ENVIRONMENT REPORT

PORT PHILLIP BAY AND RECEIVING WATER QUALITY MODELLING: SCENARIOS

Publication 1380 June 2011

A report developed for the ‘Better Bays & Waterways’ program, a water quality improvement plan for Bay and Western Port and their catchments.

PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

REPORT STATUS

Version Date Status Prepared Reviewed Approved By: By: By: V.1 May 2007 Draft SH, KPB, CB KPB KPB V.2 November 2007 Final SH, KPB KPB KPB V.3 May 2009 Final revised JO, KPB RL KPB

Disclaimer:

This document was prepared to inform the development of the Better Bays and Waterways Water Quality Improvement Plan and does not necessarily reflect the views of EPA . EPA Victoria makes no express or implied guarantees or representations to any third party as to the contents of this document. EPA Victoria accepts no liability for any reliance by any third party on the information detailed in this document.

Joint EPA and ASR Report No. 2008-EPA1 May 2009

This document was prepared to inform the development of the Better Bays and Waterways Water Quality Improvement Plan. Recommendations included in these interim reports may have been superseded by actions contained in the final plan.

Prepared by

R. Lee, K. Black and J. Oldman

PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

EXECUTIVE SUMMARY

A fully integrated suite of receiving water quality models of Port Phillip Bay and Western Port was developed for the Better Bays and Waterways (BBW) program at the request of Environment Protection Authority of Victoria (EPA Victoria) and Water. Four coupled models were created and installed on the EPA (Vic) computer with other baywide supporting simulations and a comprehensive range of graphical and data analysis tools. All software came from the package ‘The 3DD Suite™ of Marine and Freshwater Numerical Models’. The fourfold model coupling involved: • a dynamic catchment model (defining freshwater, nutrient and mud inputs) • three-dimensional salinity and temperature stratified hydrodynamics (for tidal, wind and density-induced circulation) • dispersal (for effluent, sediment transport and pollutant dispersal) • primary production (for nutrient uptake, phytoplankton and zooplankton). The models covered Port Phillip Bay, Western Port and northern . Applications ranged from the impacts of nutrients released into Port Phillip Bay to mud transport modelling in flood events. Global climate change predictions were also made in this broad-ranging study, which is the first environmental overview since the CSIRO Port Phillip Bay Environmental Study in the mid-1990s. In the current study, Western Port is also treated. The models simulate the behaviour and dispersal of key water quality indicators as prescribed by the Department of Water, Environment, Heritage and the Arts for BBW. The indicators include nutrients (total nitrogen and total phosphorus), chlorophyll-a (Chl-a), suspended solids, salinity, toxicants, pathogens and litter. The models address catchment loading in context with in-bay processing through dispersion and exchange, and ecosystem assimilation. The indicators used in this report are by no means an exhaustive assessment of marine condition in terms of broader ecosystem issues. Applications ranged from the mapping of averaged indicators over a two-year simulation period to specific footprints representing event impacts. The 2004–05 simulation period took into account a major nitrogen reduction improvement at the Western Treatment Plant in November 2004 and a significant (1-in-100-year) flood event in February 2005. This simulation period also represented prolonged drought conditions (since 1998) that differ vastly from the modelled (and observed) conditions during the CSIRO Port Phillip Bay Study in the early 1990s, when climate was significantly wetter and cooler. Model scenarios that incorporated global climate change predictions for 2030 and 2070 were run and compared with current conditions to identify regions of heightened vulnerability to climate impact. This report deals with the scenario modelling (Beaches Stage 15), which builds upon the development phase (Beaches Stage 7, see the first three reports in this series) to inform the Port Phillip Bay and Western Port Water Quality Improvement Plan on current and projected conditions. Four aspects were identified for this stage: • assessment of recent (2004–05) water quality conditions in Port Phillip Bay and Western Port • future growth and climate change scenarios • nested studies for priority beaches identified in the companion sampling program (Beaches Stage 9, EPA 2008b) • investigation of key issues for each bay through specialised modelling studies. These modelling tasks incorporate daily catchment flows and water quality loadings generated by the PortsE2 catchment model that enter the bays via 42 discrete points and diffuse discharge locations. Comparisons to in-situ data are clear and the increased spatial resolution of the models enhances the understanding of load dispersion, settlement and decay within the bays. Resultant dispersion patterns of pollutants indicate an overwhelming influence of wet-weather events to transport catchment loads to the bay. Long-term periods of wet and dry conditions, as observed during 1992–96 and 1997–2007 respectively, can cause vastly different dispersion patterns for pollutants discharged to the bay. Climate change and future growth scenarios incorporate 2030 and 2070 projections for rainfall and evaporation in the Melbourne region (using the CSIRO Mk3 A1F1 medium-sensitivity scenario from Whetton and Power 2007) and the Department of Sustainability and Environment’s (DSE) 2030 planning scheme for Melbourne. The 2030 scenarios were adopted by the PortsE2 climate change scenario modelling for the Better Bays and Waterways decision support system. There are significant differences in ocean flushing between Port Phillip Bay (approximately one year) and Western Port (up to about a month), so initial studies isolate the effect of shifting the

1 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

rainfall/evaporation balance in Port Phillip Bay. To enable this assessment the models are developed to incorporate an evaporation scheme and calibrated to in-situ salinity data. Models are compared between wetter conditions from the 1990s and projected dry conditions for 2030 and 2070, and mark a two to three PSU shift in salinity conditions from hyposaline (less salty than ocean) to hypersaline (more salty than ocean), which significantly alters the circulation and flushing of the bay. Comparing the extended drought conditions of 2004–05 with 2030 and 2070 projections, there are more subtle changes to salinity and temperature. Differences in pollutant dispersal manifest as constrained footprints from the major discharge points. Projected simulations for Western Port indicate significant increases in temperature (Western arm) and salinity (North-East arm). Companion dispersion modelling shows reduced loads increase the coastal concentration of pollutants near the discharge points and cause a subsequent downstream decrease in concentrations. Beach scenarios assess the potential of impacts at priority beaches from the PortsE2 discharge points. A nested 30 m grid model for the -Queenscliff region confirms observations that source waters from Swan Bay and ‘The Cut’ impact the beach during dry weather conditions. An investigation of the various pathogen sources in this region (bird roosting in Swan Bay, marina discharges, ferry operations) is recommended to assist in determining a mitigation strategy. The Rye-Rosebud and Altona–St Kilda priority beach regions are assessed for prevailing conditions and pathogen dispersal. In the northern region, coastally trapped waters disperse discharges alongshore, accumulating pathogens at the two priority beach regions of Altona and St Kilda. In the SE region, poor circulation in the Rye–Rosebud area causes pathogens to accumulate alongshore in the vicinity of the discharge points. This modelling highlighted the priority beaches as vulnerable regions impacted by accumulative coastal discharge. Configured for a range of local threat sources, further modelling could prioritise and track proposed mitigation strategies. Focus studies are undertaken to assess plankton–nutrient responses in Port Phillip Bay and sediment resuspension in Western Port. Phytoplankton response to current and projected conditions is explored using a coupled nutrient– phytoplankton–zooplankton model (NPZ). Further refinements of the NPZ model have advanced this into a highly evolved tool, allowing dynamic interaction of biota and nutrients within a complex system. While current and projected 2030 scenarios show similar patterns, the model indicates that significant increases in nutrients can lead to substantial changes in Chl-a levels. It has highlighted the need to investigate further the sensitivity of thresholds above which the bay could be subject to substantial and numerous algal outbreaks, particularly during favourable spring and summer conditions. The sediment dispersion model is augmented to include wave mixing and in-bay processes that are able to resuspend and transport sediment originally discharged from the catchment. Significant increases in suspended material occur across the system, with a strong preference for heightened concentrations in the eastern arm. This emphasises the importance to couple both catchment and receiving water dynamics when assessing the movement of pollutants in the shallow waters of Western Port. The modelling throughout the study, from the hydrodynamics through to the NPZ work, has challenged both the models and the modellers. The calibration results justify this effort and provide a solid basis for future management of Victoria’s waterways. It is recommended that the modelling tools, coupled with an appropriate catchment model, are incorporated into the Water Quality Improvement Plan to track and project changes associated with management actions. The models should be updated with more complex climate change projection data from CSIRO to improve resolution and certainty in water quality forecasts. To improve confidence in modelling projections for Western Port it is recommended that high-quality salinity, temperature and flow calibration data is collected at key sites to capture spatial and temporal gradients. The results of this study indicate that Port Phillip Bay conditions have been altered from hypersalinity. In the past and around the time of the CSIRO PPBES, the net flow through the entrance was outbound. This was fed by rainfall and freshwater run-off that exceeded the evaporation from Port Phillip Bay. Now, with reduced rainfall, recycling of water and reduced river flows, Port Phillip Bay may have shifted from net outbound to net inbound flow. If so, this is a fundamental transition in the state of the bay, which will lead to substantial increases in the salinity within Port Phillip Bay. Much of Port Phillip Bay has been hypersaline since 1998. Moreover, in the bay with its very long residence time (more than one to two years), the net evaporation in draws more saline water through the entrance from the sea. The salinities rise correspondingly as the evaporation selectively removes the fresh water, leaving the sodium chloride behind. This is like using salt water to replace evaporated water in a swimming pool. Eventually, the pool becomes very salty.

2 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

To understand this important process and to know whether Port Phillip Bay has flipped to be net inbound at the entrance, it is recommended that further data collection and modelling be undertaken. The implications are very substantial for the ecosystem and fisheries of Port Phillip Bay and for the future of our most important waterway.

3 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

TABLE OF CONTENTS EXECUTIVE SUMMARY ...... 1 TABLE OF CONTENTS ...... 4 TABLE OF FIGURES ...... 5 TABLE OF TABLES ...... 6 1 INTRODUCTION ...... 7 1.1 Background to the project ...... 7 1.2 Reports in the series ...... 8 1.3 The 3DD Suite ...... 8 1.4 Work plan ...... 9 2 NUMERICAL MODELLING OF THE BAYS 2004-05 ...... 11 2.1 Salinity in the bays ...... 11 2.2 Port Phillip Bay hydrodynamic simulations ...... 11 2.3 Western Port hydrodynamic simulations ...... 14 2.4 Pol3DD dispersion modelling ...... 16 2.5 Port Phillip Bay Pol3DD simulations ...... 16 2.6 Western Port Pol3DD simulations ...... 18 3 CLIMATE CHANGE SCENARIOS ...... 20 3.1 Bay conditions ...... 20 3.2 Catchment scenarios ...... 22 4 POLLUTANT DISPERSION 2030 SIMULATIONS ...... 26 4.1 Scenario for adjusted bathymetry ...... 27 4.2 Extreme 2070 climate change scenarios ...... 32 5 NESTED BEACH SCENARIOS ...... 35 5.1 Priority beaches of Port Phillip Bay ...... 35 5.2 Nested study of Swan Bay–Queenscliff ...... 35 5.3 Study of Altona–St Kilda and Rye–Rosebud regions ...... 38 5.4 2030 simulations ...... 40 5.5 POL3DD/WGEN resuspension modelling for Western Port ...... 42 6 RECOMMENDATIONS AND CONCLUSIONS ...... 44 7 ACKNOWLEDGEMENTS ...... 45 8 REFERENCES ...... 46 APPENDIX 1 PROJECT SERVICES (PROJECT BRIEF) ...... 47 Project requirements ...... 47 General model requirements ...... 47 Specific requirements of the modelling software ...... 47 Installation of the software and setup of the hydrodynamic model ...... 48 Training and model handover ...... 49 What EPA will supply...... 49 Key deliverables...... 49 APPENDIX 2: GREENHOUSE 2007 POSTER PRESENTATION...... 50

4 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

TABLE OF FIGURES Figure 1.1: Dr. Randall Lee and Nigel Nichols investigate a small sewage outfall discharging at the Nepean Peninsula west of Sorrento. . 7 Figure 2.1: Long-term salinity from central sites in Port Phillip Bay (#1229; upper panel) and Western Port (#716; lower panel) show inter-annual trends with respect to adjacent Bass Strait water (Source EPA Fixed Site network)...... 11 Figure 2.2: Residual velocity, mean temperature and salinity in Port Phillip Bay for 2004–05...... 12 Figure 2.3: Model runs for 2004-05 showing changes in salinity associated with baseline conditions and various wet-weather events. . 13 Figure 2.4: Mean temperature and salinity in Western Port for 2004...... 14 Figure 2.5: Model runs for 2004–05 showing changes in salinity associated with baseline (top) and wet-weather (bottom) conditions. . 15 Figure 2.6: Segments defined by the state environment protection policies for Port Phillip Bay (schedule F6) and Western Port (schedule F8) were designed to represent like regions of impact and marine character...... 16 Figure 2.7: Mean 2004–05 plots of TSS concentration and TSS settled particles in Port Phillip Bay...... 17 Figure 2.8: Mean 2004–05 plots of Toxicant and Pathogens concentrations in Port Phillip Bay...... 18 Figure 2.9: Two TSS event plots that contrast low and high catchment loadings (days 84 & 259 refer to salinity plots in Figure 2.3), indicate dominant loading along the east coast...... 18 Figure 2.10: Mean plots of TSS, TSS settled, toxicants and pathogens for the period 2004 in Western Port...... 19 Figure 3.1: Long-term salinity and chlorophyll-a from Port Phillip Bay (Hobson Bay site#1991) indicates sensitivity of phytoplankton to hypersaline conditions in the bay. (Source EPA Fixed Site network)...... 20 Figure 3.2: Port Phillip Bay is shown with a linear salinity gradient to Bass Strait as (a) measured as ‘fresh’ during the 1992–96 study (Harris et al 1996) and (b) mapped hypersaline conditions in Port Phillip Bay with respect to Bass Strait from 3–12 January 2007...... 21 Figure 3.3: Difference in bay salinity between (a) 2004 and projected climate change by 2030 in Port Phillip Bay (~0.2 PSU) using ‘fresh’ bay as the initial condition for both runs and (b) when the 2030 scenario is coupled with initial hypersalinity (~3 PSU)...... 21 Figure 3.4: Flushing of Port Phillip Bay (defined by the number of conservative particles remaining within Port Phillip Bay) after four months is shown for existing conditions in 2004 and projected hypersaline conditions in 2030...... 22 Figure 3.5: Cross-sectional view from north (right – Yarra ) to south (left – Bass Strait entrance) in the bay for the (a) fresh (hyposaline) and (b) 2030 hypersaline condition. Salinity contours shown with current vectors, suggest hypersalinity reverses the exchange pattern with Bass Strait...... 22 Figure 3.6: Mean Port Phillip Bay salinity, temperature and circulation conditions for 2030 compared to 2004–05...... 24 Figure 3.7: Mean Western Port salinity, temperature and circulation conditions for 2030 compared to 2004...... 25 Figure 4.1: Mean Port Phillip Bay pollutant conditions (toxicants, pathogens and TSS) for 2030 (left) compared to 2004–05 (right)...... 26 Figure 4.1: Mean Western Port pollutant conditions for 2030 compared to 2004...... 27 Figure 4.3: Mean bottom velocity during the hypersaline period compared for 2030 scenarios without and with (+CDP) the deepened entrance and southern channel...... 29 Figure 4.4: Mean surface velocity during the hypersaline period compared for 2030 scenarios without and with (+CDP) the deepened entrance and southern channel...... 30 Figure 4.5: Mean salinity during the hypersaline period compared for 2030 scenarios without and with (+CDP) the deepened entrance and southern channel. Increased oceanic flushing is evident in the circled region near Dromana...... 31 Figure 4.6: Salinity variations associated with observed climate variance (1994 and 2007) and projected increases for 2030 and 2070 scenarios (based on CSIRO Mk3, A1FI medium-sensitivity projection for Melbourne from Whetton and Power 2007)...... 32 Figure 4.7: Comparison of suspended sediments in Port Phillip Bay for modelled conditions in 2004 and the climate change scenario for 2070 (CSIRO Mk3, A1FI medium -sensitivity projection for Melbourne from Whetton and Power 2007)...... 33 Figure 4.8: Modelled 2070 salinity compared with conditions when projected scouring to 25 m at the entrance dredgeworks are included...... 34 Figure 5.1: (a) Swan Bay nested 30 m model grid and (b) 15-day (spring-neap period) residual velocity in the Queenscliff focus region. 35 Figure 5.2: The Queenscliff region has a number of sources (marina, bird roosting in Swan Bay, dredge and ferry operations) to consider that may impact the beach (to the south). Image courtesy of Google Earth, photo courtesy of A Stephens...... 36 Figure 5.3: Enterococci sampling during January 2007 along Queenscliff Beach and through The Cut (EPA unpublished data)...... 36 Figure 5.4: Mean pattern of pathogen dispersal during the modelled 15-day spring-neap period shows a clear connection for a pollutant source in Swan Bay to impact the Queenscliff beach...... 37

5 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 5.5: The 15-day (spring-neap period) residual velocity and mean pathogen dispersal in the St Kilda–Altona focus region...... 38 Figure 5.6: The 15 day (spring-neap period) residual velocity and mean pathogen dispersal in the Rye–Rosebud focus region...... 39 Figure 5.7: 3DDLife modelled Chl-a levels for 2004, 2030 and 2030 x 3 (nutrients) at and Long Reef...... 41 Figure 5.8: 3DDLife Chl-a anomaly (2004 to 2030) for central, Long Reef and Hobsons Bay sites...... 42 Figure 5.9: TSS with inclusion of wave mixing and sediment resuspension significantly increases concentrations and highlights observed maxima in the eastern section of Western Port...... 43

TABLE OF TABLES Table 1: Sequence of modelling Tasks undertaken for scenario studies ...... 10 Table 2: 2004–05 averages of fixed site and beach sampling data ...... 17 Table 3: High-risk segments in Port Phillip Bay compared for key threat sources (from EPA 2008) ...... 17 Table 4: 2004–05 averages of fixed-site and beach sampling data...... 18 Table 5: Scenario 1 (current) to Scenario 7 (2030) ...... 23 Table 6: Scenario 5 (current + management intervention) to Scenario 6 (current + management intervention + CC) ...... 23 Table 7: Scenario 13 (2030 + management intervention) to Scenario 14 (2030 + management intervention + CC) ...... 23 Table 8: Port Phillip Bay PortsE2 catchment model scenarios for current and 2030 (1 v 7) ...... 23 Table 9: Ports E2 catchment model scenarios 2030 + CC (13 v 14) for Port Phillip Bay...... 23

6 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

1 INTRODUCTION

1.1 Background to the project The need for a ‘receiving water quality model’ arose as part of a joint agreement between EPA Victoria and to develop an integrated Water Quality Improvement Plan (WQIP) for Port Phillip Bay, Western Port and the surrounding rivers and creeks of the region. The goal was to support the developments already undertaken by Melbourne Water with the eWater catchment model ‘PortsE2’ (Argent 2006; MW 2007) by developing a full suite of receiving water quality models. ASR Ltd was commissioned to undertake that task. In addition, EPA Victoria runs a series of monitoring programs assessing the water quality of rivers and the bays, along with ecological conditions in the freshwater streams. Their goal is to objectively monitor environmental health and protect the shared natural resources of Victorian waters (Figure 1.1). The coupled-catchment and receiving waters models developed in the WQIP provide an assessment tool to augment and progress the current understanding of environmental health that has been guided by the monitoring records.

Figure 1.1: Dr. Randall Lee and Nigel Nichols investigate a small sewage outfall discharging at the Nepean Peninsula west of Sorrento.

The key aims of the receiving water modelling study are summarised as being to: • create and calibrate the 3DD Suite of numerical marine models for Port Phillip Bay and Western Port, with linkage to an existing catchment model • develop models that simulate the hydrodynamic behaviour, nutrient dynamics, wave climate and sediment transport of Port Phillip Bay and Western Port • transfer the licensed models to EPA’s computer for their use • provide training in all aspects of the 3DD Suite necessary to ensure that the model, once set up and calibrated, can be operated ‘in house’ at EPA offices in Melbourne • provide user manuals for the software • document the configuration and other factors used to customise the 3DD Suite for Port Philip and Western Port. Further detail on the key objectives that guided the project development are shown as an excerpt from the contract agreement in Appendix 1. Four coupled models were created and installed on the EPA computer with other baywide supporting simulations and a comprehensive range of graphical and data analysis tools. All software came from the package ‘The 3DD Suite of Marine and Freshwater Numerical Models’. The fourfold model coupling involved • a dynamic catchment model (defining freshwater, nutrient and mud inputs) • three-dimensional salinity and temperature-stratified hydrodynamics (for tidal, wind and density-induced circulation)

7 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

• dispersal (for effluent, sediment transport and pollutant dispersal) • primary production (for nutrient uptake, phytoplankton and zooplankton). The models covered Port Phillip Bay, Western Port and northern Bass Strait. Applications ranged from the impacts of nutrients released into Port Phillip Bay to mud transport modelling in flood events. Global climate change predictions were also made in this broad-ranging study, which is the first environmental overview since the CSIRO Port Phillip Bay Environmental Study in the mid 1990s. In the current study, Western Port is also treated. 1.2 Reports in the series For this large modelling study, three additional reports were produced that covered the breadth of physical and biological processes, from the hydrodynamics to primary production. The other reports in the series are: • Port Phillip and Western Port Receiving Water Quality Modelling: Hydrodynamics (2007). ASR Technical Report No. 2007–EPA1. Authors: S Harrison, K Black and C Bosserelle. Published as EPA publication 1377. • Port Phillip and Western Port Receiving Water Quality Modelling: Lagrangian dispersal (2007). ASR Technical Report No. 2007–EPA2. Authors: S Harrison, K Black and S Bosserelle. Published as EPA publication 1378. • Port Phillip and Western Port Receiving Water Quality Modelling: Nutrient dynamics and primary production (2007). ASR Technical Report No. 2007–EPA3. Authors: S Harrison, K Black and S Bosserelle. Published as EPA publication 1379. The establishment, methods adopted and calibration of the models are presented in the above reports. One of the aims of developing the receiving water model is to understand impacts of future urban growth and climate change impacts, and to support the development of decision strategies for the implementation of a WQIP. This report presents both current and future scenarios in the two bays using the models developed during Stage 7 (see Appendix 1) to address the requirements of Beaches Stage 15. Scenarios addressing climate change, future growth and high-risk beaches were highlighted to assess selected water quality parameters for regions of vulnerability and change. The group also recognised the need for further investigation on nutrient-phytoplankton dynamics in Port Phillip Bay and sediment re-suspension in Western Port, as they are the respective overriding water quality issues in the region. This report gives details of the modelling of scenarios that integrate catchment model scenario outputs and inform the relevant Better Bays and Waterways programs. 1.3 The 3DD Suite ASR Ltd provides modelling services using in-house modelling software that has undergone over 30 years of development. The software — known as the 3DD Suite of Numerical Models — is a commercially available software package that is applicable to a broad range of hydrodynamic conditions. This includes inland waterways (rivers, estuaries and ports), beaches and embayments, the continental shelf or large expanses of open-ocean. The model 3DD has been successfully applied and verified in a diverse range of situations (Black 1987; Black 1989; Black and Gay 1991; Black et al. 1993; Young et al. 1994; Middleton and Black 1994). The model has been previously applied to investigate the parameters responsible for eddy formation behind and reefs (Black and Gay 1987; Black 1989; et al. 2000). Other example applications are beach erosion studies, numerical modelling of oil spills, tracking floating debris, assessing ocean outfalls, and ocean/atmosphere dynamics. Applications in Victoria The 3DD Suite models of Port Phillip Bay, Western Port and Bass Strait were first developed in the late 1980s and have been used many times in Victoria. Each time more data becomes available to improve the models, and computers are always improving, making it possible to do more sophisticated modelling, such as in the present study. Calibrated, verified and published in the peer-reviewed scientific literature for more than two decades, the models have also been applied in Victoria in the modelling of nutrients from the Werribee and Gunnamatta sewage treatment plants, dispersal of nutrients and E. coli from the drains around Port Phillip Bay, sediment transport studies in the Great Sands, wave-driven sediment transport studies of Port Phillip Bay beaches, wave-generation modelling, inland waterways, mixing studies in Docklands and three-dimensional salinity-stratified modelling studies of the Yarra Plume (for example, Keough and Black 1996; Black and Hatton 1994; Sokolov and Black 1999). Other applications of the models in Victoria include ship-borne dispersal of bilge water contaminants in Bass Strait, coastal-trapped wave and tidal circulation of Bass Strait, extreme water levels, salinity intrusion in the Lakes and numerous sediment transport and larval studies covering the broader region from the South Australian Bight to the NSW coast and (Black 1990; Black et al. 1990; Black and Hatton 1992).

8 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

For example, in 20 years of collaboration with Dr Greg Jenkins of DPI (MAFFRI), modelling with the 3DD suite has provided essential input for a succession of scientific papers, technical reports and research studies to understand the early life history of the King George Whiting (Jenkins et al. 1997; Jenkins et al. 1999). 1.4 Work plan To undertake this phase of modelling a stepwise approach was required to design and manage the tasks, to ensure efficiency of the modelling operation and that appropriate tests and data are available to begin the subsequent task. Calibrated catchment flows from PortsE2 are needed to feed into hydrodynamic model runs, which are subsequently employed by the pollutant dispersal and nutrient–phytoplankton–zooplankton models. They are also used to set boundary conditions for nested, small-scale models (whose outputs drive the nested simulations for pathogens). The general work plan is outlined in the following sequence. Current conditions — utilising two-year runs for the period 2004–05 Hydrodynamic models are updated with an evaporation/rainfall scheme to provide calibrated sensitivity for observed fluctuations in bay salinities. Catchment model outputs are recalibrated to gauge records across the region. The pollution dispersal model uses the calibrated PortsE2 load data and hydrodynamic time series to assess water quality parameters for average and event conditions during the simulation period. Climate change scenarios (based on 2030 projections of climate change and future growth adopted in PortsE2 scenarios [WBM 2007b]) Hydrodynamic models adopt an altered evaporation/rainfall scheme to account for projected changes in climate. Catchment flow 2030 scenarios for climate change and future growth are incorporated in the hydrodynamics and rerun for the 2004–05 period. Nested beach scenarios (to assess beaches of high priority identified in the Beaches interim WQIP program) Small-scale grids were nested using boundary conditions from the larger model runs to simulate hydrodynamics. A dispersal model tested a variety of pathogen sources in the prevailing circulation to assess beach impact. Focus studies (on key issues for the two bays) The nutrient–phytoplankton–zooplankton model for Port Phillip Bay was updated with additional nutrient source data and calibrated against cross-referenced data records (fixed-sites sampling and moored instruments). A sediment transport/settlement model for Western Port was augmented with time series wave data to account for resuspended transport and resettlement within the bay. Table 1 identifies the modelling tasks, number of modelling runs and relative efficiencies (ratio of model time to simulated time). In summary, there were 33 model runs to undertake and, as some hydrodynamic runs can take up to 12 days (for example, WP400m runs), the tasks were shared between ASR and EPA. EPA has attempted to undertake operational runs where possible, to allow ASR time to focus on model development and calibration. This required that ASR was able to provide EPA with new tested code as it became available.

9 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Table 1: Sequence of modelling Tasks undertaken for scenario studies Port Phillip Bay Model Western Port Model Comment Model/scenario runs Model/scenario runs Grid Grid Efficiency Efficiency 1. 2004–05 runs 3DD/Hydrodynamics PPB800m 1 3DD/Hydrodynamics WP400m 1 ASR updated 3DD model program to 1:140 1:60 incorporate rainfall/evaporation scheme Runs done by EPA Pol3DD/Dispersals(Tox, Path, TSS) 4 Pol3DD/Dispersals(TN, TP, Tox, Path, 5 Pol3DD runs done at EPA 800m TSS) Various 400m Various 2. Climate change 3DD/2030 Climate change 3 3DD/ 2030 Climate change 3 Altered rainfall/evap scheme applied to 3DD 800m 400m (for CC projections) Runs jointly by ASR/EPA to 1:140 1:60 check sensitivity Pol3DD (Tox, Path, TSS) + PortsE2 4 Pol3DD (TN, TP, Tox, Path, TSS) + 5 Pol3DD runs done at EPA 2030 CC/Growth scenarios PortsE2 CC/Growth scenarios 800m 400m 1:200 1:150 3. Nested beaches 3DDnest + Pol3DD /Nested Beaches 3 No priority beaches in Western Port NA Queenscliff, St. Kilda-Altona, and Rye-Rosebud 30, 50 & 200m nests done by ASR/EPA 1:120 4. Focus studies 3DDLIFE/NPZ (Nutrients & 1 Pol3DD + WGEN/Sediment res- 1 ASR adding atmospheric flux (from EPA data) PhytoPLANKTON and zooplankton) suspension model into NPZ model runs. ASR incorporate shading PPB3200m 400m effect into 3DDLife model using turbidity from 1:600 1:200 fixed sites (or PoMC data).

ASR Run WGEN for waves as input to Pol3DD 3DDLIFE/ 2030 Climate Change 2 NA NA Using climate change hydrodynamic file are projections (Nutrients & supplied 3DDLife runs done by ASR phytoplankton and zooplankton) 3200m 1:600

10 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

2 NUMERICAL MODELLING OF THE BAYS 2004-05

2.1 Salinity in the bays Long-term records in Port Phillip Bay and Western Port (Figure 2.1) show strong seasonal and interannual variations in salinity that can be both above or below the adjacent waters of Bass Strait. There may be also a trend for increasing salinity since 1998, caused by low rainfall, higher evaporation and reduced inflows due to diversion of water by recycling and other human uses. Much of Port Phillip Bay has been hypersaline since 1998. Differences in salinity are likely to further affect the exchange efficiency with Bass Strait, because of the altered density-driven circulation. This would change the flushing capacity of the bays. Moreover, in Port Phillip Bay with its very long residence time (more than one to two years), the net evaporation in the bay draws more saline water through the entrance from the sea and the salinities correspondingly rise as the evaporation selectively removes the fresh water, leaving the sodium chloride behind. Catchment discharges differ significantly both seasonally and interannually and are proportional to the salinity patterns observed in the bays. Of concern are the climate change predictions that suggest more evaporation and less rainfall and therefore increasing salinity in Port Phillip Bay. The Model 3DD schemes to simulate rainfall/evaporation were validated in Report 1 (Hydrodynamics) of the series (2007–EPA1). Two-year 2004–05 hydrodynamic model runs were undertaken with 3DD using an 800 m, eight-layer grid for Port Phillip Bay and 400 m, six-layer grid for Western Port, as developed in the Stage 7 Beaches interim projects (2007–EPA1). Here, the results coming from the models are considered, for both present and future scenarios.

Figure 2.1: Long-term salinity from central sites in Port Phillip Bay (#1229; upper panel) and Western Port (#716; lower panel) show inter-annual trends with respect to adjacent Bass Strait water (Source EPA Fixed Site network).

2.2 Port Phillip Bay hydrodynamic simulations Models show average conditions for 2004–05 for residual flow, temperature and salinity in Figure 2.2. Reference is also made to the marine segments from the WQIP risk assessment (EPA 2008), to indicate regions that are poorly flushed and experience temperature and salinity extremes.

11 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 2.2: Residual velocity, mean temperature and salinity in Port Phillip Bay for 2004–05.

A series of events during the 2004–05 series is shown in Figure 2.3 for salinity. Although Port Phillip Bay is hypersaline for the majority of the two-year period, wet-weather events bring significant reductions in salinity that track along the east coast to the Great Sands region.

12 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 2.3: Model runs for 2004-05 showing changes in salinity associated with baseline conditions and various wet-weather events.

13 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

2.3 Western Port hydrodynamic simulations Model results are shown as average conditions for 2004 for residual flow, temperature, and salinity in Figure 2.4. Although models were run for the entire 2004–05 period, poor salinity calibration data to assess accurate evaporation flux (especially for and areas of high turbidity) limited further model development to account for evaporative losses, and hence confidence in the 3D model stability. Detail on salinity calibrations can be found in the hydrodynamic report from this series. A series of events during the 2004–05 period is shown in Figure 2.5 for salinity.

Figure 2.4: Mean temperature and salinity in Western Port for 2004.

14 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 2.5: Model runs for 2004–05 showing changes in salinity associated with baseline (top) and wet-weather (bottom) conditions.

15 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

2.4 Pol3DD dispersion modelling Dispersion modelling for Port Phillip Bay and Western Port was run using the two-year hydrodynamic data from the respective Port Phillip Bay 800 m and WP400m 3DD model runs. Results are presented as mean concentrations for the two-year period. Significant event results are also shown for key parameters. Discrete marine segments are used by state environment protection policy to reflect catchment loading and bay circulation patterns. These eight segments are monitored at nine long-term sampling sites. They are provided in Figure 2.6 for reference to the patterns generated by the dispersion model.

Figure 2.6: Segments defined by the state environment protection policies for Port Phillip Bay (schedule F6) and Western Port (schedule F8) were designed to represent like regions of impact and marine character.

2.5 Port Phillip Bay Pol3DD simulations Pol3DD was used to simulate total suspended solids (TSS), pathogens and toxicants in Figure 2.7 using the catchment PortsE2 data prepared for the 2004–05 period (WBM 2007a). TSS is shown for total number of particles settled over the two years and as a mean concentration of material in suspension. Events for TSS are also shown for snapshot periods matching the salinity time series in Figure 2.8. Events predominantly distribute the TSS load along the bay’s east coast, due to the SW wind conditions that prevail during winter–spring storms. Extreme wet-weather events in summer are often correlated with NW or SE winds. Table 2 provides the two-year averages of measurements from EPA’s fixed sites (refer Figuire 2.6) and Beach Report programs, to compare with the mean plots in Figure 2.7. Relative differences are matched by the model results. Pathogens and toxicants are constrained to the coastal boundary layer, which is driven by the predominant clockwise circulation of the bay (combining dominant SW winds and Yarra outflow).

16 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Table 2: 2004–05 averages of fixed site and beach sampling data

Water quality parameter Central Hobsons Werribee Dromana Corio #1229 #1991 #369 #1282 #1911 Non-filterable residue (mg/L) 2.8 5.2 4.8 5.1 4.2 Enterococci (75%ile) (cfu/100ml) NA 217(St.Kilda) 52 52 55 (Eastern Beach) Chl-a mg/m3 0.7 2.2 1.5 0.7 0.75 TN g/m3 0.142 0.218 0.246 0.142 0.187 TP g/m3 0.069 0.088 0.145 0.059 0.103 NOx g/m3 0.007 0.023 0.044 0.00005 0.001 NH3 g/m3 0.005 0.007 0.015 0.0005 0.005 Tox (Pb & Zn) NA NA NA NA NA

Reference to the WQIP marine risk assessment in Table 3 (EPA 2008), the high-risk segments are identified predominantly by catchment-related threat sources, and this is mimicked in the mean distribution maps of pollutants in Port Phillip Bay in figures 2.8 and 2.9).

Table 3: High-risk segments in Port Phillip Bay compared for key threat sources (from EPA 2008)

Hobsons # Corio # Werribee # Sth Eastern # River discharge 24 Large point source 24 Large point source 12 Small aggregated 48 point sources Lack of flushing 18 Small aggregated point 24 Small aggregated 12 Septic tanks 30 sources point sources Small aggregated point 18 Lack of flushing 12 sources Coastal infrastructure 17 Litter 18

Figure 2.7: Mean 2004–05 plots of TSS concentration and TSS settled particles in Port Phillip Bay.

17 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 2.8: Mean 2004–05 plots of Toxicant and Pathogens concentrations in Port Phillip Bay.

The baseline and wet-event condition shown in Figure 2.9 for TSS highlights how flood events provide the overwhelming catchment load to Port Phillip Bay.

Figure 2.9: Two TSS event plots that contrast low and high catchment loadings (days 84 & 259 refer to salinity plots in Figure 2.3), indicate dominant loading along the east coast.

2.6 Western Port Pol3DD simulations Pol3DD was used to simulate total suspended solids (TSS), pathogens, toxicants and nutrients in Western Port, using the catchment PortsE2 data in Figure 2.10. Results indicate a concentration of suspended sediments in the eastern arm of Western Port, and dominant discharge of pollutants out of the western arm with enhancement in coastal areas adjacent to discharge points. Table 4 provides the two-year averages of measurements from the EPA fixed sites (refer Figure 2.6) and WQIP Interim Beaches programs, to compare with the mean plots in Figure 2.10. Coastal concentration of pathogens is evident in model results around Merrick (Mornington coast) and Cowes (NE Phillip ), and maximum TSS concentrations occur in the East Arm near the Corinella #724 sample site.

Table 4: 2004–05 averages of fixed-site and beach sampling data.

Water quality parameter Central#716 Hastings#709 Corinella#724 Merricks Cowes TSS non-filterable residue 5.1 6.2 38.4 (mg/L) Enterococci (75%ile) NA NA <10 25 18 (cfu/100ml)

18 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Tox (Pb & ZN)s NA NA NA NA NA Chl-a 1.10 1.11 2.94 TN g/m3 0.160 0.137 0.278 TP g/m3 0.008 0.011 0.029 NOx g/m3 0.001 0.003 0.015 3 NH3 g/m 0.003 0.004 0.005

Figure 2.10: Mean plots of TSS, TSS settled, toxicants and pathogens for the period 2004 in Western Port.

19 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

3 CLIMATE CHANGE SCENARIOS

3.1 Bay conditions Port Phillip Bay provides an interesting case for the manifestation of climate change in a temperate marine system. The bay, although large in area, is constrained by a relatively small entrance and extensive neighbouring sand banks. In combination, tidal volumes are reduced by more than 90 per cent, resulting in long residence times within the bay of a year or more. As a result, the density structure of this large bay is dominated by the balance between surrounding catchment inflows and within-bay evaporative fluxes. With a majority of catchment loading at the northern end of the bay, a linear density gradient typically progresses to the open waters at the southern end (the hyposaline condition). Since 1947, Port Phillip Bay monitoring indicates that, during strong and prolonged periods of drought, bay salinities exceed that of the ocean (a condition known as hypersalinity) and can extend throughout the bay (Brown and Davies 1991). Unbroken records dating back to 1984 (see Figure 2.1) show that there has been a prolonged hypersaline condition since the strong 1997–98 El Niño. In contrast the adjacent embayment of Western Port experiences predominantly seasonal variations in salinity, due to greatly reduced flushing times. An influence of prolonged hypersalinity on water quality in Port Phillip Bay is shown in Figure 3.1. A comparison of salinity and chlorophyll-a data in Port Phillip Bay at Hobsons Bay suggests hypersaline conditions are associated with constrained plankton growth.

Figure 3.1: Long-term salinity and chlorophyll-a from Port Phillip Bay (Hobson Bay site#1991) indicates sensitivity of phytoplankton to hypersaline conditions in the bay. (Source EPA Fixed Site network). The hypersaline condition for Port Phillip Bay was mapped in 2007 (Figure 3.2b), when bay salinities were consistently 1 PSU higher and up to 2.5 PSU greater in the shallow western arm. Under these conditions, river plumes such as from the Yarra are quickly arrested, and mixing dynamics with the open ocean are essentially inverted, with likely changes to overall residence times. Projections of up to a 25 per cent reduction in rainfall for the region by 2070 (CSIRO 2007) as a consequence of climate change indicate the hypersaline condition for Port Phillip Bay may be expressed more frequently. The data evidence and regional climate projections motivated a detailed modelling assessment for climate change scenarios in Port Phillip Bay and attracted interest through the publication of initial findings in Appendix 2.

20 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Bay model simulations in Figure 3.3a compare a more conservative projected 2030 climate change evaporation (+4.7%) and rainfall (–2.7%) with baseline conditions in 2004. Both simulations use the same initial linear saline condition from Figure 3.2a. These highlight a strong likelihood of hypersaline conditions across the bay. Enhanced hypersalinity in Figure 3.3b occurs when the 2030 simulation is coupled with current mapped hypersalinity from Figure 3.2b as an initial condition. Under future climate change scenarios it is predicted more water will enter Port Phillip Bay from Bass Strait than under present conditions. Results from the hydrodynamic simulations predict that an extra 150 Mm3 of Bass Strait water would enter Port Phillip Bay annually. As shown in Figure 3.5, as well as a net effect on entrance fluxes, there are changes in the predicted circulation pattern at the entrance — the 2030 case shows Bass Strait inflow at the surface and bay water exchange at depth.The subtle balance between the increased net flux of Bass Strait waters into Port Phillip Bay, changes to the pattern of exchange at the entrance, decreased river flows and increased evaporation under climate change scenarios will not only change the flushing rate of Port Phillip Bay, but may lead to an increased likelihood of hypersaline conditions in the future. For example Figure 3.4 shows the rate at which conservative particles are transported out of Port Phillip Bay for both the 2030 hypersaline scenario and current conditions in the bay. This suggests that the combination of processes in Port Phillip Bay are altered under climate change scenarios and that residence time for the bay will become longer due to climate change. The rate at which salinity may increase within Port Phillip Bay will depend on how freshwater inputs to the bay are managed in the future and the subtle balance between changes to exchange between Port Phillip Bay and Bass Strait and the likely increase in evaporation into the future.

Figure 3.2: Port Phillip Bay is shown with a linear salinity gradient to Bass Strait as (a) measured as ‘fresh’ during the 1992–96 study (Harris et al 1996) and (b) mapped hypersaline conditions in Port Phillip Bay with respect to Bass Strait from 3–12 January

2007. Figure 3.3: Difference in bay salinity between (a) 2004 and projected climate change by 2030 in Port Phillip Bay (~0.2 PSU) using ‘fresh’ bay as the initial condition for both runs and (b) when the 2030 scenario is coupled with initial hypersalinity (~3 PSU).

21 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 3.4: Flushing of Port Phillip Bay (defined by the number of conservative particles remaining within Port Phillip Bay) after four months is shown for existing conditions in 2004 and projected hypersaline conditions in 2030.

Figure 3.5: Cross-sectional view from north (right – Yarra ) to south (left – Bass Strait entrance) in the bay for the (a) fresh (hyposaline) and (b) 2030 hypersaline condition. Salinity contours shown with current vectors, suggest hypersalinity reverses the exchange pattern with Bass Strait.

3.2 Catchment scenarios PortsE2 catchment flows were simulated for both 2030 climate change and future growth scenarios (WBM 2007b). Ports E2 climate change scenario modelling adopted a blanket 4.7 per cent increase in evaporation and a 2.7 per cent decrease in rainfall, based on widely varying CSIRO estimates for 2020 and 2050 (though essentially based upon the CSIRO Mk3 2030 A1F1 scenario medium sensitivity run for annual conditions in the Melbourne region). This was manifested as an average of about a 10 per cent reduction in flow from the Port Phillip Bay and Western Port catchments. Tables 5 to 7 document a range of PortsE2 catchment model scenarios undertaken by WBM (2007a).

22 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Table 5: Scenario 1 (current) to Scenario 7 (2030)

Embayment Flow TN TSS Port Phillip Bay 1.9% (1795–1829) 3% (4831–4977) 5% (79–83) Western Port 1% (547–554) 4% (875–910) 5% (20–21)

Table 6: Scenario 5 (current + management intervention) to Scenario 6 (current + management intervention + CC)

Embayment Flow (GL/yr) TN (t/yr) TSS (t/yr) Port Phillip Bay –4.2% (1795–1720) –3.2% (1915–1854) –2.9% (35–34) Western Port –9.1% (547–497) –10.6% (606–542) –16.6% (12–10)

Table 7: Scenario 13 (2030 + management intervention) to Scenario 14 (2030 + management intervention + CC)

Embayment Flow (GL/yr) TN (t/yr) TSS (t/yr) Port Phillip Bay –4.3% (1829–1750) –3.5% (1965–1897) –5.5% (36–34) Western Port –9.2% (554–503) –10.4% (616–552) –16.6% (12–10)

The adopted climate change projection was based on the spatially coarse projection data freely available at the time for the region (from CSIRO’s Ozclim model simulator, and subsequently from Whetton and Power 2007). Experimental modelling with CSIRO’s TAPM (The Air Pollution Model) has subsequently generated projection meteorological data to 5 km resolution. Additional work at CSIRO (McInnes et al) is also being done to improve scale and reliability for wind projections (both speed and direction). It is planned that this information will be incorporated into the RWQM and projections rerun to provide guidance during the implementation phase of the WQIP.

Table 8: Port Phillip Bay PortsE2 catchment model scenarios for current and 2030 (1 v 7)

Catchment Flow (GL/yr) TN (t/yr) TSS (t/yr) Yarra 1.5% (546–554) 4.8% (878–920) 3.8% (26–27) Werribee* (incl. WTP) 1.6%(861–875) 1.3% (3190–3232) -3% (33–34) Patterson 2.9% (306–315) 9.2% (512–559) 7.7% (13–14) Maribyrnong 4.4% (45–47) 5.5% (145–153) 0 % (4–4) SE total (Barwon-Bellarine) 0% (24–24) 1.6% (63–64) 0% (2–2)

Table 9: Ports E2 catchment model scenarios 2030 + CC (13 v 14) for Port Phillip Bay.

Catchment Flow (GL/yr) TN (t/yr) TSS (t/yr) Yarra –10.8% (554–494) –9.6% (560–506) –9% (11–10) Werribee* (incl WTP) –0.5% (875–871) –0.4% (917–913) 0% (18–18) Patterson –2.5% (315–307) –1.4% (363–358) 0% (5–5) Maribrynong –8.5% (47–43) –5.5% (72–68) 0% (1–1) SE total (Barwon-Bellarine) –8.3% (24–22) –3.2% (31–30) 0% (1–1)

Total results in tables 5 to 7 are skewed by inclusion of the Western Treatment Plant (WTP) in climate change scenarios, as suggested by comparing results in tables 8 and 9 for Yarra and WTP. If the WTP is removed from total Port Phillip Bay results, the respective climate change reductions are flow 7.8 per cent, TN 6.2 per cent and TSS 22 per cent. TSS differences are widely varying, as the quantities are low, so percentage uncertainties are likely to be greater. Future growth scenarios based on the ‘Melbourne 2030’ planning for land use have been used. The DSS project was unable to rerun specific future growth scenarios for 2004–05 (daily data). Net changes across the region were derived from results in the catchment modelling scenario report (WBM 2007a) The 2004–05 simulation was modified with the 2030 future growth scenario, resulting in a net two per cent increase in flows.

23 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Hydrodynamic models were run with both the adjusted in-bay 2030 evaporation/rainfall conditions and PortsE2 inputs. Results are compared to the 2004–05 model runs in figures 3.6 and 3.7 for mean salinity, temperature and circulation conditions. Due to the differences in bay depths, the differences in Port Phillip Bay are more subtle than those shown in Western Port.

Figure 3.6: Mean Port Phillip Bay salinity, temperature and circulation conditions for 2030 compared to 2004–05.

24 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 3.7: Mean Western Port salinity, temperature and circulation conditions for 2030 compared to 2004.

25 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

4 POLLUTANT DISPERSION 2030 SIMULATIONS

Using the 2030 climate change and future growth scenarios from the PortsE2 catchment modelling, new loadings were generated and input via the 42 points/diffuse lines to the dispersion models. Results for toxicants, pathogens and TSS are shown in figures 4.1 and 4.2 respectively for Port Phillip Bay and Western Port. These figures compare the 2030 scenarios with the current scenarios.

Figure 4.1: Mean Port Phillip Bay pollutant conditions (toxicants, pathogens and TSS) for 2030 (left) compared to 2004–05 (right).

26 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 4.1: Mean Western Port pollutant conditions for 2030 compared to 2004.

To focus on effects of increased evaporation, climate change patterns in Western Port were constrained to an initial six-month period when major wet-weather events were absent. Significant increases in salinity (+1 PSU in the NE arm) and temperature (+1 °C across the western arm) are apparent from Figure 3.7. Results from the Pol3DD dispersion model runs indicate the reduced loads cause an increase in coastal concentration of TSS near the discharge points and a subsequent downstream decrease in concentrations. There are negligible differences for toxicants and pathogens. 4.1 Scenario for adjusted bathymetry The recently approved and implemented Channel Deepening Project (CDP) for Port Phillip Bay has deepened the Port Phillip Bay entrance channel and the Southern channel through the Great Sands. With increased cross-section to the Bass Strait and a deeper passage through the Great Sands, an increase in tidal flushing is expected (PoMC 2007).

27 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

For hypersaline climate change scenarios this is of interest in assessing whether the channel deepening will have any offset effect on reducing the hypersaline condition (in other words, improve the flushing of the bay). To test this hypothesis, the 2030 scenario was run for six spring-neap tidal periods during baseline catchment flow conditions. The same inputs were rerun with bathymetry adjusted according to PoMC (2007) quoted dredge locations and construction depths (Figure 4.3). The mean velocities for the bottom layer shown in Figure 4.3 indicate there are differences at the eastern end of the channel where it connects to the main section of the bay. The scenario without CDP shows an isolation of flows to the west of the Great Sands from those on the bayward (eastern) side. The CDP 2030 scenario shows a net inflow and increased flow into the channel from the Great Sands (to the north). On a larger scale, the residual flow patterns through the combined southern channel and entrance show minimal differences (Figure 4.4). The resultant difference is shown in Figure 4.5 for mean salinity in Port Phillip Bay with greater flushing evident in the SE region near Dromana. In summary, the increased transparency to Bass Strait does marginally improve flushing and offset the hypersaline bay condition in the vicinity of Dromana, but is not significant enough to improve the baywide condition.

28 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 4.3: Mean bottom velocity during the hypersaline period compared for 2030 scenarios without and with (+CDP) the deepened entrance and southern channel.

29 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Dromana

Figure 4.4: Mean surface velocity during the hypersaline period compared for 2030 scenarios without and with (+CDP) the deepened entrance and southern channel.

30 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 4.5: Mean salinity during the hypersaline period compared for 2030 scenarios without and with (+CDP) the deepened entrance and southern channel. Increased oceanic flushing is evident in the circled region near Dromana.

31 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

4.2 Extreme 2070 climate change scenarios Based upon the same A1F1 medium sensitivity run for the Melbourne region (Whetton and Power 2007), 2050 and 2070 scenarios were considered in this section to provide a longer term context of climate change impacts. 2050 and 2070 conditions show, respectively, –11 and –18 per cent decreases in average annual rainfall. This is paired with respective increases in potential evaporation of +6.5 and +8 per cent. In the absence of further climate change scenario runs of the PortsE2 model, projections for 2070 were estimated based on proportional flow reductions for the 2030 scenario with respect to projections for rainfall and evaporation. Model runs were undertaken with adjusted rainfall response from the catchment model output and adjusted evaporation-rainfall balance in the 3DD hydrodynamics of the bay. Resultant plots of salinity are shown in Figure 4.6 for 2030 and 2070 scenarios compared to modelled observations for 1994 and 2007. Adjusted TSS loads from PortsE2 for a 2070 scenario were driven by the 2070 hydrodynamics to map TSS (mean log concentrations) in the bay in Figure 4.7. There are overall clear reductions in TSS concentrations for 2070 related to reduced loads predominantly from the . This is manifested not only as a reduced footprint in the Hobsons Bay region to the north, but also as significant reductions along the east and south-east coast of the bay. The western side of the bay (near Werribee) shows an increase in TSS for 2070, most likely resulting from reduced dilution and advection of the WTP discharge with increased hypersaline conditions in the bay.

1994 2007

2030 2070

Figure 4.6: Salinity variations associated with observed climate variance (1994 and 2007) and projected increases for 2030 and 2070 scenarios (based on CSIRO Mk3, A1FI medium-sensitivity projection for Melbourne from Whetton and Power 2007).

32 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 4.7: Comparison of suspended sediments in Port Phillip Bay for modelled conditions in 2004 and the climate change scenario for 2070 (CSIRO Mk3, A1FI medium -sensitivity projection for Melbourne from Whetton and Power 2007).

33 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

The more extreme salinity scenario for 2070 in Figure 4.6 was also tested for adjusted bathymetry, this time using extreme projections of scouring to 25 m at the Port Phillip Bay entrance dredging works. Plots in Figure 4.8 suggest similar increases in flushing at the Dromana site as for the 2030 + CDP scenario in Figure 4.5 and a slight increase at the entrance. Both effects are local and do little to offset the overall hypersalinity of the bay set up by the extreme–heat–low–rainfall 2070 scenario.

Figure 4.8: Modelled 2070 salinity compared with conditions when projected scouring to 25 m at the entrance dredgeworks are included.

34 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

5 NESTED BEACH SCENARIOS

5.1 Priority beaches of Port Phillip Bay From the Beaches Interim program (EPA 2008b) four beaches were identified (Queenscliff, Altona, St Kilda and Rye/Rosebud) as priorities, based on dry and wet-weather concentrations of measured pathogens (E.coli and Enterococci) from the WQIP sampling program and long-term EPA Beach Report records (EPA 2008b). The modelling program selected these beaches to run scenarios over a summer spring-neap period (approsimately 15 days). Residual velocity plots are shown to identify the prevailing net circulation in the study area. Pathogen dispersion from the PortsE2 point and diffuse-catchment discharges use a default T90 die-off of 20 hours and typically depict a coastally constrained outflow (Figure 4.1). Where required, finer scale models were developed by nesting a smaller grid within the Port Phillip Bay model grid, using the 3DDNest module from the 3DD model suite (Black 2001). 5.2 Nested study of Swan Bay–Queenscliff A 30 m grid incorporating the Swan Bay region was developed to assess likely sources that were routinely impacting the Queenscliff beach area (Figure 5.1a). The 30 m resolution was needed to define a significant artificial channel known as ‘The Cut’ that links Swan Bay to the bay. Swan Bay is a shallow marine reserve with significant populations of resident water birds roosting. Within this channel a significant marina operation also resides. The residual velocity plot for a 15-day spring-neap period (Figure 5.1) indicates the flow from The Cut, the dredge spoil ground and the ferry terminal operations all impact the priority beach (to the south of the channel entrance). Observations in Figure 5.2 and sampling in Figure 5.3 also indicate the likelihood that the plume discharged from The Cut impacts the priority beach with above-background levels of pathogens (20–80 cfu/100ml). The nested hydrodynamic model was coupled to the Pol3DD dispersion model to assess the impact of a pathogen source from Swan Bay affecting the priority beach region. Pathogens were released at the Swan Bay entrance to The Cut and run for the 15-day spring-neap period. Results in Figure 5.3 indicate the majority of the released material impacts the priority beach region to the south.

(a) (b)

Figure 5.1: (a) Swan Bay nested 30 m model grid and (b) 15-day (spring-neap period) residual velocity in the Queenscliff focus region.

35 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 5.2: The Queenscliff region has a number of sources (marina, bird roosting in Swan Bay, dredge and ferry operations) to consider that may impact the beach (to the south). Image courtesy of Google Earth, photo courtesy of A Stephens.

Figure 5.3: Enterococci sampling during January 2007 along Queenscliff Beach and through The Cut (EPA unpublished data).

36 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 5.4: Mean pattern of pathogen dispersal during the modelled 15-day spring-neap period shows a clear connection for a pollutant source in Swan Bay to impact the Queenscliff beach.

37 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

5.3 Study of Altona–St Kilda and Rye–Rosebud regions The Altona–St Kilda and Rye–Rosebud priority beach regions were also assessed for prevailing conditions and pathogen dispersal. In the northern region coastally trapped waters dispersed discharges alongshore, accumulating pathogens at the two priority beach regions of Altona and St Kilda (Figure 5.5). In the SE region, weak prevailing flows in the Rye–Rosebud area caused pathogens to accumulate alongshore in the vicinity of the discharge points (Figure 5.6).

Altona St. Kilda

Figure 5.5: The 15-day (spring-neap period) residual velocity and mean pathogen dispersal in the St Kilda–Altona focus region.

38 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Rye-Rosebud

Figure 5.6: The 15 day (spring-neap period) residual velocity and mean pathogen dispersal in the Rye–Rosebud focus region.

39 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

5.4 2030 simulations The primary production modelling was undertaken using a new set of boundary conditions that define the mass of nutrient entering the bay from the three main sources: Yarra River, WTP and . The mass (g/s) was calculated by multiplying the flows and concentrations of nutrients provided by the catchment model PortsE2 2030 scenarios and coupled with 2030 3DD hydrodynamics. In addition, a background level of 10 g/s was added to all values from the model for nitrates and 2 g/s for inorganic phosphorus. In these simulations, both the nitrogen and phosphorus availability were used to define the growth rate of the phytoplankton. That is:

NN= sum(N03+NH4) Ratio_N= NN/(NN+xkn) where xkn=0.01, the half saturation constant And the same for inorganic phosphorous (P): PP=P Ratio_P= PP/(PP+xkp) where xkp=0.01, the half saturation constant The growth scaling factor was then: SCAL= Ratio_N * Ratio_P Other scaling factors (such as light intensity) were unchanged. Phytoplankton maximal growth rate factor was set to 2.5, instead of 3 as in other runs. The non-predatory mortality factor was set to 0.01 (one per cent per day). Figure 5 shows the phytoplankton response of the 3DDLife model during 2004 and a 2030 scenario extracted for sites at Hobsons Bay and Long Reef. Enhanced nutrient loads (threefold) for the 2030 were included to test system response and recovery (during major loading events). While the patterns for 2004 and 2030 are similar, the scenario for enhanced nutrient loads doubles growth peaks and extends recovery period (back to baseline conditions). The effect of hypersalinity constraining phytoplankton populations (as suggested by observational data in Figure 3.1) is tested in Figure 5 by comparing the Chl-a anomalies (2004 to 2030 runs) for the three continuously monitored locations in Port Phillip Bay. All sites indicate constrained production for the enhanced bay salinities in 2030, with the Hobsons Bay site clearly showing the greatest difference, especially when comparing responses to major loading events during the spring (days 230–320).

40 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

3DDLife - Chl-a: Hobsons Bay (top)

50 45 PHYTO 2030 x 3 40 PHYTO-2030 35 PHYTO- 2004 30 25 20

Chl-a ug/L Chl-a 15 10 5 0 0 50 100 150 200 250 300 350 400 Days 2004

3DDLife - Chl-a Long Reef (top)

9 PHYTO - 2003 x 3 8 7 PHYTO - 2030 6 PHYTO - 2004 5 4

Chl-a ug/L Chl-a 3 2 1 0 0 50 100 150 200 250 300 350 400 Days 2004

Figure 5.7: 3DDLife modelled Chl-a levels for 2004, 2030 and 2030 x 3 (nutrients) at Hobsons Bay and Long Reef.

41 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

PPB Chl-a Anomaly(2004-2030)

Central Long Reef Hobsons Bay

1

0.8

0.6

0.4

0.2

Chl-a anomaly (ug/L) Chl-a 0 0 50 100 150 200 250 300 350 400 Days

Figure 5.8: 3DDLife Chl-a anomaly (2004 to 2030) for central, Long Reef and Hobsons Bay sites.

Overall, the model is highly sensitive to a broad range of competing empirical parameters and so the quality of the calibration over a full year across all of Port Phillip Bay is good, especially given that the modelling was done on a coarse resolution grid. The study was very fortunate to have long time series of Chl-a data at three sites in Port Phillip Bay. However, further measurements of nutrients and Chl-a at more diverse sites (for example, , Bass Strait, Great Sands and Bellarine) would make it possible to greatly increase the accuracy of the empirical relationships. Other improvements would occur with high-resolution measurements around a single discharge (such as the WTP, Yarra River or Patterson River). The different discharges have very different ratios of N and P and so they would usefully help with the discrimination of the relative dependence of Chl-a growth. While this project has achieved its goals, there remains the need to continue with refinements of the model and to use new datasets. One of the most difficult aspects of modelling is calibration, because it consumes considerable time with data preparation and numerous model tests of the many competing coefficients. But the model testing provides great insight into the important ‘base of the food chain’ aspect of primary production, and calibrations and knowledge gains clearly justify the effort so far. Above all else, it was highly evident from the model that increases in nutrients lead to substantial changes in Chl-a levels and so it is necessary to know the thresholds above which the bay could be subject to substantial and numerous algal outbreaks, particularly during the favourable spring and summer conditions. Further refinements of the NPZ model will continue to advance this highly evolved tool. The modelling throughout the study, from the hydrodynamics through to the NPZ work, has challenged both the models and the modellers. The calibration results justify this effort and provide a solid basis for future managements of Victoria’s waterways. 5.5 POL3DD/WGEN resuspension modelling for Western Port The results from the sediment transport/settlement model for Western Port (shown in Figure 4.2) represent the distribution of sediment discharged from catchments, as a result of momentum absorbed by the bay’s receiving waters. To account for resuspension of material from in-bay processes such as strong tidal currents and wave mixing, the model was augmented with time series wave data to generate mixing using the WGEN module in the 3DD modelling suite. A threshold velocity for sediment bedload transport (0.07 m/s) and resuspension (0.1 m/s) was applied to account for resuspended transport and resettlement for the predominant mud/silt fractions in Western Port (mean grain size of 0.01565 mm). Figure 5.9 compares the results incorporating resuspension with those from Figure 4.2, indicating significant increases in TSS (approximately two orders of magnitude) with resuspension included, and enhanced focus of high turbidity in the eastern arm of Western Port.

42 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Figure 5.9: TSS with inclusion of wave mixing and sediment resuspension significantly increases concentrations and highlights observed maxima in the eastern section of Western Port.

43 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

6 RECOMMENDATIONS AND CONCLUSIONS

The scenario modelling has built on the development phase (Beaches Stage 7) to assess current conditions in Port Phillip Bay and Western Port marine systems, using high-resolution simulations for the two-year period 2004–05. These results have incorporated daily catchment discharges generated by the PortsE2 catchment model that enter the bays via 42 discrete points and diffuse discharge locations. Resultant dispersion patterns of pollutants indicate an overwhelming influence of wet-weather events to transport catchment loads to the bay. Comparisons to in-situ data are clear and the increased spatial resolution of the models enhances the understanding of load dispersion, settlement and decay within the bays. Climate change scenarios have been incorporated using CSIRO 2030 projections for rainfall and evaporation in the Melbourne region to modify the two-year record inputs. As there are significant differences in ocean flushing between Port Phillip Bay (one year) and Western Port (20–40 days), an initial focus examines the effect of shifting the rainfall/evaporation balance in Port Phillip Bay. To enable this assessment, the models were developed to incorporate heat loss and evaporation terms (as an evaporation scheme) and then calibrated against long-term salinity records. When models are compared using wetter conditions from the 1990s and projected dry conditions for 2030, there is a shift in salinity regime from hyposaline (less salty than ocean) to hypersaline (more salty than ocean) by two to three PSU, which significantly alters the circulation patterns and flushing of the bay. Due to prolonged drought conditions since 1998, the modeled 2004–05 conditions for Port Phillip Bay are hypersaline and thus projected changes to salinity and temperature, while evident, are more subtle. Differences in pollutant dispersal are mostly evident as a reduction in loading from the major discharge points (Yarra River and WTP). There was significant uncertainty in the model stability in Western Port when incorporating shifting evaporation/rainfall balance terms, as there was no appropriate time series record to calibrate outputs. As Western Port has large expanses of shallow mud flats, its response to an altered evaporation scheme would be significantly different from a scheme developed for Port Phillip Bay. To focus on effects of increased evaporation, climate change patterns in Western Port were constrained to an initial six-month period when major wet-weather events were absent. Hydrodynamics simulations indicate significant increases in temperature (Western arm) and salinity (NE arm). Companion dispersion modelling results indicate reduced loads caused an increase in coastal concentration of TSS near the discharge points and a subsequent downstream decrease in concentrations. There were negligible differences for toxicants and pathogens. To improve confidence in modelling projections for Western Port, it is recommended that high-quality salinity, temperature and flow calibration data is collected at key sites to capture spatial and temporal gradients. Beach scenarios were undertaken to assess the potential of impacts at priority beaches (as defined in the WQIP Interim Beach sampling project) from the PortsE2 discharge points. A nested 30 m grid model for the Swan Bay–Queenscliff region confirmed observational information that had suggested source waters from Swan Bay and ‘The Cut’ impacted the beach during dry weather conditions. An investigation of the various pathogen sources in this region (bird roosting in Swan Bay, marina discharges, ferry operations) is recommended to assist in determining a mitigation strategy. The Rye–Rosebud and Altona–St Kilda priority beach regions were also assessed for prevailing conditions and pathogen dispersal. In the northern region, coastally trapped waters dispersed discharges alongshore, accumulating pathogens at the two priority beach regions of Altona and St Kilda. In the south-east region, weak prevailing flows in the Rye–Rosebud area caused pathogens to accumulate alongshore in the vicinity of the discharge points. Continued scenario and calibration testing of the NPZ (3DDLife) model included further effort in relation to seabed type, shading by turbidity and dependence of phytoplankton on nutrients. Testing of additional empirical relationships from the scientific literature and toxicity/growth effects of elevated salinity (hypersalinity) on phytoplankton and zooplankton also occurred. It is likely that further work on the hydrodynamics will provide only small gains, as the HD models have become very advanced and well calibrated. One aspect of the HD model that remains troublesome is the rate of evaporation and salinity changes in very shallow water, such as the intertidal flats of Western Port. Different evaporation rates occur over these flats than in the deeper waters of Port Phillip Bay. It is recommended that the modelling tool, coupled with an appropriate catchment model, is incorporated into the Water Quality Improvement Plan to track and project changes associated with management actions. The models should be updated with more complex climate change projection data from CSIRO to improve resolution and certainty in water quality forecasts.

44 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

To improve confidence in modelling projections for Western Port, it is recommended that high-quality salinity, temperature and flow calibration data is collected at key sites to capture spatial and temporal gradients.

7 ACKNOWLEDGEMENTS

Additional information for model calibration and verification was provided by Andy Longmore (DPI). Photography is courtesy of Andy Stephens (EPA). Supportive information on mapped water quality was provided from the Two Bays studies undertaken in January and December 2007 (EPA). Advice on priority beaches was provided by Nigel Nicholls and Alison Kemp (EPA).

45 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

8 REFERENCES

Argent 2006, PortsE2: A decision support system for Whetton P and Power S 2007, Climate change in the Port Phillip Bay and Western Port Water Quality , Technical Report, CSIRO and Bureau of Improvement Plan, CEAH Report 01/06 to MW. Meteorology, ISBN: 9781921232930, 148 pp. Black KP 2001, The 3DD Suite of Numerical Process Models. ASR Ltd, New Zealand. www.asrltd.co.nz Bowie GL, Mills WB, Porcella DB, Campbell CL, Pagenkopf JR, Rupp GL, Johnson KM, Chan PWH, Gherini SA and Chamberlin CE 1985, Rates, constants, and kinetic formulations in surface water quality modelling, US Environmental Protection Authority, Athens, Ga. EPA 2008a, Marine risk assessment for the Better Bays and Waterways program. EPA 2008b, Beaches interim program report series. Eppley RW 1972, ‘Temperature and phytoplankton growth in the sea’, Fisheries Bulletin 70, 1063–85. Harris G, Bately G, Fox D, Hall D, Jernakoff P, Molloy R, Marray A, Newell B, Parslow J, Skyring G and Walker S 1996, Port Phillip Bay Environmental Study Final Report, CSIRO, Canberra. Harrison S, Lee RS and Black KP 2007, ‘An integrated catchment and receiving model for Port Phillip Bay and Western Port to address pressing environmental concerns’, proceedings of Coasts and Ports 2007. Jenkins GP, Black KP, Wheatley MJ and Hatton DN 1997, Temporal and spatial variability in recruitment of a temperate -associated fish is largely determined by physical processes in the pre- and post- settlement, Marine Ecology Progress Series, 148:23– 35. Jenkins GP, Black KP and Keough MJ 1999, The role of passive transport and the influence of vertical migration on the pre-settlement distribution of a temperate, demersal fish: numerical model predictions compared with field sampling, Marine Ecology Progress Series, 184:259–71. Longdill P 2008, Environmentally sustainable aquaculture: An eco-physical perspective, PhD thesis, University of Waikato, NZ. Melbourne Water 2007, Port Phillip Bay and Western Port ‘PortsE2’ model application and assessment project – existing land use case assessments – summary report, WBM Oceanics report to Melbourne Water. PoMC 2007, Supplementary Environmental Effects Statement, Channel Deepening Project, Volume 1. WBM 2007a, Scenario modelling report using PortsE2 catchment model, report to Melbourne Water. WBM 2007b, 2004–05 PortsE2 catchment model outputs for the receiving water quality model, report to EPA Victoria.

46 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

APPENDIX 1 PROJECT SERVICES (PROJECT BRIEF)

The development of a receiving water quality model for the Port Phillip Bay and Western Port Water Quality Improvement Plan (WQIP) to link with the DSS’s E2 catchment model. The study area for this WQIP task is the receiving waters of Port Phillip Bay and Western Port and associated estuaries, the Port Phillip Bay and Westernport catchment and ocean boundary conditions of Bass Strait. Project requirements • Adopt proven calibrated hydrodynamic model(s) linked to water quality and transport modules to simulate receiving water quality in Port Phillip Bay and Western Port. • Link the operational receiving water quality model (RWQM) with E2 catchment model inputs (loads). • Run baseline and event model scenarios over bay flushing and seasonal cycles in tandem with catchment scenario loading. • Run future growth scenarios that may include changes to in-bay conditions (e.g. marina development, dredging, aquaculture). • Run climate change scenarios for sea-level rise and increased storm frequencies (model may need capacity to incorporate wave modelling for erosion estimates). • Co-develop the model with an in-house user group and the E2 catchment modellers. • Provide a licensed model with user configurable components and modules for an ongoing monitoring tool to the in-house user group. • Provide ongoing modelling support to the in-house user group (when the WQIP undertakes scenario testing of pilot offset schemes in 2007). • Fit to current WQIP time frame (initial operational model in approximately two months, full calibrated model with initial scenarios in six months).

General model requirements A model for both Port Phillip Bay and Western Port (or separate models for each, if required) will need to have the ability to model: • the water quality constituents of TN, TP, TSS, salt (EC), zinc, lead and E.coli/enterococci, Chl-a and gross pollutants (litter) • processes, including constituent fate and behaviour • linkage with catchment model through process representation of constituent fate and behaviour in estuaries • point sources and their impacts with an adaptive model grid (i.e. a nesting capability) • with time starts using previous model operational files information (e.g. from the CSIRO model/s and Westernport Sednet model) • sediment transport, flux and resuspension. The E2 catchment model will be developed by April 2006. The receiving water quality model(s) need to be developed soon after, to use outputs from the catchment model to then be used as part of the scenario-testing process, which is the next stage of the DSS project (taking place between April and August 2006). Specific requirements of the modelling software • The software must be installed on a desktop computer designated by EPA for the modelling tasks. The fully calibrated and verified modelling package will reside permanently on EPA’s computer after completion of all work by the successful tenderer. • Backup copies are required of the modelling programs and all associated parameter files necessary to enable the model to be reinstalled with full functionality. • The hydrodynamic model must be able to simulate the vertical and horizontal salinity gradients, thermal stratification and current shear that are likely to exist in the study area. • The hydrodynamic model must allow for the input of meteorological data so that evaporation, rainfall, air pressure, wind speed and direction, and solar radiation are included in the calculations. • The hydrodynamic model must be able to simulate the wetting and drying of the tidal flats in the study area. • The hydrodynamic model must be able to accept input data (loads, flows and their coordinates) as ASCII text from the catchment model or link to the E2 catchment model outputs.

47 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

• The water quality modelling module (WQ module) must be capable of simulating advection and dispersion of water quality parameters and predicting their concentrations at various levels in the water column. • The nutrient component modelling must be capable of accounting for nitrogen and phosphorus fluxes from sediments. The sediment characteristics vary throughout the bays and it is likely that the potential nutrient fluxes will also vary spatially. The model must be capable of accommodating differing rates of nutrient release and deposition across the entire sea floor of the study area. This includes the sediments of the tidal flats and . • The nutrient component modelling software must be capable of handling nutrient inputs from the atmosphere, having regard to the possibility that the atmospheric loads will vary across the estuary. • The phytoplankton component modelling should be capable of simulating chlorophyll concentrations and broad phytoplankton composition (for example, diatoms and dinoflagellates) at different depths in the water column. • The phytoplankton component modelling software must include light availability (photosynthetically active radiation) as one of the parameters used to determine chlorophyll production or loss. • The toxicant and pathogen component modelling should adopt appropriate rates of decay, as outlined in supporting literature. • The gross pollutant component modelling will include positively, neutrally and negatively buoyant fractions for a range of sizes and materials. • The model must be able to accommodate point-source nutrient inputs (and flows), some of which will be constant — for example, wastewater treatment plant effluent discharges — and some of which will be highly variable and related to rainfall. • The sediment transport module must be capable of simulating sediment transport in situations involving bed friction that varies spatially and temporally. • The package should include preprocessing software that facilitates the creation and modification of the modelling grid or mesh, as well as all other data preparation necessary for the model setup. • A specific wave modelling module or built-in capability to include effects of coastal erosion (especially for future scenarios). The package should include postprocessing software that allows: • incorporation of a map • display of isobaths (bathymetry) • display of water velocity vectors • display of isopleths for any of the modelled parameters • display of percentage impact according to a defined threshold concentration of a modelled parameter (such as from ANZECC) • contours to be displayed in colour or as line plots • representation of the behaviour of various parameters in the estuary using colour graphics and animation. Installation of the software and setup of the hydrodynamic model The contractor will: • assist in all aspects of the modelling software installation, including the compilation of source code (if applicable) • establish the appropriate computational grid for the model • integrate the bathymetric and sediment data into the model • set the appropriate values for any of the parameters used by the hydraulic modelling software • ensure the proper interfacing of the nutrient, wave and sediment transport modules with the hydrodynamic modelling software • calibrate the hydrodynamic model based on measurements derived from historical data • calibrate the nutrient dynamic modelling based on historical data • calibrate the sediment transport modelling based on historical data • calibrate the phytoplankton growth/chlorophyll modelling • verify the outputs of the hydrodynamic model • verify the outputs of the nutrient dynamic modelling • verify the outputs of the sediment transport modelling • verify the phytoplankton growth/chlorophyll modelling.

48 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

Training and model handover Explanatory note: EPA will want to continue to use the hydrodynamic model set up for Port Phillip Bay and Western Port on an ongoing basis for scenario testing. It may also need to use the modelling software to simulate the behaviour of other constituents in the future (for example, heavy metals). The contractor will: • provide training in all aspects of the suite of modelling software necessary to ensure that the model, once set up and calibrated, can be operated ‘in house’. The training is to include all aspects of the calibration process. The training will be provided at the EPA offices in Melbourne during various stages of the development of the model. • provide a user manual(s) for the software suite. • document the configuration (including parameter settings) and all other factors used to customise the modelling suite for Port Phillip Bay and Western Port.

What EPA will supply EPA will provide the following: • A dedicated PC to be used for the installation and setup of the modelling software. (e.g. Dell 670 workstation Dual 2.8GHz Xeon processors with up to 3.80GHz with 2MB L2 cache and an appropriate graphics card). • Digitised bathymetric data of the area to be modelled. • Tidal data. • Meteorological data. • Data acquired from field surveys conducted specifically for model setup and calibration. • A full-time officer to assist with all stages of the modelling process. • A workstation for the contractor, as well as internet access and an email account. Key deliverables a) A licensed, calibrated RWQM compatible with the WQIP catchment model. b) Software modules that will be sufficient to allow EPA to achieve its modelling objectives of Port Phillip Bay and Western Port. c) Sufficient training and skills transfer to enable EPA to use and modify the modelling software for scenario testing in Port Phillip Bay and Western Port using internal resources. d) Models that simulate the hydrodynamic behaviour, nutrient dynamics, wave climate and sediment transport of Port Phillip Bay and Western Port. e) Documentation of the parameter values and any other customisation used for the model setup. f) User manual(s) for each component of the modelling software. g) A presentation outlining results of the model linking phase. h) A presentation/discussion presenting scenario results. i) Draft report of the project for comment/discussion. j) Five copies of the final written report. k) A PDF of the final report.

49 PORT PHILLIP BAY AND WESTERN PORT RECEIVING WATER QUALITY MODELLING: SCENARIOS

APPENDIX 2: GREENHOUSE 2007 POSTER PRESENTATION

50