Harper B A (2001b) Queensland climate change and community vulnerability to tropical cyclones - ocean hazards assessment - Stage 1A – Operational Manual, Report prep by Systems Engineering Australia Pty Ltd, Queensland Government, March, 75pp.

Harper B.A. (2001c) Natural Hazards and the Risks they Pose to South-East Queensland (Chapter contributions on Flood, Tropical Cyclones, East Coast Lows and Severe Storms). Edited by Granger K. and Hayne M., Geoscience Australia in conjunction with the Bureau of Meteorology, Aug.

Harper B.A. (2004) Queensland climate change and community vulnerability to tropical cyclones – ocean hazards assessment: Synthesis Report, Queensland Government, Aug, 38pp.

Harper, B. A. and Holland, G. J. (1999) An updated parametric model of the tropical cyclone. Proc. 23rd Conf. Hurricanes and Tropical Meteorology, American Meteorological Society, Dallas, Texas, 10-15 Jan, 1999.

Harper, B. A. and McMonagle, C. J. (1985) Storm tide statistics - methodology, Blain Bremner and Williams Pty Ltd, Beach Protection Authority of Queensland, January, 120 pp.

Harper, B.A., Mason, L.B. and Bode, L. (1993) Tropical cyclone Orson - a severe test for modelling, Proc. 11th Australian Conf. on Coastal and Ocean Engin., Institution of Engineers Australia, Townsville, Aug, 59-64.

Harper B.A., Sobey R.J. and Stark K.P. (1977) Numerical simulation of tropical cyclone storm surge along the Queensland coast - Part X: Gold Coast, Department of Civil and Systems Engineering, James Cook University, Nov, 90pp.

Harper B.A., Stroud S.A., McCormack M. and West S. (2008) A review of historical tropical cyclone intensity in north-western Australia and implications for climate change trend analysis. Australian Meteorological Magazine, Vol. 57, No. 2, June, 121-141.

Holland G. J. (1980) An analytic model of the wind and pressure profiles in hurricanes, Monthly Weather Review, Vol 108, No.8, Aug, pp 1212-1218.

Holland G. J. (1981) On the quality of the Australian tropical cyclone data base, Aust Met Mag, Vol.29, No.4, Dec, pp. 169-181.

Holland G. J. (1997a) Personal communication. Holland G.J. (1997b). The maximum potential intensity of tropical cyclones. J. Atmos. Sci., 54, Nov, 2519-2541.

Holland G.J., Lynch A.H. and Leslie L.M. (1987) Australian east coast cyclones. Part I: Synoptic Overview and Case Study. Monthly Weather Review, 115, 3024-3036.

Hopkins L.C. and Holland G.J. (1997) Australian heavy-rain days and associated east coast cyclones: 1958-92. Journal of Climate, 10, April, 621-635.

IPCC (2007) Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 996 pp.

Storm Tide Study 79 Final Report Addendum

Kemp R.L and Douglas D.A. (1981) A coastal storm climatology for engineers. Proc 5th Australian Conference on Coastal and Ocean Engineering, IEAust, 230-233.

Knutson T.R., McBride J.L., Chan J., Emanuel K., Holland G., Landsea C., Held I., Kossin J.P., Srivastava A.K. and Sugi M. (2010) Tropical cyclones and climate change. Nature Geoscience, 3, 15 –163.

McInnes K., Macadam I., Abbs D., O’Farrell S.P., O’Grady J., and Ranasinghe R., (2007) Projected changes in climatological forcing conditions for coastal erosion in NSW. Tech Report, CSIRO, Melbourne, Australia, 38pp.

MSQ (2012) Semi-diurnal Tidal Planes. Published online.

NCCOE (2004) Guidelines for responding to the effects of climate change in coastal and ocean engineering. National Committee on Coastal and Ocean Engineering, Engineers Australia, EA Books, Barton, ACT, 76 pp.

Nicholls N. (1992) Historical El Nino / Southern Oscillation variability in the Australian region. In Diaz, H.F. and Margraf, V. (Eds), El Nino, historical and paleoclimate aspects of the Southern Oscillation. Cambridge University Press, Cambridge, UK pp151-174.

Nielsen P. and Hanslow D.J. (1991) Wave runup distributions on natural beaches. J Coastal Res, Vol 7, No 4, pp 1139-1152.

Pielke R.A. (Jr) and Landsea C.W. (1999) La Niña, El Niño, and Atlantic hurricane damages in the United States, Bulletin of the American Meteorological Society, 80, 2027-2033. Power, S., Casey, T., Folland, C., Colman, A. and Mehta, V. (1999) Inter-decadal modulation of the impact of ENSO on Australia. Climate Dynamics, 15, 319-324.

PWD (1985), Elevated ocean levels, storms affecting NSW coast 1880 – 1980. Report prepared by Blain, Bremner and Williams Pty Ltd for NSW Public Works Department, Report No. 85041, Dec.

QDoT (2010) Tide tables for Queensland with notes on boating. Department of Transport, Brisbane.

SEA (2002) Parametric tropical cyclone wave model for Hervey Bay and South East Queensland. Queensland Climate Change and Community Vulnerability to Tropical Cyclones: Ocean Hazards Assessment - Stage 2. Prepared by Systems Engineering Australia Pty Ltd for JCU Marine Modelling Unit and the Bureau of Meteorology, March.

Sobey R.J., Harper B.A. and Stark K.P. (1977) Numerical simulation of tropical cyclone storm surge along the Queensland coast, Department of Civil and Systems Engineering, Research Bulletin No. CS14, James Cook University, May, 300pp.

WMO (2006) Statement on tropical cyclones and climate change. Outcome of the Sixth International Workshop on Tropical Cyclones (IWTC-VI), Costa Rica, WMO-TCP, Geneva, Nov, 13pp.

Storm Tide Study 41/22526/412243 Final Report Addendum

Appendix A Technical Note on the Interpretation of Statistical Return Periods

Storm Tide Study 1 Final Report Addendum

A Note on the Interpretation of Statistical Return Periods This study has presented its analyses of risk in terms of the so-called Return Period (or average recurrence interval ARI). The return period is the “average” number of years between successive events of the same or greater magnitude. For example, if the 100-year return period storm tide level is 3.0 m AHD then on average, a 3.0 m AHD level storm tide or greater will occur due to a single event once every 100 years, but sometimes it may occur more or less frequently than 100 years. It is important to note that in any “N”-year period, the “N”-year return period event has a 64% chance of being equalled or exceeded. This means that the 3.0 m storm tide has a better-than-even chance of being exceeded by the end of any 100-year period. If the 100-year event were to occur, then there is still a finite possibility that it could occur again soon, even in the same year, or that the 1000 year event could occur, for example, next year. Clearly if such multiple events continue unchecked then the basis for the estimate of the 100 year event might then need to be questioned, but statistically this type of behaviour can be expected.

A more consistent way of considering the above (NCCOE 2003) is to include the concepts of “design life” and “encounter probability” which, when linked with the return period, provide better insight into the problem and can better assist management risk decision making. These various elements are linked by the following formula (Borgman 1963): Tr = - L / ln [1 - p ] where p = encounter probability 0  1

L = the design life (years)

Tr = the return period (years)

This equation describes the complete continuum of risk when considering the prospect of at least one event of interest occurring. More complex equations describe other possibilities such as the risk of only two events in a given period or only one event occurring. Figure A-1 illustrates the above equation graphically. It presents the variation in probability of at least one event occurring (the encounter probability) versus the period of time considered (the design life). The intersection of any of these chosen variables leads to a particular return period and a selection of common return periods is indicated. For example, this shows that the 200-year return period has a 40% chance of being equalled or exceeded in any 100-year period.

The level of risk acceptable in any situation is necessarily a corporate or business decision. Table A-1, based on Figure A-1, is provided to assist in this decision making process by showing a selection of risk options. Using Table A-1, combinations of design life and a comfortable risk of occurrence over that design life can be used to yield the appropriate return period to consider. For example, accepting a 5% chance of occurrence in a design life of 50 years means that the 1000-year return period event should be considered. A similar level of risk is represented by a 1% chance in 10 years. By comparison, the 100 year return period is equivalent to about a 10% chance in 10 years. AS1170.2 (Standards Australia 1989), for example, dictates a 5% chance in 1000-year criteria or the 1000-year return period as the minimum risk level for wind speed loadings on engineered structures.

Storm Tide Study 1 Final Report Addendum

References NCCOE (2003) Coastal Engineering Guidelines for Working with the Australian Coast in an Ecologically Sustainable Way. The National Committee on Coastal and Ocean Engineering, ENGINEERS AUSTRALIA, Canberra, in draft.

(1963) Risk Criteria. Journal of the Waterway, Port, Coastal and Ocean Division, ASCE, Vol 89, No. WW3, Aug, 1 - 35.

Standards Australia (1989) AS1170.2 - 1989 SAA Loading Code. Part 2: Wind Loads, 96pp.

100

10 y 10 20 y 50 y 100 y 200 y 1 500 y 1000 y 2000 y 5000 y

Encounter Probability (%) Probability Encounter 10000 y 0.1 Equivalent Average Recurrence Interval

0.01 1 10 100

Project Design Life or Planning Horizon (y) Figure A-1: Relationship between Return Period and Encounter probability

Table A-1: Risk selection based on encounter probability concepts. Considered Design Life Chosen Level of Risk of at Least One Event Occurring or Planning % Chance Horizon y 1 2 5 10 20 30 Equivalent Return Period (y) 10 995 495 195 95 45 29 20 1990 990 390 190 90 57 30 2985 1485 585 285 135 85 40 3980 1980 780 380 180 113 50 4975 2475 975 475 225 141

Storm Tide Study 41/22526/412243 Final Report Addendum

Appendix B Tropical Cyclone Dataset Summary

Storm Tide Study 1 Final Report Addendum

Start Finish At Maximum Intensity Within Radius At Closest Approach Name Date Lat Long Date Lat Long p0 Date Dist Bear Vfm Theta p0 Date Dist Bear Vfm Theta deg deg deg deg hPa km deg m/s deg hPa km deg m/s deg

CY0288_1960 27-Feb-60 -16.7 155.3 3-Mar-60 -28.7 158.3 964 1-Mar-60 527 50 4.3 161 990 3-Mar-60 441 96 1.0 119 CY0296_1961 26-Jan-61 -14.3 161.2 1-Feb-61 -31.7 162.0 998 31-Jan-61 529 68 3.8 158 998 31-Jan-61 528 70 3.8 158 Cy310_1962 29-Dec-62 -17.7 150.8 31-Dec-62 -26.0 151.7 978 31-Dec-62 230 349 6.0 270 978 31-Dec-62 230 349 6.0 270 CY0317_1963 3-Feb-63 -16.0 151.5 6-Feb-63 -32.0 161.3 994 5-Feb-63 511 28 12.5 169 998 5-Feb-63 322 80 12.0 169 CY0323_1963 30-Mar-63 -22.3 153.0 2-Apr-63 -24.3 165.0 1000 31-Mar-63 532 9 3.9 121 1000 31-Mar-63 503 24 7.7 113 Audrey_1964 6-Jan-64 -10.2 141.5 14-Jan-64 -30.4 153.9 984 13-Jan-64 426 266 24.1 116 986 14-Jan-64 214 210 24.0 116 Judy_1965 25-Jan-65 -11.6 133.0 5-Feb-65 -31.5 164.5 990 3-Feb-65 472 77 4.5 217 993 3-Feb-65 365 106 4.5 156 Dinah_1967 22-Jan-67 -12.7 163.8 31-Jan-67 -35.2 161.5 945 28-Jan-67 529 353 5.3 164 953 29-Jan-67 194 51 5.9 141 Barbara_1967 17-Feb-67 -13.1 163.5 21-Feb-67 -28.8 152.6 987 21-Feb-67 62 191 3.3 252 987 21-Feb-67 57 168 7.0 257 Elaine_1967 13-Mar-67 -14.7 149.3 19-Mar-67 -32.0 164.0 996 17-Mar-67 364 44 12.7 158 996 17-Mar-67 330 70 12.7 158 Glenda_1967 26-Mar-67 -12.5 155.3 5-Apr-67 -31.7 159.3 988 4-Apr-67 533 77 4.3 212 988 4-Apr-67 471 85 5.2 169 Cy562_1967 6-Dec-67 -15.5 151.6 10-Dec-67 -27.7 163.7 996 8-Dec-67 443 62 5.5 138 997 8-Dec-67 424 45 6.6 134 Cy563_1967 9-Dec-67 -24.9 154.5 12-Dec-67 -21.9 156.4 998 10-Dec-67 337 57 2.5 146 1002 9-Dec-67 326 43 8.3 133 Cy680_1969 11-Apr-69 -10.6 164.9 16-Apr-69 -31.5 160.0 998 16-Apr-69 495 115 7.7 131 1001 15-Apr-69 313 79 6.8 168 Cy575_1969 14-Nov-69 -20.1 154.0 15-Nov-69 -32.4 152.5 1004 15-Nov-69 359 13 9.6 163 1004 15-Nov-69 107 102 16.3 192 Dora_1971 10-Feb-71 -19.5 152.7 17-Feb-71 -25.7 151.9 995 17-Feb-71 97 42 7.6 302 995 17-Feb-71 95 34 7.6 302 Fiona_1971 16-Feb-71 -16.0 140.8 28-Feb-71 -20.8 161.8 994 21-Feb-71 425 341 2.3 90 995 22-Feb-71 412 349 2.9 79 Lena_1971 13-Mar-71 -12.4 154.8 19-Mar-71 -24.0 167.8 988 16-Mar-71 493 12 2.8 123 990 16-Mar-71 474 19 1.4 42 Althea_1971 19-Dec-71 -10.9 159.0 29-Dec-71 -34.8 164.7 978 27-Dec-71 322 19 5.2 78 988 27-Dec-71 265 355 5.6 84 Wendy_1972 4-Feb-72 -16.0 165.2 9-Feb-72 -25.8 156.0 1001 9-Feb-72 492 64 4.2 270 1001 9-Feb-72 356 45 4.9 288 Daisy_1972 5-Feb-72 -14.9 150.0 13-Feb-72 -27.4 158.1 959 10-Feb-72 508 14 2.4 270 997 12-Feb-72 84 11 1.9 74 Emily_1972 27-Mar-72 -11.0 157.5 4-Apr-72 -34.4 153.2 993 2-Apr-72 456 329 3.6 172 1006 2-Apr-72 46 244 10.4 159 Kirsty_1973 24-Feb-73 -14.6 157.4 1-Mar-73 -34.3 160.6 980 27-Feb-73 443 30 9.9 169 984 27-Feb-73 292 77 8.9 168 Wanda_1974 20-Jan-74 -17.7 148.8 25-Jan-74 -27.3 149.9 997 24-Jan-74 283 353 4.2 222 1003 24-Jan-74 220 316 4.6 243 Pam_1974 3-Feb-74 -19.9 163.1 6-Feb-74 -29.9 157.8 972 5-Feb-74 480 78 5.3 194 974 5-Feb-74 439 94 8.7 182 Zoe_1974 6-Mar-74 -18.8 154.3 14-Mar-74 -32.0 158.8 982 11-Mar-74 501 11 3.6 180 983 12-Mar-74 19 91 5.3 194 Alice_1974 21-Mar-74 -22.6 154.3 22-Mar-74 -29.7 161.1 1010 21-Mar-74 473 25 8.0 140 1010 21-Mar-74 430 48 7.6 137 Beth_1976 13-Feb-76 -16.5 149.9 22-Feb-76 -24.9 151.3 992 20-Feb-76 527 17 3.5 252 996 21-Feb-76 379 341 4.2 263 Colin_1976 25-Feb-76 -10.3 155.5 4-Mar-76 -33.8 158.9 955 1-Mar-76 488 26 4.6 174 961 2-Mar-76 295 69 3.8 159 Dawn_1976 3-Mar-76 -17.4 145.6 6-Mar-76 -30.4 155.7 988 5-Mar-76 508 341 9.7 133 992 6-Mar-76 125 69 10.1 156 Watorea_1976 25-Apr-76 -9.5 152.6 28-Apr-76 -27.1 158.9 990 28-Apr-76 426 22 22.1 127 991 28-Apr-76 413 39 22.2 127 Paul_1980 2-Jan-80 -15.1 137.1 8-Jan-80 -30.0 159.6 989 7-Jan-80 443 27 11.2 141 990 7-Jan-80 404 51 11.2 141 Simon_1980 21-Feb-80 -17.0 153.8 28-Feb-80 -30.5 160.5 960 25-Feb-80 532 351 1.1 24 978 27-Feb-80 238 43 7.5 132 Cliff_1981 9-Feb-81 -11.1 171.6 15-Feb-81 -26.0 146.5 985 13-Feb-81 497 58 4.8 257 985 14-Feb-81 290 16 6.8 287 Abigail_1982 22-Jan-82 -25.7 154.3 5-Feb-82 -26.2 166.5 1006 22-Jan-82 371 24 5.9 28 1007 22-Jan-82 273 18 4.8 42 Pierre_1985 18-Feb-85 -11.8 143.3 24-Feb-85 -23.8 160.0 1001 22-Feb-85 505 355 4.8 96 1001 22-Feb-85 495 6 4.8 96 Nancy_1990 28-Jan-90 -18.3 156.0 4-Feb-90 -34.5 155.0 975 1-Feb-90 452 47 8.4 266 980 2-Feb-90 34 75 2.1 167 Daman_1992 15-Feb-92 -13.1 168.5 19-Feb-92 -31.6 157.0 975 18-Feb-92 405 81 7.0 215 978 18-Feb-92 302 118 5.8 207 Fran_1992 9-Mar-92 -18.6 168.3 17-Mar-92 -25.5 159.0 980 17-Mar-92 356 27 9.3 93 987 16-Mar-92 320 5 5.2 95 Roger_1993 12-Mar-93 -10.0 157.0 21-Mar-93 -21.3 160.9 982 17-Mar-93 487 29 1.5 180 985 17-Mar-93 458 31 2.1 0 Rewa_1993 28-Dec-93 -9.5 165.5 21-Jan-94 -29.0 158.0 980 20-Jan-94 499 350 5.2 126 985 21-Jan-94 296 46 7.7 122 Violet_1995 3-Mar-95 -16.0 152.5 8-Mar-95 -29.2 155.1 980 6-Mar-95 425 129 8.2 280 989 7-Mar-95 82 132 2.1 41 Gertie_1995 17-Dec-95 -13.2 125.5 24-Dec-95 -23.0 163.0 994 23-Dec-95 233 13 8.7 53 994 23-Dec-95 233 13 8.6 53 Yali_1997 18-Mar-98 -12.2 165.9 31-Mar-98 -68.5 161.0 989 27-Mar-98 495 115 10.3 180 989 26-Mar-98 477 100 8.3 189 Kerry_2005 8-Jan-05 -18.2 159.8 18-Jan-05 -27.5 157.8 998 15-Jan-05 272 74 2.1 180 998 16-Jan-05 264 84 2.5 114 Hamish_2009 4-Mar-09 -12.5 148.5 11-Mar-09 -22.0 153.1 946 9-Mar-09 486 7 4.0 137 975 10-Mar-09 420 24 3.4 335

Storm Tide Study 1 Final Report Addendum

Appendix C Calibration of the RHD Model

Water Surface Elevations Instantaneous Flow Discharges

41/22526/412243 Storm Tide Study 1 Final Report Addendum

Figure C-1 RHD model performance – comparison of predicted (blue) versus modelled (red) water levels at station Gold Coast Seaway, RMSE for 7 month simulation period =0.06. Figure C-2 RHD model performance – comparison of predicted (blue) versus modelled (red) water levels at station Brisbane Bar, RMSE for 7 month simulation period = 0.10. Figure C-3 RHD model performance – comparison of predicted (blue) versus modelled (red) water levels at station Tangalooma Point, RMSE for 7 month simulation period = 0.10. Figure C-4 RHD model performance – comparison of predicted (blue) versus modelled (red) water levels at station Bongaree Jetty, RMSE for 7 month simulation period = 0.09. Figure C-5 RHD model performance – comparison of predicted (blue) versus modelled (red) water levels at station Mooloolaba, RMSE for 7 month simulation period = 0.07. Figure C-6 RHD model performance – comparison of measured (black) versus modelled (red) water levels at station B4, RMSE = 0.11 for a four-month period (17 November 2004 to 8 March 2005). Figure C-7 RHD model performance – comparison of measured (black) versus modelled (red) water levels at station B5, RMSE = 0.20 for a one-month period (20 November 2004 to 24 December 2004). Figure C-8 RHD model performance – comparison of measured (black) versus modelled (red) water levels at station B9, RMSE = 0.15 for a one-month period (11 November 2004 to 18 December 2004). Figure C-9 RHD model performance – comparison of measured (black) versus modelled (red) water levels at station B10, RMSE = 0.13 for a four-month period (18 November 2004 to 09 March 2005). -3.0

Figure C-10 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station L1 (12 November 2004 to 16 February 2005), RMSE=0.15. Figure C-11 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station L1 (1 April 2005 to 30 May 2005). Figure C-12 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station B3 (12 November 2004 to 16 February 2005), RMSE = 0.14. -3.0

Figure C-13 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station B3 (1 April 2005 to 30 May 2005). Figure C-14 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station B7, RMSE = 0.09 for a four-month period (20 November 2005 to 07 March 2005). Figure C-15 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station C2, RMSE = 0.10 for a four-month period (13 November 2005 to 11 March 2005). Figure C-16 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station Monterey Keys, RMSE = 0.07 for a one-month period (01 February 2005 to 07 March 2005). Figure C-17 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station Coomera Shores, RMSE = 0.09 for a one-month period (28 January 2005 to 04 March 2005). Figure C-18 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station Brygon Creek, RMSE = 0.10 for the period 01 to 15 February 2005. Figure C-19 RHD model performance – comparison of measured (black) versus modelled water levels (red – RHD and purple - integrated 1-D links) at station B11, RMSE = 0.13 for a four-month period (11 November 2004 to 09 March 2005). Figure C-20 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station N1, RMSE = 0.09 for the period 24 December 2004 to 31 January 2005. Figure C-21 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station N2, RMSE = 0.12 for the period 24 December 2004 to 31 January 2005. Figure C-22 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at station N3, RMSE = 0.10 for the period 13 November 2004 to 09 March 2005. Figure C-23 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) water levels at Coombabah Lake. Figure C-24 RHD model (A) performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge at station Jumpinpin. Figure C-25 RHD model (A) performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge at Gold Coast Seaway. Figure C-26 RHD model (A) performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge at Gold Coast Seaway – Northern Channel. Figure C-27 RHD model (A) performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge at Gold Coast Seaway – South Channel. Figure C-28 RHD model performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge at the mouth of the Coomera River. -3.0

Figure C-29 RHD model performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge in the northern arm of the Coomera River. Figure C-30 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) instantaneous flow discharge in the northern arm of the Coomera River. Figure C-31 RHD model performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge in the southern arm of the Coomera River. Figure C-32 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) instantaneous flow discharge in the southern arm of the Coomera River. Figure C-33 RHD model performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge at Victoria Point. Figure C-34 RHD model (A) performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge in the western channel of Russell Island. Figure C-35 RHD model (A) performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge in the east channel of Sovereign Island. Figure C-36 RHD model (A) performance – comparison of measured (black) versus modelled (red) instantaneous flow discharge in the west channel of Sovereign Island. -3.0

Figure C-37 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) instantaneous flow discharge at the mouth of the Nerang River. Figure C-37 1-D link integrated into the RHD model – comparison of measured (black) versus modelled (purple) instantaneous flow discharge at the mouth of the Logan River.

Appendix D Calibration to Historical Tropical Cyclone Events

41/22526/412243 Storm Tide Study 1 Final Report Addendum G:\41\22526\WP\Appendix D Historical Storms.doc 2 1.1 Introduction The deterministic performance of the numerical wind, wave and storm tide models has been demonstrated here by numerical hindcasts of the effects of tropical cyclones (TC) Dinah in 1967 and the unnamed 1954 storm. These are the most significant tropical cyclones known to have directly impacted this section of the coastline. However, a lack of quantitative data significantly restricts the extent to which comparisons can be made with the modelled scenarios. The descriptions of impacts from these storms and some of the meteorological information have been provided by Mr Jeff Callaghan.

1.2 TC Dinah January 1967 Background and Reported Impacts Tropical Cyclone Dinah is one of the significant tropical cyclones to affect the Queensland and Northern New South Wales coastlines. It caused severe damage at Heron Island - initially inundation from large NE swells and a day later from strong winds. The storm then re-curved and its centre passed about 6 km west of the Sandy Cape Lighthouse, which recorded a minimum central pressure of 945 hPa and where the ocean reached levels on the beach estimated to be some 10 m “above normal”. The Sandy cape lighthouse suffered some damage during this event with paint removed from buildings and access tracks blocked with fallen trees. Although Dinah was well off the mainland coast, tree damage extended from Rockhampton to Grafton and houses were unroofed at Bundaberg and Maryborough. Huge seas and associated storm surge caused severe erosion at Emu Park, Yeppoon, and in the Maryborough, Bundaberg area. Along the Sunshine coast, storm surge effects inundated cane farms at Bli Bli and water was reportedly knee-deep in Hastings St Noosa. In Moreton Bay near Sandgate, seawater 1.5 metres deep came into properties. More than one hundred were flooded and at Cribb Island one house was washed into the sea. Storm surge also affected the Gold Coast and water lapped the decking of the old Jubilee Bridge, which was estimated to be about 1.5 metres above Highest Astronomical Tide. A section of the esplanade collapsed at Surfers Paradise. A similar storm tide occurred on the Tweed River, isolating Fingal. The most affected area of northern NSW was near Brunswick Heads where several banana plantations were destroyed and hundreds of acres of sugar cane in the Tweed Valley were damaged. A wind gust even reportedly overturned a car at Evans Head. Available Storm Data The BoM National Climate Centre (NCC) Official track for this event (Figure 0-1) was obtained and compared with other references provided by Jeff Callaghan. This shows a generally steady path towards the coast until the storm reached the tip of Fraser Island, when it curved eastwards and continued moving in a south-easterly direction. The minimum barometric pressure recorded at the Sandy Cape Lighthouse during the passage of the eye was 944.8 hPa on 29/01/1967 at 07:30 UTC A series of synoptic views of the storm, from 25/01/1967 at 11:00 UTC to 30/01/1967 at 23:00 UTC, is provided in Figure 0-2.

G:\41\22526\WP\Appendix D Historical Storms.doc 1 Figure 0-1: Official Bureau of Meteorology Track for TC Dinah.

G:\41\22526\WP\Appendix D Historical Storms.doc 2 G:\41\22526\WP\Appendix D Historical Storms.doc 3 Figure 0-2: Synoptic Situation for TC Dinah (provided by Mr Jeff Callaghan).

G:\41\22526\WP\Appendix D Historical Storms.doc 4 Figure 0-3: Track of Dinah showing the official and modified storm track near Brisbane.

G:\41\22526\WP\Appendix D Historical Storms.doc 5 Wind and Pressure Field Calibration

The wind model calibration process aims to find an optimal set of Holland model R and Bo parameters that, assuming the BoM central pressure estimates are reliable, would minimise the error between modelled and measured wind and pressure across all the available sites. While reliable barometer readings are available from the Sandy Cape and Cape Moreton lighthouses and Brisbane Airport (Aero), unfortunately the only reliable wind measurements for Dinah are from the Brisbane Airport. This site is around 175 km west of the storm centre and is known to suffer attenuation in SE conditions (Jeff Callaghan, personal communication). Wind estimates are available from the lighthouse records but these have proven unreliable in our view, as described below. After much testing of a range of parameter choices, the surface pressure data at Cape Moreton and Brisbane Aero appeared inconsistent with the track of the storm as it passed south of Sandy Cape. After inspection, it was decided to adjust the raw BoM track fixes by including additional fixes offshore Brisbane that would provide a more gradual (smoother) turning of the track. This had the effect of moving the storm centre some 40 km further west and succeeded in bringing the various modelled and measured surface pressure records into reasonable agreement, as shown in the figures that follow. The adopted best fit values of R and Bo were 58 km and 6.8 when the storm was at closest approach to Brisbane. Firstly Figure 0-4 summarises the model performance at Sandy Cape Lighthouse. The top panel compares modelled mean winds with the estimated winds noted by the observer, which unfortunately do not appear at all favourable. The difficulty here is that the observer could only estimate the wind visually based on the Beaufort Scale system. If the reported mean winds of around 90 knots were true, these would be physically inconsistent with the accurately measured surface pressure of 945 hPa. Accordingly it is argued that the observer was likely to have reported winds more closely associated with the peak gusts than the 10- minute means (e.g. 40% higher), notwithstanding that any observer has a limited experience of such high winds in any case and the location and elevation of the lighthouse would likely result in some topographic enhancement. The middle panel shows the comparison of the modelled and estimated directions, which are generally favourable except for some deflection after the eye passage. The bottom panel however shows a very good agreement with the objectively measured barometric pressure, especially during the storm approach. The separation after the eye passes tends to suggest that the storm may have continued to increase in size slightly more than has been assumed in the modelling. Next, Figure 0-5 presents the corresponding results at Cape Moreton Lighthouse, where similar problems exist in interpreting the reported wind speeds. In a recent wind tunnel modelling study of Cape Moreton conducted by the James Cook Cyclone Testing Station for the Bureau of Meteorology (Ginger and Harper, 2004), topographic amplification was estimated to be as much as a factor of 3 in some directions due to the steep-sided site. The reported winds are shown in the top panel as a series of points, while a topographic correction to those is shown as the dashed line. This is still well above the modelled values and the remaining mismatch is again attributed to the observer likely estimating peak gust wind speeds rather than mean wind speed. Again the wind directions are reasonable and the objective comparison of the pressure values is very good. Finally, Figure 0-6 has the results for Brisbane Aero, where the winds have been objectively estimated from an anemometer record, although difficulties remain still in reading the chart- based mean windspeed record from the Dines Anemograph in use at the time.

G:\41\22526\WP\Appendix D Historical Storms.doc 6 Estimated Modelled

50

45

40

likely peak gusts 35

30

25

20 Mean Wind Speed m/s Speed Wind Mean

15

10

5

0 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Modelled Estimated

360

320

280

240

200

160 Wind Direction deg Direction Wind 120

80

40

0 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Modelled Measured 1014

1004

994

984

974 MSL Pressure hPa Pressure MSL

964

954

944 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Figure 0-4: Sandy Cape Lighthouse comparisons for TC Dinah

G:\41\22526\WP\Appendix D Historical Storms.doc 7 Estimated (corercted) Estimated (Uncorrected) Modelled

50

45

40

likely peak gusts 35

30

25

20 Mean Wind Speed m/s Speed Wind Mean

15

10

5

0 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Modelled Estimated

360

320

280

240

200

160 Wind Direction deg Direction Wind 120

80

40

0 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Modelled Measured 1020

1015

1010

1005

1000

995 MSL Pressure hPa Pressure MSL 990

985

980

975 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Figure 0-5: Cape Moreton Lighthouse comparisons for TC Dinah

G:\41\22526\WP\Appendix D Historical Storms.doc 8 Measured Modelled 20

18

16

14

12

10

8 Mean Wind Speed m/s Speed Wind Mean

6

4

2

0 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Modelled Measured

360

320

280

240

200

160 Wind Direction deg Direction Wind 120

80

40

0 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Modelled Measured 1020

1015

1010

1005

1000 MSL Pressure hPa Pressure MSL

995

990

985 26/01/1967 12:00 26/01/1967 00:00 27/01/1967 12:00 27/01/1967 00:00 28/01/1967 12:00 28/01/1967 00:00 29/01/1967 12:00 29/01/1967 00:00 30/01/1967 12:00 30/01/1967 00:00 31/01/1967 12:00 31/01/1967 Time (UTC)

Figure 0-6: Brisbane Aero comparisons for TC Dinah

G:\41\22526\WP\Appendix D Historical Storms.doc 9 This overestimation by the model is consistent with the expected attenuation at the airport site (Jeff Callaghan, personal communication) and so seems reasonable. It also compares favourably with observations from a vessel (Eastern Moon) located near the Pile Light early on the 29th (UTC) as the winds were building. However, similar to the issues with the Sandy Cape and Cape Moreton wind estimates, later observations of hurricane force winds from this vessel have been discounted as not being consistent with the observed regional pressures and the noted absence of any significant wind damage. It is again likely that the Ships Master has been inclined to respond to the force of the peak gusts rather than the true mean wind conditions. Taken overall, these results confirm the ability of a Holland-like wind and pressure field model to represent tropical cyclone forcing, at least within the influence of the central vortex (3 to 5 times R). The apparent underprediction of estimated winds at several locations needs to be balanced with the model’s overprediction of winds at the only reliable wind measurement site, where attenuation is expected. The pressure matches are good throughout. Storm Tide Modelling As mentioned previously, objectively measured water level data for this event is scarce, with the Brisbane Bar tide gauge being a principal reference. There is no measured wave data available but some photographs (refer below) indicate conditions in Moreton Bay. In order to reproduce the measured water level at this location, the numerical model must first be able to match the predicted tides during this period. This in itself can be a difficult exercise, as demonstrated in the main report and Figure 0-7 here shows that there is a residual model error in this case of the order of 0.2 m. This is based on using the Mike21 tidal constituents on the open boundaries and was not able to be improved within the time available.

2.0 Tide m AHD HA T m A HD Tide Error m Predicted Tide m AHD

1.5

1.0

0.5

Water Level m AHD m Level Water 0.0

-0.5

-1.0 28/01 28/01 29/01 29/01 30/01 30/01 31/01 31/01 01/02 00: 12: 00: 12: 00: 12: 00: 12: 00: Time UTC

Figure 0-7: Brisbane Bar tide modelled and predicted for TC Dinah

G:\41\22526\WP\Appendix D Historical Storms.doc 10 Notwithstanding the modelled tide error during this event, the model was used to simulate the storm surge component response at Mean Sea Level and also coupled with the modelled tide boundaries. The latter experiment however could not be completed due to unresolvable model instabilities and so had to be abandoned. This left only the separate tide and surge (at MSL) model simulations able to be linearly combined, which due to the aforementioned tide modelling errors, has been superceded here by the predicted tide for illustration purposes. Figure 0-8 summarises this result, whereby the black dashed line is the modelled surge + (predicted) tide and the blue dashed line is the actual measured water level at the tide gauge. HAT is also indicated (approximately 1.5 m AHD) plus the predicted tide (black dotted) and the measured tidal residual is shown together with the modelled surge component (solid black and blue respectively). It can be seen that the modelled surge has the character of the measured tidal residual signal, namely a positive surge followed by a negative surge as the storm passes and winds reverse, but is somewhat lower in magnitude. The high tide on the 29th can be seen to have the highest excursion, exceeding HAT by about 0.15 m at the gauge, while the modelled response is lower by about 0.25 m. The time of the Courier Mail photos at Sandgate (shown later in Figure 0-10) is also indicated on this graph.

Tide + Surge m AHD Surge m 2.0 Tide m AHD HAT m A HD photos Measured Tide m AHD Measured Surge m 1.5

1.0

0.5

Water Level m AHD m Level Water 0.0

-0.5

-1.0 28/01 28/01 29/01 29/01 30/01 30/01 31/01 31/01 01/02 00: 12: 00: 12: 00: 12: 00: 12: 00: Time UTC

Figure 0-8: Brisbane Bar water levels modelled and predicted for TC Dinah

Notwithstanding the model is slightly low it is useful to consider the various water level component estimates in this vicinity and the degree of consistency with the observed local impacts. For example, the top panel of Figure 0-9 shows the modelled total storm tide (tide + surge + wave setup) and the estimated wave runup potential at the nominal Sandgate site, where a low seawall was present (refer Figure 0-1). Also shown is the approximate ground elevation behind the seawall, which is close to the HAT. This modelled runup/overwash potential is consistent with the photographed impacts on the 28th and confirms that conditions would have been more severe over the following few days.

G:\41\22526\WP\Appendix D Historical Storms.doc 11 7.0 Tide + Surge m AHD 6.0 Total Storm Tide m AHD Total Wave Runup m AHD 5.0 HAT m A HD Sandgate houses m AHD 4.0 potential only

3.0 photos 2.0 Water Level m AHD m Level Water

1.0

0.0

-1.0 28/01 28/01 29/01 29/01 30/01 30/01 31/01 31/01 01/02 00: 12: 00: 12: 00: 12: 00: 12: 00: Time UTC

14.0

Hs m 12.0 Setup m Runup m 10.0 Tp s

8.0

6.0 Wave Parameter 4.0

2.0

0.0 28/01 28/01 29/01 29/01 30/01 30/01 31/01 31/01 01/02 00: 12: 00: 12: 00: 12: 00: 12: 00: Time UTC

Figure 0-9: Brisbane Bar modelled storm tide and wave conditions for TC Dinah

G:\41\22526\WP\Appendix D Historical Storms.doc 12 The lower panel shows the predicted wave conditions at the Pile Light site, which although not identical to the Sandgate site, are considered here to be broadly representative. The graph indicates Hs, Tp and the wave setup and runup components. The resulting persistent overwash combined with the exceedance of HAT appears consistent with the reported impacts.

Figure 0-10: Courier Mail photos of high tide at Sandgate 28/03/1967 (via Jeff Callaghan)

Consideration is now given to the likely conditions in the Gold Coast area during TC Dinah, whereby Figure 0-11 shows the modelled tide, surge and tide+surge values at the “Gold Coast Waverider” site in approximately 18 m of water. This shows a quite small storm surge component of about 0.2 m, not inconsistent with the fact that the storm centre is over 100 km distant. These are all directly taken from the model in this case, which without local tide gauge records, cannot be tested in any objective way. This leaves only qualitative comparisons with the reported impacts of the total water level and wave conditions, summarised in Figure 0-12.

G:\41\22526\WP\Appendix D Historical Storms.doc 13 2.0 Tide + Surge m AHD Surge m 1.5 Tide m AHD HAT m AHD

1.0

0.5

0.0 Water Level m AHD Level Water

-0.5

-1.0 28/01 28/01 29/01 29/01 30/01 30/01 31/01 31/01 01/02 00: 12: 00: 12: 00: 12: 00: 12: 00: Time UTC

Figure 0-11: Gold Coast modelled water levels for TC Dinah The reported impacts speak of extensive beach erosion and incursions of waves and seawater into low lying properties along the entire southern Queensland and Northern NSW coastline. There is also a report that water levels in the Broadwater/Nerang River rose at one time to the deck level of the old Jubilee Bridge near Main Beach. This is described as being 1.5 m above HAT but the water level may also have been affected by riverine flooding to an unknown extent due to the associated heavy rainfall. Also, the tidal regime and ocean connectivity at the time (prior to the construction of the Seaway and the extensive canal estates) makes it difficult to interpret such a level. Nevertheless, Figure 0-12 (top panel) suggests that, at the ocean shoreline near the present Spit, mean water levels as a result of tide+surge+setup probably reached to 3 m AHD, which is almost 2 m above HAT. This would certainly have created significant erosion of the beachface. Additionally, wave runup effects were possible up to 6 m AHD, limited only by the available dune crest heights. Typical estimated values of the present day dune crest along the Spit are shown also, together with the implied Jubilee Bridge deck level. It remains possible that overwash from the ocean and even wave setup may have added to water levels in the Broadwater, but this is speculative. The lower panel of Figure 0-12 shows the modelled nearshore Hs is of the order of 11.5 m and has a Tp of similar magnitude in seconds, with shoreline setup of the order of 2 m magnitude and runup potential of 5.5 m. Conclusion While lacking quantitative value in many respects, the modelling approach demonstrates an ability to reasonably represent an event such as TC Dinah. It is acknowledged that the modelled surge is approximately 0.25 m low at the Brisbane Bar, but this is not significant in the context of the small measured amplitude, which is to be expected for an event this far offshore and moving parallel to the coast. Importantly, as discussed in the main report, other broadscale forcing also likely contributed to this event and is responsible for this small mismatch. We have provided a transparent and reproducible account here and resisted the temptation to adjust the wind and pressure forcing in any way in order to force better agreement.

G:\41\22526\WP\Appendix D Historical Storms.doc 14 7.0 Tide + Surge m AHD Total Storm Tide m AHD 6.0 Total Wave Runup m AHD HA T m A HD potential only 5.0 Spit Crests m AHD Jubilee Bridge Levels m AHD 4.0

3.0

2.0 WaterLevel m AHD 1.0

0.0

-1.0 28/01 28/01 29/01 29/01 30/01 30/01 31/01 31/01 01/02 00: 12: 00: 12: 00: 12: 00: 12: 00: Time UTC

14.0

Hs m 12.0 Setup m Runup m 10.0 Tp s

8.0

6.0 Wave Parameter 4.0

2.0

0.0 28/01 28/01 29/01 29/01 30/01 30/01 31/01 31/01 01/02 00: 12: 00: 12: 00: 12: 00: 12: 00: Time UTC

Figure 0-12: Gold Coast modelled storm tide and wave conditions for TC Dinah

G:\41\22526\WP\Appendix D Historical Storms.doc 15 1.3 TC 1954 Background and Reported Impacts In early February 1954, northern Queensland had already been severely impacted by tropical cyclones that caused inundation in large areas, then, on the 17th, a new storm developed in the Coral Sea and moved towards the central Queensland coast. Upon nearing the coast, the storm changed path and turned south bringing with it gales and heavy rain, which affected the east coast right down into northern NSW. On the 19th February there was an indication that the storm was moving in a south-easterly direction, away from the Queensland coast. However, it swung south again and passed within 100 km of Brisbane, with the city experiencing wind-gusts of over 100 km/h. On the 20th of February around 10 pm the storm centre made landfall at the border Twin Towns of Coolangatta, Tweed Heads. This led to very damaging waves, storm surge, and a destructive wind zone near the centre. The weather charts indicate that the cyclone approached the Gold Coast-Tweed area from the north northeast to be overland west of Coffs Harbour by 3 PM 21 February 1954. Record rainfall accompanied the cyclone with 900 mm were recorded at Springbrook in the 24 hour period up to landfall. Floods combined with storm surge and cyclonic winds resulted in some 26 to 30 people losing their lives. In the worst hit areas of Northern NSW, flood waters began to rapidly rise around 7 PM 20 February which was some 3 hours before landfall.

 The storm crossed the coast at Coolangatta with a pressure reading in the eye of 973 hPa. Some reports from the Coolangatta/Tweed Heads area had pressure readings to 962 hPa.

 The cyclone centre passed well to the east of Brisbane however a record low pressure reading of 982.7 hPa was recorded at the Weather Bureau office in the City. At the Airport the mean seal level pressure was measured at 6pm at 981.8hPa (1.4hPa lower than the more western City reading at the time). At Cape Moreton Lighthouse (outside the eye) the lowest mean sea level pressure read was 978.0 hPa at 3pm.

 There were two barometers at the Condong Sugar Mill situated inland from Coolangatta, one an aneroid registered 28.8 inches (975 hPa), while the other mercury barometer read 973 hPa. The eye took two hours to pass over the mill around 11pm.

 The ship Kaipara (from New York) berthed at New Farm in Brisbane on the afternoon of 21 February 1954. According to the captain the ship had fared badly in the storm with cabins flooded and some superficial damage to the ship. Some (likely estimated-Beaufort) observations from Kaipara with mean winds and (likely measured-barometer) pressure are given below:-

 0600UTC 19/2/1954 26.7S 154.0E 150/50 knots bar 996.7hPa

 0900UTC 19/2/1954 26.6S 154.1E 150/50 knots bar 994.2hPa

 1200UTC 19/2/1954 26.6S 154.3E 130/50 knots bar 993.7hPa

 0200UTC 20/2/1954 27.4S 153.6E 160/65 knots bar 974.1hPa

 0600UTC 20/2/1954 27.0S 153.6E 210/40 knots bar 967.1hPa The storm caused widespread structural damage on the Gold Coast, Sunshine Coast and around Brisbane. Some of the more significant impacts from the storm are listed below;

 At Noosa, cyclonic southerly winds pushed the waters of Lake Weyba 0.6 m high over the road north of the lake.

G:\41\22526\WP\Appendix D Historical Storms.doc 16  At Moreton Bay, a storm surge of 0.64 m was measured at the tide gauge. However, it was worse on the Beachmere foreshore with boats lodged in mangrove tree tops.

 The sea came into the shopping area of Coolangatta and waves at Kirra brought 2 m of water onto the highway picking up cars. Storm surge caused some fifty families to be evacuated from the Broadwater on the Gold Coast and residents had to be rescued from Macintosh Island.

 In Cudgen (NSW) some houses were destroyed and trees more than 1 m in diameter were twisted out of the ground.

 Byron Bay was flooded with sea water and the outer section of the jetty was swept away taking with it all 22 vessels comprising the fishing fleet.

 The Richmond River at Kyogle peaked during the early hours on Sunday 21st February 1954, recording an unprecedented height of 19.1 Metres. Ten people drowned on this day. The cyclone had produced enormous amounts of rain and wind and caused destruction to bridges, which were swept away or destroyed. Apart from the houses that were washed away, many were swept off their stumps or badly damaged.

 A record flood hit the Coffs Harbour region where some of the heaviest rainfall was recorded

 It is thought that in total, between 26 and 30 lives were lost as a result of the 1954 cyclone.

Available Storm Data

The published BoM track for the “1954” storm was used as the basis of the reconstruction, Figure 0-13, but following advice from Mr Jeff Callaghan, a reanalysis prepared by the Queensland Regional Office Severe Weather Section (Figure 0-14) was used to digitise a revised storm track, which increased the estimated peak intensity of the storm beyond the official BoM value. This unofficial reanalysis of TC 1954’s intensity reveals that it may well have been one of the more intense tropical cyclones in this region. Using the updated interpretation, the maximum intensity of the storm was assessed during the period 19/02/1954 10:30 UTC to 20/02/1954 00:00 UTC when its central pressure may have been as low as 968 hPa. The track parameters were therefore adjusted to accommodate a diversion to the revised maximum intensity over a 24 h period, returning to the official estimates largely at 20/02/1954 23:00 UTC.

G:\41\22526\WP\Appendix D Historical Storms.doc 17 Figure 0-13: BoM official track for TC 1954

Figure 0-14: TC 1954 re-analysis by Brisbane Severe Weather Section staff

G:\41\22526\WP\Appendix D Historical Storms.doc 18 Wind and Pressure Calibration

The wind model calibration process attempted to find an optimal set of R and Bo that would minimise the error between modelled and measured wind and pressure across all the available sites. Data from three monitoring stations (Brisbane Regional Office, Brisbane Aero and Cape Moreton Lighthouse) was used for calibration. This proved to be a difficult task particularly for wind calibration due to the scarcity of objectively measured data. In the case of Brisbane Regional Office the peak winds were recorded in the city centre when the storm was some 61 km north-east. For Brisbane Aero peak winds were recorded when the storm was some 138 km south-east of the site and for Cape Moreton Lighthouse peak winds were recorded when the storm was 57 km east of site (Figure 0-15). Based on the supplied reanalysis, the storm was estimated to have been at its maximum intensity from 19th 10:30 UTC to 20th 00:00 UTC when its estimated central pressure was 968 hPa (Figure 0-16). As the wind data was erratic and scarce, the key calibration point from a pressure calibration perspective is at the closer Cape Moreton, where the pressure has been reasonably well matched. Analyses showed that an R value of around 65 km and a Bo value of 7.05 were capable of reproducing estimated wind speeds well over a wide range and provided a close match to the shape of the landfall pressure profiles available both north and south of the track. Pressure values remote from the storm centre generally remain over-predicted. Wind values have been slightly over-predicted in case of Brisbane Aero and under predicted in the case of Brisbane Regional Office. There was negligible wind data for Cape Moreton, affected also by the issues identified earlier with Dinah, so conclusive comparisons could not be made. Figure 0-17, Figure 0-18 and Figure 0-19 summarise the generally good accuracy obtained in fitting the numerical mean wind and pressure model to the available data for Brisbane Regional Office, Brisbane Aero and Cape Moreton Lighthouse respectively. The solid blue line represents the modelled results whereas the red dotted line relates to measured data. In addition, the Cape Moreton graphs show reports from the Kaipara at various stages of its hazardous voyage when offshore near Cape Moreton. Although the reported winds from the vessel are again suspect, the barometer readings are of special interest and suggest that the storm may have been even more intense than even the re-analysis used here suggests. Storm Tide Modelling We were unable to source the Brisbane Bar tide gauge record for this event to verify the reported 0.64 m surge, but if this is correct the modelled surge peak of 0.17 m is approximately 0.5 m lower. Instead, we offer graphs and commentary on the modelled storm tide values for the “Gold Coast Waverider” location and compare these with the reported impacts. Figure 0-20 shows that a storm surge component of about 0.5 m is generated by the model, peaking almost simultaneously with the high tide late in the morning (UTC) of 20/02 and, although not exceeding HAT, was likely responsible for the degree of damage reported. Next, Figure 0-21 shows that the potential wave runup was of the order of the Spit dune crest but significant overtopping is not indicated on this basis, although erosion would be extensive. An indicative level for Macintosh Island is also indicated, suggesting that it may have been within reach of wave setup effects, assuming again the bar configuration at the time may have permitted some setup to penetrate into the Broadwater. More than likely however, the reported severe riverine flooding would have contributed to the situation that required evacuations. It can be noted that the estimated wave conditions from the 1954 event are actually somewhat lower than those during Dinah, even though it was a direct hit on the Gold Coast. This is related to the difference in storm tracks.

G:\41\22526\WP\Appendix D Historical Storms.doc 19 Figure 0-15: Track Plot (with Time UTC) of TC 1954 in proximity to the three monitoring stations.

Figure 0-16: Track Plot (with Time UTC) of TC 1954 during maximum intensity (Central Pressure = 968 hPa).

G:\41\22526\WP\Appendix D Historical Storms.doc 20 Measured Modelled

35

30

25

20

15 Mean Wind Speed m/s Speed Wind Mean

10

5

0 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (GMT)

Modelled Measured

360

320

280

240

200

160 Wind Direction deg Direction Wind 120

80

40

0 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (UTC)

Measured Modelled

1010

1005

1000

995

990

985

980 MSL Pressure hPa Pressure MSL

975

970

965

960 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (UTC)

Figure 0-17: Brisbane Regional Office comparisons for TC 1954

G:\41\22526\WP\Appendix D Historical Storms.doc 21 Measured Modelled

35

30

25

20

15 Mean Wind Speed m/s Speed Wind Mean

10

5

0 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (UTC)

Modelled Measured

360

320

280

240

200

160 Wind Direction deg Direction Wind 120

80

40

0 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (UTC)

Measured Modelled

1010

1005

1000

995

990

985

980 MSL Pressure hPa Pressure MSL

975

970

965

960 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (UTC)

Figure 0-18: Brisbane Aero comparisons for TC 1954

G:\41\22526\WP\Appendix D Historical Storms.doc 22 Estimated Kaipara Report Modelled

35

30

25

20

15 Mean Wind Speed m/s Speed Wind Mean

10

5

0 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (UTC)

Modelled Measured Kaipara Estimated

360

320

280

240

200

160 Wind Direction deg Direction Wind 120

80

40

0 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (UTC)

Measured Kaipara Measured Kaipara Modelled Modelled

1010

1005

1000

995

990

985

980 MSL Pressure hPa Pressure MSL

975

970

965

960 18/02/1954 12:00 18/02/1954 00:00 19/02/1954 12:00 19/02/1954 00:00 20/02/1954 12:00 20/02/1954 00:00 21/02/1954 12:00 21/02/1954 Time (UTC)

Figure 0-19: Cape Moreton Lighthouse comparisons for TC 1954

G:\41\22526\WP\Appendix D Historical Storms.doc 23 1.5

Tide + Surge m AHD Surge m 1.0 Tide m AHD HAT m AHD

0.5

0.0 Water Level m AHD

-0.5

-1.0 19/02 00: 19/02 12: 20/02 00: 20/02 12: 21/02 00: Time UTC

Figure 0-20: Gold Coast modelled water levels for TC 1954

Conclusion The 1954 storm wind and pressure fields appear to have been reasonably well represented by a Holland-style model, at least within the limits of the available information. A lack of objective surge and wave data prevents any practical assessment of the model predictions on the Gold Coast but the outcomes are not inconsistent with the reported impacts. However, at the Brisbane Bar it is again acknowledged that the modelled surge is approximately 0.4 m lower than the reported peak level. Without access to the original tide gauge chart it is not possible to better assess this aspect. Importantly, as discussed in the main report, other broadscale forcing also likely contributed to this event and may be largely responsible for this mismatch. More detailed modelling of this event would be of benefit, with special attention given to Moreton Bay, but such a level of effort is outside the scope of this study.

G:\41\22526\WP\Appendix D Historical Storms.doc 24 6.0 Tide + Surge m AHD Total Storm Tide m AHD 5.0 Total Wave Runup m AHD HA T m A HD 4.0 Spit Crests m AHD Macintosh Island m AHD

3.0

2.0 WaterLevel m AHD 1.0

0.0

-1.0 19/02 00: 19/02 12: 20/02 00: 20/02 12: 21/02 00: Time UTC

12.0

Hs m Setup m 10.0 Runup m Tp s

8.0

6.0 Wave Parameter 4.0

2.0

0.0 19/02 00: 19/02 12: 20/02 00: 20/02 12: 21/02 00: Time UTC

Figure 0-21: Gold Coast modelled storm tide and wave conditions for TC 1954

G:\41\22526\WP\Appendix D Historical Storms.doc 25

Appendix E Location of Named Model Sites

41/22526/412243 Storm Tide Study 1 Final Report Addendum

GDA94°

Site Lat Lon

Swan_Bay -27.7334 153.4340

Jumpinpin_Entrance -27.7424 153.4540

Couran_Cove_Ocean -27.8282 153.4390

Sheraton_Mirage -27.9681 153.4350

Main_Beach -27.9817 153.4350

Narrow_Neck -27.9862 153.4350

Surfers_Paradise -27.9997 153.4350

Broadbeach -28.0223 153.4400

Mermaid_Beach -28.0448 153.4450

Nobby_Beach -28.0629 153.4500

Miami_Beach -28.0764 153.4500

Burleigh_Heads -28.0899 153.4610

Tallebudgera_Ck -28.0989 153.4660

Palm_Beach -28.1170 153.4760

Currumbin_Point -28.1259 153.4910

Tugun -28.1484 153.5070

Billinga -28.1574 153.5170

Kirra_Beach -28.1619 153.5270

Coolangatta -28.1619 153.5320

Greenmount -28.1618 153.5420

Rainbow_Bay -28.1618 153.5530

Point_Danger -28.1663 153.5530

Leticia_Spit -28.1798 153.5630

Fingal -28.1978 153.5730

Rocky_Point -27.6975 153.3580

Little_Rocky_Point -27.7111 153.3580

Cabbage_Tree_Point -27.7156 153.3580

2 Storm Tide Study 41/22526/412243 Final Report Addendum

GDA94°

Site Lat Lon

Pimpama_Island -27.7246 153.3580

Steiglitz -27.7291 153.3580

Cabbage_Tree_Point -27.7336 153.3630

Steiglitz -27.7382 153.3630

Sandy_Beach -27.7427 153.3680

Jacobs_Well -27.7697 153.3930

Couran -27.8147 153.4140

Couran_Cove -27.8238 153.4040

Sovereign_Islands -27.8779 153.4140

Paradise_Point -27.8870 153.4140

Currigee -27.9005 153.4190

Runaway_Bay -27.9140 153.4090

Lands_End -27.9231 153.4140

Porpoise_Point -27.9276 153.4190

Nerang_Head -27.9321 153.4190

Gold_Coast_Seaway -27.9366 153.4190

Wave_Break_Island -27.9411 153.4190

The_Broadwater (The_Spit) -27.9501 153.4190

Seaworld -27.9637 153.4190

Southport -27.9682 153.4190

MOB_Tide_Gauge -27.9727 153.4250

Koureyabba -27.7096 153.3930

Jumpinpin -27.7425 153.4340

Labrador -27.9411 153.4140

Nerang_R._Bundall (tidal_station) -27.9992 153.4160

N3_Royal_Pines Harbour -28.0092 153.3720

N2_Campbell St_Sorrento -28.0289 153.4050

N1_Evandale -28.0023 153.4190

41/22526/412243 Storm Tide Study 3 Final Report Addendum

GDA94°

Site Lat Lon

B11_Nerang_Mouth -27.9731 153.4230

C4_Monterey_Keys (Saltwater_Creek) -27.8798 153.3450

C1_Coomera_River (North_Arm) -27.8358 153.3910

C6_Brygon_Creek -27.8848 153.2970

C2_Marine_Precinct (Coomera) -27.8626 153.3400

C5_Coomera_Shores -27.8480 153.3610

Coomera_River_Mouth_(south) -27.8692 153.3990

Coombabah_Lake -27.9088 153.3530

L2_Logan_River_289 Rotary_Park -27.6976 153.2500

L1_Logan_River_Boat_Hire_Jetty -27.6920 153.2920

B3_Logan_River_Mouth -27.6956 153.3290

B6_Pimpama -27.8203 153.3810

TALLEBUDGERA Schuster Park -28.1246 153.4440

TALLEBUDGERA Elanora WQ Control Centre -28.1248 153.4420

CURRUMBIN Elanora Community Center -28.1436 153.4670

4 Storm Tide Study 41/22526/412243 Final Report Addendum

GHD

145 Ann Street Brisbane QLD 4000 GPO Box 668 Brisbane QLD 4001 T: (07) 3316 3000 F: (07) 3316 3333 E: [email protected]

© GHD 2012 This document is and shall remain the property of GHD. The document may only be used for the purpose for which it was commissioned and in accordance with the Terms of Engagement for the commission. Unauthorised use of this document in any form whatsoever is prohibited.

Document Status

Rev Reviewer Approved for Issue Authors No. Name Signature Name Signature Date

41/22526/412243 Storm Tide Study Final Report Addendum