Source 2016

Robert Carr

Source 2016 Workshop Opening

Richard Beecham, Dr. Christobel Ferguson

NSW Implementation of Source

The NSW Government has used river system simulation models for water policy and planning assessment since the 1980s, starting with bespoke monthly models, and evolving to the daily models implemented in the IQQM platform from 1993. DPI Water is now actively transitioning two regulated valleys to the Source platform; the Border Rivers in collaboration with Queensland, and the Murrumbidgee, and has undertaken foundation work in five others. The transition provides an opportunity to improve on our current IQQMs, a consideration that will become important in gaining crucial stakeholder acceptance. The approach taken to implement the models will be discussed, as well as progress and indicative results. Model progress will be reviewed early 2017 to decide on future directions.

Seker Mariyapillai, Uttam Manandhar (Unlicensed), yongli, Andrea Ballinger

Implementation of Source in Victoria

A monthly time step SOURCE model of Victorian trial catchment, namely the Goulburn, Broken, Campaspe and Loddon (GBCL) valleys, has been developed in the past using Network Linear Programming (NetLP) ordering algorithm. With the availability of new rules-based ordering algorithm in SOURCE, Victoria has switched to developing daily time step rules based GBCL SOURCE model and explored modelling of operational and management rules in individual valley models. For this, development of daily input derivation methods and derivation of around 114 years of daily inputs have been completed. Our key current activities for the development of daily GBCL model includes testing of Source functionality and implementation of

(i) annual accounting resource assessment including carryover in the model,

(ii) developing Irrigator demand models for Victorian irrigation areas,

(iii) trialling wetland functionality using a wetland in Goulburn system and

(iv) modelling environmental demand along the Goulburn river.

Victoria has worked collaboratively with eWater and MDBA to resolve more than 60 issues, and made significant progress on first two activities.

Victoria is continually working closely with eWater and MDBA to resolve the remaining issues, which are mainly related to consistency in SOURCE modelling in Victorian annual accounting resource assessment and carryover accounting, bi-directional flows and ordering priorities. Together with these, Victoria will undertake works on wetland and environmental demand modelling activities. The daily GBCL model will also be progressively enhanced to incorporate those features including operational and management rules. Stage 1 of the daily model is scheduled to be completed within a year and will have capabilities that are viable with SOURCE functionality.

alistairkorn-admin

The Source Murray Model

The Source model of the Murray and Lower Darling System (SMM) reflects a major investment over the last 15 years by the MDBA and jurisdictions in developing a daily model for the Murray System. The SMM is envisaged to replace the MDBA’s current modelling framework MSM-Bigmod and as such needs to be able to be used to meet multiple purposes including legislative responsibilities (BSMS, BDL, SDL scenarios), policy analysis and river operations. The development of the SMM has been difficult due to the inherent complexity of the Murray system, the requirement to convert monthly to daily rules and customisation of the Source modelling framework.

This presentation provides a high level overview of the current status of SMM, including its conceptualisation, configuration, calibration, testing, and review process. It will also consider some of the future improvements and challenges that exist in further developing the SMM and Source modelling framework. The SMM is currently configured to represent a Without Development scenario and the Baseline Diversion Limit for the Murray and Lower Darling System. The SMM is also being developed as an operational tool and is undertaking salinity forecasts.

David Waters and Craig Johansen

eWater Source application in Queensland to date and into the future

Craig Johansen Queensland Government

The Queensland Government has been a long term supporter of the development of the eWater Source platform.  Queensland Government is currently transitioning from existing model platforms to eWater Source platform to support technical assessments across Queensland.

There has been a long development period in building the capability in the eWater Source platform and staff gaining proficiency in the platform to address a broad range of issues dealt with in Queensland.

Queensland was an early adopter of the Water Quality aspect of the Water Source platform with Great Barrier Reef reporting.  Queensland will utilise Water Quantity aspect of the eWater Source platform in new models to address in
                - developing Queensland Water Resource Plans
                - developing strategies to protect urban water supplies during drought
                - addressing legislative requirements in the Murray-Darling Basin Plan
                - assessing Climate Change Impacts
                - assessing extreme events through Stochastic Modelling.

The adoption has had several challenges that had to be dealt with the transition to the new model platform.  However, the adoption of the platform has provided new functionality that was not available in the existing platforms.

eWater Source application in the Great Barrier Reef, Queensland

Dave Waters, Rob Ellis Queensland Government

The eWater Source model has been used to report annually on the progress being made towards the Great Barrier Reef (GBR) water quality targets as a result of state and federal government investment in improved land management practices. The flexible nature of the eWater Source framework has enabled the Queensland Government to develop and refine a range of tailored plugins, over the past decade, to represent the complex range of climate, landuse, land management, and constituent interactions across the 423,000 km2 draining to the GBR combined make up the GBR Source model.

In addition to the annual report card load reductions, modelled outputs are now being used for a range purposes including linking to receiving water models for scenario assessment, assisting regional Natural Resource Management groups with prioritisation of funds, assessing future land management scenarios to improve water quality and prioritising new monitoring sites. Improvements to the modelling approach include reduced run times, updating the hydrology calibration approach, refinement of constituent budgets to align with load monitoring program data, the use of python scripting to speed up model parameterisation and constituent budgets and updating input data sets such as remotely sensed seasonal cover.

A number of important learnings have come out of the past decade namely: the need to develop your own in house skills in plugin development is the most cost effective approach. Plugins require maintenance and updating, as the internals of Source change, hence development requires a longer term commitment. It is important to apply a consistent methodology across basins, maintaining base data sets and model algorithms for a fixed period before updating and the importance of regular external peer review of the modelling approach to guide prioritisation of work.

A lack of base data (eg gully extent), and the science that underpins the models (eg representation of stream bank erosion processes), remain the major limitation to improved model predictions. Coupling improved algorithms and processing better input data sets would not be limited by the Source software.

A future challenge for eWater is how to minimise the impact that internal changes to the software can have on external plugins. Experience in this project has shown that a substantial amount of the plugin maintenance effort is devoted to keeping up with seemingly unrelated changes within Source, even when the model structure of the plugin is unchanged. Even small internal changes can have major implications on external plugins developed built with prior assumptions.

Claire Sims, Matt Gibbs (Unlicensed), Mahdi Montazeri (Unlicensed)

Source Modelling in South Australia

South Australia is currently using eWater Source to undertake catchment and flow routing modelling in the Bungala catchment on the Fleurieu Peninsula, part of the Western Mount Lofty Ranges Prescribed Water Resources Area (WMLR PWRA). The purpose of the modelling is to assess whether environmental flow metrics can be achieved with fewer low flow bypass devices on farm dams within the catchment. Two scenarios in a Source project were developed: 1) a catchment rainfall-runoff model, and 2) a farm dam flow routing link model using the Farm Dam plugin. The catchment rainfall-runoff model is used to generate inflows to the farm dam flow routing model, then farm dam management scenarios can be simulated using the farm dam scenario.

Using a modified version of the MDBA eWater Source Lower River Murray model, DEWNR routinely undertake lake level and barrage release modelling for a given forecast of flow to South Australia to support and inform River Murray Operations management of the Lower Lakes, Coorong and Murray Mouth site. The model assesses the expected lake level profile and barrage releases based on a series of (45 year) historic climatic and fixed diversion patterns, including SA Water metro and country towns, and South Australian irrigation diversions. 

Joel Hall, Michael Braccia

Western Australia and eWater Source: overview, challenges and plans for the future

The Department of Water Western Australia has successfully developed two of the State’s main reservoir models into eWater Source. The Wellington reservoir model in the states south-west, originally developed in REALM, has been used to set allocation limits and define the reservoir releases. Using the flexibility of the user defined functions in Source, a water quality component has been added to represent the stratification and mixing of salt in the reservoir. 

The Ord reservoir model in the Kimberly region, originally developed in MIKE BASIN has been rebuilt in Source and used in supply planning to assess water availability. The model will be used in the next round of planning to assess how to best manage the system to meet environmental and development outcomes. The Source model platform has made the model more accessible and understandable without losing any functionality.

The Regional Estuaries Initiative is a four-year, $20 million Royalties for Regions funded program to improve the health of six Western Australian estuaries. The majority of West Australians live on or around estuaries - these water bodies are integral to our way of life. Unfortunately, estuary condition is compromised by drainage from agricultural, industrial and urban catchments. Catchment models, which will deliver flow, sediment and nutrient will be developed in eWater Source, and will be boundary conditions to state-of-the-art hydrodynamic and biogeochemical estuary models. The catchment models will be used to determine the impact of a suite of drainage and agricultural management scenarios on the flows and nutrient loads to the estuary.

Carl Daamen (Unlicensed), Karina Redpath

Integrated water resource management in the Upper Godavari Basin, Maharashtra, India

Equitable distribution of water in river basins is a target for improvements to water management in the State of Maharashtra, India. However, since equitable distribution has many dimensions, its implementation in river basins has been difficult.  There is a demand from stakeholders that both distress and water be equitably shared in the basin and this is leading to frequent confrontations amongst upstream and downstream beneficiaries.

The Government of Maharashtra noted that the situation in the Godavari basin is similar to issues arising in the Murray-Darling Basin in Australia and hence a Schedule for ‘Actions to Improve River Basin Management’ was agreed in August 2014 under an MOU between the Government of Maharashtra, India and the Government of New South Wales, Australia.  Under this Schedule a joint project team was formed to build river basin models for the Upper Godavari and to exchange experience and knowledge in water management.

For the purposes of the project the Upper Godavari basin is divided into smaller sub-basins. At present, a river basin model for the Mula-Pravara sub-basin has been built using eWater Source to demonstrate use of Australian tools relevant to investigation of water management issues in Maharashtra.   Construction of a larger model of the whole Upper Godavari is progressing well and it will be used to compare the effects of new water management options on water use distribution within the basin.  This will provide important information to support development of new water management policies for this basin.  The aims of the project will be summarised and some preliminary model results presented.  

Juanita Moolman, rachelb (Deactivated)

Integrated modelling in the Mekong River Basin with eWater Source

Integrated Water Resources Management (IWRM) in the Mekong River Basin is challenging due to factors ranging from the physical characteristics of the basin through to social and economic trans-boundary issues. An advantage of computer-based modelling for IWRM is that models can form the basis for developing a shared understanding of a river system that is transparent and testable. Once a “baseline” understanding is established and agreed-upon, models can then be used to test the likely future impacts of changes in water resources management policy and provide a defensible and scientific basis for decision-making.

We present two projects underway in the Mekong Basin that demonstrate alternative approaches to integrated modelling using eWater Source. One uses Source’s built-in algorithms to create a truly integrated model, while the other uses Source as an integrative framework to combine results of other models.

Cameron Dougall Cameron.Dougall@dnrm.qld.gov.au

Great Barrier Reef Source Catchment's modelling: Optimising workflow, with Jupyter, Python and the Veneer Plugin
C. Dougall a, J. Rahman, D. Waters a. S. Darra.
a Department of Natural Resources and Mines, Queensland
b Flow Matters Pty Ltd, PO Box 272, Jamison, ACT, 2614

The Great Barrier Reef (GBR) plan 2013 outlines water quality targets to address declines in water quality discharging into the GBR Lagoon, with Source Catchments modelling used as one line of evidence in assessing progress toward those targets. Two key challenges for the modelling are reporting annually the impact of ameliorative catchment actions and continually improving model performance. The annual workflow is underpinned by four main areas
1. Model parametrisation, 
2. Run model, 
3. Reporting and assessing results 
4. Selecting a refined parameter set, (which in turn leads the user back to step 1)
Logically this is an iterative process and it follows that the faster one can iterate (with appropriate feedback) the greater the potential for improving workflow. The eWater Source Catchments framework comes with an impressive array of in built tools and options that assist workflow, however, it is impossible to tailor solutions to fit all users’ demands. In response, we hypothesised that GBR Source catchments model workflow could be better optimised with the integration of Jupyter, Python and the Venner Plugin.
Jupyter Notebook is a web application that allows you to create and share documents that integrate live code, and graphical result windows. The project chose the python language to use within Jupyter, due to its wide use in the scientific community, impressive libraries, and its emphasis on code readability, while the interaction with the Source model, is managed by the Veneer plugin. The work team, had limited coding background, and the approach was found to be an ideal learning environment. The majority of workflow gains were in reporting and assessing results such as assessment of sediment and nutrient budgets. With the development of parameter quality assurance scripts between scenarios’ being particularly valuable. Importantly the time taken for assessment of model results and reparameterisation was reduced in some instances by half. 
In conclusion the use of Jupyter, Python and the Venner Plugin shows considerable promise at further improving GBR source catchments model workflow, however, a significant time investment is initially required for those without a programming background. Nonetheless, the long-term benefits are likely to outweigh any initial loss of productivity.
Keywords: Source Catchments, Workflow, Veneer

GS, gk (Unlicensed)

Lachlan Valley eWater Source Model Development

The Lachlan River Valley, located in central western NSW, occupies around 85,000 km2 or about 10% of the Murray-Darling Basin. The Lachlan River is a complex regulated river system having numerous anabranches, several headwater and re-regulating storages, wetlands, major irrigation developments, Town water supply and various environmental needs. Regulation of the Lachlan River has allowed irrigation to thrive in the catchment. Water is stored upstream in Wyangala and Carcoar dams and released downstream based on irrigation and environmental flow requirements. Lake Brewster and Lake Carregiligo also regulate the flows as re-regulation storages within the basin.

In the grip of the Millennium Drought, various studies were completed that identified the need for additional storage as the current storages were found inadequate to meet Water Security in the region. Subsequently, the NSW Government commissioned Water NSW in 2014-15 to undertake an investigation into potential storages in the Lachlan Valley, including the Belubula River. The modelling objective was to develop fit for purpose models to be used for the new water infrastructure investigations. Firstly, the tasks undertaken involved hydrologic modelling to provide a detailed coverage of the long term tributary inflows that will be used for building the Sacramento models, within the Source were used to calibrate/validate the rainfall runoff models based on gridded rainfall, evaporation data provided. A number of challenges in the hydrologic investigation are presented in this paper including the modelling of inflow contribution from the large inland catchments.

The next stage involved building water demand and water allocation planning models for the Lachlan valley. The paper discusses some of the key challenges, issues and success in building the eWater SOURCE water demand and water allocation models based on the requirements set out in the Water Sharing Plan for the Lachlan River valley. The paper also presents results on modelling water demand and river diversions including environmental flows in a complex regulated river system that includes some significant wetlands (including Booligal and floodplain red gum swamps).

Udaya Kularathna

Water resources modelling at Melbourne Water: Future transition to Source

Melbourne Water (MW) is developing a Source model as a future replacement of its current suite of water supply headworks system simulation and optimisation modelling tools based on REALM and OptimizerWSS. The current headworks model is used to assess the implications on the water supply headworks system of future climate, streamflow and demand scenarios and identify the best management strategies. The model informs water supply strategy development, water resource planning, annual operation planning, drought response planning, water allocation assessment and environmental flow planning. 

There have been multiple goals and challenges in Source model development that include (1) developing the capability to use one model for all of the above business processes, (2) calibrating a complex model using the functionalities available in a new modelling platform that has not previously been used to model a complex urban water supply system, (3) improving the processing of model runs and the analysis of outputs in modelling a wide range of future climatic scenarios, (4) developing multi-objective optimisation modelling capability to optimise water resource system planning and operations under uncertain future scenarios, and (5) enhancing the headworks model further to support integrated water management. 

A project undertaken for Melbourne Water by eWater in 2016 further improved the Melbourne Source headworks model by focusing on the first four goals and challenges. A parallel project, also undertaken by eWater, prepared a high-level design for addressing the fifth goal. The presentation aims to share the experiences of MW so far in developing the Melbourne Source model.

Stephanie Ashbolt

Optimising short-term operating rules for a water grid

This presentation discusses the use of Source in recent PhD research to simulate and optimise short-term operating rules for a water grid case study. This case study water grid is based on the South East Queensland Water Grid and includes 28 dams and weirs, 2 groundwater borefields, a potable wastewater recycling scheme, and desalination plant. These water sources are connected to 48 demands via a network of 7 two-way pipelines, as well as one-way pipelines and streams. Source's Insight module is used to optimise 16 operating rules that govern operation of the desalination plant, wastewater recycling scheme and two-way pipelines over a five-year planning horizon. These rules are optimised to meet three objectives: maximising water security, minimising operational cost, and minimising spills from reservoirs. This research also investigates the potential of optimising operating rules across multiple scenarios of forecast streamflow, to identify operating rules that are both robust and tailored to expected conditions.

Joel Rahman

Andrew Herron

Secure Yield Study

Jacobs was contracted by Port Macquarie Hastings Council to develop a Source model representing the current system, compare future headworks options and ultimately support long-term water supply planning in Port Macquarie, NSW. In this rapidly developing region, the Council responsible for undertaking a Secure Yield Study (this is the first application of Source for a secure yield project in NSW) to ensure sufficient water supply headworks capacity under current and future climate conditions.  The project involved the development of calibrated Source Rivers models to calculate the secure yield under existing and future infrastructure and climate scenarios. This was combined with economic analysis to inform decision on water restriction and infrastructure development.  

This presentation will discuss the complexity of the existing supply system and future augmentation options in terms of translating physical operations into a modelling representation. Specifically around the use of multiple storages, backflow arrangements and complex operating rules that are used overcome water quality challenges inherent in this unfiltered system.   

Andrew undertook this work while at Jacobs and at Eco Logical Australia (on behalf of Jacobs).

Tony Weber

Pushing the boundaries in Source -  subdaily timesteps and improving stream routing

Tony Weber – Alluvium Consulting

Badin Gibbes – University of Queensland

In several recent projects there has been a requirement to use subdaily timesteps in Source to attempt to capture more rapid responses in urban streams and to improve stream routing responses.  Source is flexible enough to now allow for native sub-daily timesteps to be run within the framework and this presentation will outline the potential to do this, issues that may arise in doing so and the potential uses for running finer temporal scale.  This has been exemplified in the application of Source to the mid Brisbane River in order to investigate the development of a rapid response model to inform emergency management of water quality issues (chemical and sewer spills, salinity pulses and high turbidity events).  In undertaking this research, we identified that there are only a limited number of high temporal resolution, long-term rainfall records available, and also limited flow gauging data at sub-daily timesteps.  There are also only a limited number of hydrologic models within Source that are able to handle native sub-daily data.  The calibration of sub-daily models was challenging given the longer processing time required and that the calibration was dominated by larger daily events rather than sub-daily short term responses which were of most concern to the project.  From this, we have gained a greater understanding of the potential of the framework to be used in this manner and this will be discussed in the presentation.

Application of daily SedNet for modelling catchment-scale sediment generation and transport: New Zealand case study

In catchments that have experienced long-term land clearance and agricultural intensification, sedimentation can have a significant impact on the health of receiving water environments. Since the early 2000s, the “Sediment River Network Model” (SedNet) has been used as regional scale tool for estimating long-term annual sediment budgets that identify patterns in erosion processes and transport of sediment in surface waters. SedNet has been applied in many catchments in Australia and New Zealand to assess management actions and prioritise soil conservation investments to improve water quality. Although SedNet provides comprehensive spatial resolution, there are instances where a finer temporal resolution than an annual sediment budget is required to simulate event-based sediment export (e.g. impact of high flows) at different times within a year and to investigate within-year variations. Daily SedNet (dSedNet), a time-stepping spatially-distributed sediment budget model for predicting daily sediment loads, was developed to fit such catchment modelling requirements. The dSedNet sheet and gully erosion rates are simulated based on disaggregation of mean annual rates by daily rainfall and runoff. Sediment generation models are assigned to hydrological units representing landuse types, split into hillslope, streambank and landslide erosion sources. Fine and coarse sediments are transported through the river node-link network, where in-stream deposition occurs. This paper discusses the development and calibration of a dSedNet model coupled to an integrated catchment hydrological model for simulation of erosion processes in the Porirua River Catchment, New Zealand.

phillip.jordan@harconsulting.com.au

Towards the development of an in-stream processing model plugin for nitrogen species

Abstract: Nitrogen species in catchment models constructed using eWater Source have conventionally assumed that the Nitrogen generated from the subcatchments is conserved as it flows with the water through the streams in a catchment. Less commonly, simple exponential decay models have been applied to the Total Nitrogen concentration and/or to one or more of the Nitrogen species. Where decay models have been applied to multiple species, the Source user interface has limited the modeller to applying these independently to each species, which ignores the conservation of mass involved in transformation between species and may also lead to the simulated Total Nitrogen not equalling the sum of its simulated component species.

A new plugin model was developed in Visual Studio C#.Net to implement in-stream processing of Nitrogen species in Source. The model implements an aggregated model for several constituents: Total Nitrogen, Nitrate and Nitrite Nitrogen, Ammonia, Organic Nitrogen and Temperature. It therefore tracks the transformation in forms of Nitrogen in-stream, the temperature dependent de-nitrification process and it maintains Total Nitrogen concentration as the total of its constituent forms.

The model was tested using flow and nutrient monitoring data for the Latrobe River, Gippsland, Victoria, where in-stream processes were exhibited in the monitoring data. The performance of the model in matching in-stream concentrations to monitoring data will be demonstrated.

Alice Brown (Unlicensed), Matthew Bethune, alistairkorn-admin and Simon Gallant

Practice Notes for Source Model implementation

To support the adoption of a ‘consistent Basin water modelling platform’ within the Murray-Darling basin (MDB), CSIRO and MDBA are collaborating on a project to support documentation of current approaches, and undertake research into the advantages and disadvantages of alternative approaches for building models to represent water sharing plans in the MDB. These methods are being documented as a series of “practice notes”, short technical documents defining recommended practice for a specific aspect of building a model. As each jurisdiction may have a different recommended practice, there may be a number of methods documented representing the same process in the practice notes.

Using a set of test reaches from across the MDB, research is being undertaken to assess if and when one method offers an advantage. When complete the documented methods will be shared between jurisdictions and other Source users. Establishing an environment where methods are documented and exposed for peer review will improve the consistency of the model development and model transparency. This presentation covers the progress so far in this two year project, and details the test reaches, example practice notes and progress on the research component.

gpodger (Unlicensed)

Catchment Climate Data Extractor tool

Preparation of precipitation, potential evapotranspiration and minimum and maximum temperature time series inputs for Source from gridded datasets can be very time consuming. There is existing functionality in the Flow Calibration Wizard to do this but this is particularly slow for large gridded datasets, where there is a large number of functional units and an extended period of time. As part of rainfall runoff and snow and glacial melt modelling in South Asia CSIRO has had to prepare inputs for hundreds of catchments from a range of gridded data sets including Princeton, Aphrodite, WATCH, ANUSPLIN and ESDIIM, which are in different formats, and spatial resolution and coverages. To facilitate this process CSIRO has built a standalone climate data tool which takes a sub-catchment raster file of FUs and gridded climate datasets and produces a Source compatible CSV time series input file. The program also includes climate change scaling factors for RCPC4.5 and 8.5 scenarios from 42 GCMs so that climate change scenarios can also be created. The tool is also capable of working out average elevations of sub-catchment FUs from a DEM to support elevation corrections within rainfall runoff models. The tool currently works for South Asia for a range of gridded dataset formats but could be readily adopted for Australian gridded datasets and climate change scaling factors.

Geoff Davis

Source Software Developments

Agnes Sakal

Development of eWater Source - Sacramento model for the Namoi River Valley

The newly established WaterNSW embarked on rural catchment modelling using the eWater Source modelling Framework for the development of asset solutions to meet customer needs. The first pilot model development was undertaken on the Namoi River Valley. 
The Namoi River is one of the Murry-Darling Basin’s major NSW sub-catchments. It covers a total area of about 42,000 km2 from the Great Dividing Range near Tamworth to the Barwon River near Walgett. The Peel River is a major regulated tributary to the Namoi with an area of around 4,700 km2. 
For rainfall run-off modelling the Sacramento model was selected form the suite of rainfall run-off models available through eWater Source. The Sacramento model requires catchment rainfall and potential evapotranspiration to simulated flows. These data was derived from the BOM areal rainfall and potential evaporation estimates.
Then, the Namoi River Valley was delineated into 56 sub-catchments based on the selected gauging stations. The following Figure illustrates the sub-catchments in the Source model and the calibrated flow duration curve at one of the calibration sites.

WaterNSW is planning to use this rainfall run-off model as a base for the Namoi Planning model in the future. Some of lesson learnt during the development of the models will be presented.

Daniel Bothelo

Using Source to estimate microbial loads from a water-supply catchment

Water Management in the Kingdom of Morocco

Uniadi Mangidi

Developing an Urban Vegetation Water Demand framework to mitigate urban heat island (UHI) effects

Jannah Hardefeldt

Is SOURCE suitable? Identifying critical source areas of diffuse pollution to drive management actions

The Richmond River Catchment covers 6,850km² and drains the largest coastal floodplain in the highest rainfall area in NSW. The dominant agricultural land uses include sugar cane, dairying, broad acre agriculture and grazing and these are considered the main factors responsible for sediment and nutrient delivery to water bodies. Understanding the relative contribution of these land uses to the total catchment pollutant loading will enable targeted efforts of on-ground works at the farm-scale and industry collaboration to reduce diffuse source pollution. Critical source areas (CSA’s) are areas within a catchment that contribute disproportionate amounts of pollutants such as nitrogen & phosphorus to the total catchment load. These areas can be identified using GIS and hydrological modelling. It is hypothesized that identification of CSA’s in the Richmond River Catchment can be achieved with the SOURCE model and will optimize management of catchment land use and water quality improvements.