Receptor impact modelling

Executive summary

This submethodology describes the process for assessing impact on, and risk to ecological assets in bioregional assessments (BA) due to coal resource development.

The potential impact on and risk to ecosystems and water-dependent ecological assets due to coal resource development is quantitatively assessed using receptor impact models. They translate predicted changes in hydrology at specified points in time into a distribution of ecological outcomes that may arise from those changes. A receptor impact model predicts the response of a receptor impact variable (an ecological indicator, such as annual mean percent canopy cover of woody riparian vegetation), to one or more hydrological response variables (for example, dmax, maximum groundwater drawdown due to additional coal resource development). Receptor impact models in a bioregion or subregion are developed at landscape class level, and any ecological implications for a water-dependent asset subsequently consider predictions of receptor impact variables for landscape classes that intersect that asset. For BA purposes, a landscape class is defined as an ecosystem with characteristics that are expected to respond similarly to changes in groundwater and/or surface water due to coal resource development.

Economic water-dependent assets are confined to groundwater and surface water management zones or areas and comprise specific water access entitlements or rights and other water supply features or infrastructures. The scope of BA precludes detailed analysis of economic impacts, and is limited to the assessment of changes to the availability and reliability of the relevant groundwater or surface water. Potential impacts to economic assets are tied more directly to hydrological response variables and specific management thresholds (e.g. cease-to-pump rules), than for ecological assets. As a result, receptor impact models are not considered for economic assets.

Water-dependent sociocultural assets are classified into ‘Heritage site’, ‘Indigenous site’ or ‘Recreation area’ classes. Receptor impact modelling is not considered directly for sociocultural assets and the impact and risk is limited to characterising the hydrological change that may be experienced by those assets. However, a receptor impact model for ecological assets may also play an important role in assessing any change for those sociocultural assets that are ecological in character.

This submethodology is constructed to address the complexity of ecological systems, propagate and communicate the uncertainty, and synthesise the current state of knowledge. The specified time points of interest include a reference assessment year in 2012, a short-term future assessment year in 2042 and a long-term future assessment year in 2102. These are selected to represent the current state of the ecosystem, the ecosystem close to peak production and the ecosystem following coal resource development. The receptor impact model construction process depends crucially on the definition of a landscape class. Each bioregion or subregion has multiple landscape classes, and only those that may be potentially impacted from hydrological change due to additional coal resource development are considered for receptor impact modelling.

For a given landscape class, a qualitative mathematical modelling approach is used with independent ecological experts to identify the key ecosystem components and dependencies, and importantly, their link to groundwater and surface water variables. These hydrological variables are then translated into hydrological response variables that can be estimated by the groundwater and surface water hydrological models. The qualitative models are used to assess potential candidate receptor impact variables and hydrological response variables to consider within a receptor impact model.

A suite of hydrological scenarios based on plausible combinations of hydrological response variables is developed for each receptor impact model. These scenarios are presented to independent ecological experts as part of an expert elicitation process, and the distribution of receptor impact variable under each scenario elicited. A statistical model is constructed based on the experts’ responses to estimate the relationship between the receptor impact variable and the relevant hydrological response variables, and to enable predictions to be made for combinations of hydrological response variables not considered in the elicitation. Within a landscape class, probabilistic predictions from the model can then be made at specific locations based on the modelled changes in those hydrological response variables at that location in different time periods. These predictions may be extended to regions, such as reaches of river, by applying the receptor impact models at different locations and using the changes in hydrological response variables at each of those locations. It is important to note that predictions of a receptor impact variable at a specific location is a prediction of the likely response across the landscape class for the change in hydrology at that location.

The receptor impact model for a landscape class therefore permits the prediction of how cumulative hydrological changes may change a receptor impact variable at key time points in the BA futures. This represents an essential component of the impact and risk analysis (product 3‑4), and is a key vehicle by which impacts on and risks to water-dependent ecosystems and assets are assessed. Distributions of change for hydrological response variables and receptor impact variables are summarised in BAs by a limited series of percentiles (or quantiles), nominally 5 percent increments between the 5th and 95th percentiles, for different futures (baseline and coal resource development pathway (CRDP)) and the specified time frames of interest. For a specific water-dependent asset, predictions of receptor impact variables represent an important line of evidence, but need to be considered in conjunction with the qualitative mathematical models, the wider suite of hydrological changes, detailed information about the presence and condition of that asset, and other more local information.


Last updated:
18 October 2018