The Office of Water uses a range of modelling techniques to understand how our river and groundwater systems behave. The resultant models help predict what will happen in a variety of scenarios, including water sharing, compliance and the effects of climate change, and factors that affect water availability.

The complexity of modelling approaches can vary from very simple conceptual models through to very detailed and data rich approaches. Models can be applied at scales varying from very small scale specific sites or study levels, through to regional scales. The utility of any of these models is constrained by basic limitations in our knowledge and the availability of data to build the model. Models cannot generate knowledge; they only combine what we know into useful forms.

Surface water model for water sharing and management

The main surface water model used for water sharing and management is the Integrated Quantity and Quality Model (IQQM). IQQM has been developed to assess the impacts of different management strategies on all water users. The models have been developed to simulate the major hydrological processes in river valleys along with relevant management rules. These models have been calibrated to match reservoir levels, diversions and flows over the calibration periods. The models are set up in such a way as to reproduce the average long term behaviour of the river system for planning purposes and not specifically to reproduce individual daily flow behaviour in any particular year, or to forecast any future year.

IQQM models have been developed for most inland river basins and some coastal river valleys. The models can be used to obtain a range of information on simulated river system behaviour ranging from average summary statistics to specific event or sequence details.

These models are used in different water management areas such as:

Groundwater models

A range of regional groundwater models have also been established to investigate, audit and derive sustainable yield estimates for the state's groundwater systems. The models have been used in water sharing plans, land and water management plans together with monitoring of groundwater use throughout the state.

Data-centric groundwater modelling project for coal seam gas and large coal mining developments

This project will combine groundwater data from a range of sources using new methods to develop more sophisticated groundwater models in terms of predictions, including the potential risks associated with assessing and monitoring coal seam gas exploration and large coal mining developments.

The overall aim of project is to bring all useful groundwater data together into a single integrated model, to enable 'best science-based' groundwater decision making.

Now in the research phase, it is hoped this project will eventually provide an accurate risk-based assessment of current and future groundwater activities in NSW.

This project is a partnership between the Department of Primary Industries - NSW Office for Water and National ICT Australia (NICTA) - Australia's Information and Communications Technology Research Centre of Excellence.

Project overview - Enhanced groundwater modelling – Coal seam gas and large coal mines (PDF 82 KB)

Progress to date

Initial research around water trading assessments has helped to construct a base case and provide insight into the important aspects in making groundwater impact assessments and how difficult it is to construct parameter fields that capture realistic uncertainties related to the sub-surface environment, especially where the is limited data as is generally the case with predicting coal seam gas and large coal mining groundwater impacts.

Research into groundwater flow models and specifically Modflow simulation software has shown that it is flexible and modular which makes it suitable to be applied across a range of different scenarios. At its core is numerical component which solves partial differential equation which is typically used for saturated groundwater modelling.

However, its predictions are heavily based on assumptions such as structural geology and material properties. In essence it uses a single set of inputs to make a single prediction without taking the uncertainty of those inputs into account. Best practice today is a combination of expert interpretation of hydrogeology and observed characteristics from bore holes.

NICTA has received and analysed the Namoi Water Study conducted by independent experts in 2012. This study went a long way to producing a regional stratigraphic model for the region. This built on a range of industry and government data and produced a groundwater model built on the numerical Modflow software platform.

The values that defined the hydrogeological properties were taken from available impact assessment reports from industry proponents or from the NSW Office of Water models where they existed. However, these were taken on face value and extrapolated to a regional scale without fully considering the uncertainties that this could cause. The study did identify through a sensitivity analysis on parameters the effect of various parameters on predictions.

NICTA is now building on this sensitivity analysis to develop tools that will enable better probability distributions over parameter values such that a more robust prediction of possible impacts can be made. Modflow does not question the parameters, it does its job as a numerical tool to solve the problem based on inputs. NICTA is now focusing on how the inputs should be chosen and how uncertainty can be quantified capturing available data and existing expert knowledge.

Additional industry data has now been acquired and analysed providing more detailed stratigraphic units and a better understanding of the internal lithological heterogeneity preparing us for the next phase of the project.

Next steps

NICTA and collaborators at the University of New South Wales are now experimenting on the effect unobservable small scale heterogeneity and its effect on uncertainty in large scale extrapolations. In other words, the range of values that should be input into Modflow to give a more realistic understanding of potential impacts to an aquifer.

While uncertainty is considered to some extent in industry practice today, it is fundamental to improve the understanding of the range of possible realistic values that need to be used and how these can be constructed from typical data sets and expert interpretation.

Ultimately it is the goal of the project to deliver tools that will allow users to better represent uncertainty in groundwater modelling, recognising the sparse and patchy nature of hydrogeological and sub-surface data.

Impact predictions of coal seam gas extraction requires the prediction of any reduction in water availability in shallow aquifers caused by the extraction of water and gas from deep coal seams. This type of prediction scenario is generally data poor as there is often very limited number of monitoring bores and hydraulic test results for the deep geological formations between the target coal seams and the shallower aquifers. For this problem the project is now developing statistical modelling methods to better capture the possible variation in the large scale hydrogeological characteristics, such as hydraulic conductivity of these formations that typical govern the magnitude of these impacts. These methods will enable the geologic and hydraulic test data, to be combined with expert knowledge of the formations depositional processes and the post-depositional features that may be present in the formations, such as fractures.

Stacked water source model

A conceptual MODFLOW groundwater model methodology was used to represent a stacked water source system to study the impact of groundwater pumping on water availability. The impacts have been studied in a variety of scenarios in the report, Impact of groundwater pumping on stacked water sources (PDF 633 KB).

Climate and hydrological modelling

The effects of climate variability and climate change have been a particular focus of the Office. We have used a range of climate and hydrological modelling approaches to help translate estimates of rainfall and evaporation changes from climate change scenarios into impacts on the surface water regimes of river basins across NSW.