2 Introduction to conceptual modelling in bioregional assessments

2.1 General introduction to conceptual modelling in bioregional assessments

The Methodology for bioregional assessments of the impacts of coal seam gas and coal mining development on water resources (the BA methodology; Barrett et al., 2013) defines a conceptual model as:

a qualitative description of the systems and subsystems within a bioregion. It describes the set of hypotheses as to how these systems interact with impacts of CSG and coal mining development and links closely with the qualitative, semi-quantitative and quantitative models used to describe impacts on receptors. Conceptual models in the BAs describe the causal pathway from CSG and coal mining development to the direct, indirect and cumulative impacts on receptors. They may comprise broad-scale coarse-resolution conceptual models within which fine-scale conceptual sub-models are nested to take into account the range of scales over which processes occur.

For the purpose of an individual development, the IESC (2015) in their guidelines for development proposals define the conceptual model as a:

descriptive and/or schematic hydrological, hydrogeological and ecological representation of the site showing the stores, flows and uses of water, which illustrates the geological formations, water resources and water-dependent assets, and provides the basis for developing water and salt balances and inferring water-related ecological responses to changes in hydrology, hydrogeology and water quality.

More generally, conceptual models are abstractions or simplifications of reality. They describe the most important components and processes of natural and anthropogenic systems, and their response to interactions with extrinsic activities or stressors. They provide a transparent and general representation of how complex systems work, and identify gaps or differences in understanding. They are often used as the basis for quantitative modelling, form an important backdrop for assessment and evaluation, and typically have a key role in communication.

Many types of conceptual models are available to serve many different purposes, whether enhancing the understanding of a system, managing ecosystems or communicating how systems work. Models can represent processes (physical, biological) and subjects (rivers, aquifers) at a range of resolutions including spatial (local to landscape), temporal (seconds to decades) and organisational (simple to complex). They can represent relationships (connectivity) and measurement methods (sampling). Examples include control versus stressor models (Gross, 2003), state and transition models (Fischenich, 2008), and conceptual ecological models (driver, stressor, effect, attribute) (Ogden et al., 2005).

Conceptual models can consist of all or some of the following: (i) verbal descriptions, (ii) pictorial and schematic representations of process models (e.g. formal science communication diagrams), or (iii) schematic box and arrow diagrams. The latter are also known as influence diagrams and can be considered ‘a series of working hypotheses connected together by arrows to indicate relationships’ (Commonwealth of Australia, 2014, p. 78). Other types of models include signed digraphs, directed acyclic graphs and other mathematical expressions of dependencies (Hayes et al., 2012).

A conceptual model for a BA consists of:

  • a clearly documented purpose
  • a documentation of the process used to develop the conceptual model
  • the elements of the overall conceptual model (e.g. narrative text, pictorial diagrams, influence diagrams)
  • conceptual sub-model or sub-models for finer scale representations
  • the evidence base underpinning the conceptual model.

2.2 Conceptual modelling philosophy

Bioregions and subregions are large, heterogeneous and complex systems, and the clear articulation of potential causal pathways plays a critical role in focusing the attention of the Assessment teams on the most plausible and important impact modes, the manner in which a hazardous chain of events (initiated by an impact cause) could result in an effect (change in the quality and/or quantity of surface water or groundwater). These pathways describe a series of cause-effect relationships, and underpin the construction of the groundwater and surface water models that are subsequently used to assess the severity and likelihood of impacts. Causal pathways can be an emergent property of these models.

Assumptions about the geological, hydrogeological and hydrological systems are made on the basis on the best available knowledge, and the groundwater and surface water models integrate this understanding and determine if a potential causal pathway is plausible. The pivotal nature of the conceptual model of causal pathways is emphasised by their central position in the workflow in Figure 4 where they sit before the final surface water and groundwater model results.

The conceptual models of causal pathways have the following purposes within BAs:

  • summarise the existing scientific knowledge by collating system knowledge and understanding, and describing how the bioregion or subregion is hypothesised to work and is likely to respond to coal resource development. This requires identifying the important components and processes of the hydrological systems, as well as gaps or differences in understanding. This summary underpins the surface water or groundwater modelling and ensures that all relevant components of the coal resource development and bioregion or subregion are captured in the subsequent stages of the BA
  • provide a general narrative or hypothesis of how the hydrological systems are likely to respond to the coal resource development pathway (CRDP). This requires predicting how the system’s components and processes might react, or, as importantly, not react to the activities associated with coal resource development. These narratives underpin Component 3: Impact analysis and Component 4: Risk analysis
  • identify differences in thinking and model structure uncertainty. Developing and documenting conceptual models with interdisciplinary teams, and with all relevant stakeholders, identifies differences in the mental models that these groups necessarily create. This is important for ensuring that there is adequate breadth in the risk analysis and no surprises occur in later stages of the BA
  • identify the potential for antagonistic and synergistic interactions between system components or processes and coal resource development, thereby providing the first insights into possible system feedbacks and cumulative impacts
  • contribute to the evidence base for the selection of appropriate hydrological response variables and receptor impact variables that will be considered during the receptor impact modelling
  • develop a shared understanding of the science and goals among participants and disciplines within a BA
  • communicate the science behind the BA, the gaps in the science, and the uncertainty associated with it, to users and other interested parties. The conceptual models of causal pathways are a key communication vehicle for the BAs, for explaining complex scientific processes to workshops that are part of BAs, as well as to a broader community (e.g. using two‐ and three‐dimensional visualisation techniques). This should be considered in the presentation of the information.

Documenting the mental constructs of how a system works and how it is affected by coal resource development makes these constructs available for discussion, evaluation and refinement. This is important to achieve a high level of transparency, a key principle of BAs.

Conceptual models summarise current knowledge, hypotheses and assumptions about the bioregion or subregion. However, these models do not represent ‘the truth’, are not final or unmodifiable, and are not expected to be complete or include the entire ecosystem. Multiple or alternative conceptualisations of the system are also possible, and are consistent with the desire to be transparent about the evidence base and knowledge gaps or uncertainties. Conceptual models represent a flexible construct that should evolve as understanding of the system increases (Maddox et al., 2001). BAs cover large spatial extents, and the choices about the nesting and scale of the models are even more important than for conceptual models of an individual coal resource development.

While the focus of BAs is regional, it will be important to learn from ‒ and reconcile with ‒ local conceptualisations (including those included in environmental impact statements) about the hydrogeological and hydrological components, processes and interactions from individual coal resource developments where possible.

The conceptual model of causal pathways synthesises and summarises the important components, processes and pathways in a bioregion or subregion. It is the primary mechanism by which the various contributors across disciplines develop a shared understanding of the BA’s goals. The detail of important components, processes and pathways will typically sit in a number of products but it is valuable to have a concise summary in product 2.3 (conceptual modelling; see Section 4. 1 in this submethodology). Getting the right level of resolution, and avoiding excessive detail and complexity, is essential because it focuses effort going forward. What is not in the model is equally important and needs to be balanced with the need to ensure that other plausible pathways are represented and available for consideration in the risk analysis.

Clarity on the evidence base and potential knowledge gaps or uncertainties can add substantially to the credibility of product 2.3 (conceptual modelling) and the entire BA. Where a link or dependency between ecosystem components can be supported through literature or other information sources, then it should be documented; Nichols et al. (2011) describes an approach for documenting evidence to support cause-effect relationships that could be used. They should also leverage off existing and appropriate conceptual models where possible.

2.3 Causal pathways

A causal pathway is the logical chain of events ‒ either planned or unplanned ‒ that link coal resource development and potential impacts on water resources and water-dependent assets. As an example, a potential pathway between several open-cut coal mines and a natural spring might be initiated by the intentional dewatering of aquifers from mining, leading to a local drawdown of the watertable, which in turn reduces connectivity and groundwater availability for the natural spring. Multiple, often related, causal pathways might potentially be relevant for water-dependent assets and need to be accounted for in the conceptual model. The conceptual model of causal pathways will typically represent multiple and nested systems at different resolutions.

The knowledge about these chains of events – that is, how these impacts might occur – is formally documented in the conceptual models compiled for each BA. Constructing and representing these conceptual models is the primary focus of this submethodology.

In a BA, the identification and definition of causal pathways is supported by a formal hazard analysis, known as Impact Modes and Effects Analysis (Ford et al., 2016). The conceptual models are based on the outcomes of this hazard analysis and the current understanding of the way geological, hydrogeological, and hydrological systems and subsystems in the bioregion or subregion work and interact with each other and the ecosystems and landscape classes in the bioregion or subregion. Constructing and representing these models are discussed in more detail in Chapter 3.

Ultimately, the causal pathways will be represented spatially during the impact analysis. For the purposes of the analysis, it will be as important to say where pathways do not exist as where they do as that may rule out parts of the bioregion or subregion where there is no potential for connection or impact and thus address some of the community’s concern. The causal pathways must be constructed with this spatially-based method of analysis in mind.

2.4 Baseline coal resource development and coal resource development pathway

At the heart of a BA is a comparison of two potential futures, the baseline coal resource development (baseline) and the coal resource development pathway (CRDP). The difference in results between CRDP and baseline is the change that is primarily reported in a BA and it is critical that it is captured in the conceptual model of causal pathways. This change is due to the additional coal resource development – all coal mines and CSG fields, including expansions of baseline operations, that are expected to begin commercial production after December 2012. To understand the potential implications of that difference, the changes over time occurring under the baseline will need to provide important context where possible. For instance, the implications of a difference in drawdown of 1 m may alter substantially if drawdown under baseline is 20 m compared to 0.2 m.

The chain of events in a causal pathway can be considered a series of conditional (or cause-effect) relationships. For instance, the CRDP might impact the physical system and affect aspects of water quantity and quality (as represented by the hydrological response variables), which in turn may affect assets and/or landscape classes as represented by receptor impact variables. While feedback loops are possible, for example where changes in ecology may alter the hydrology, the model implementation assumes that the causal pathway can be compartmentalised and that the conditional relationships occur without feedback loops given the short time frames involved. If feedback loops are expected for some reason it is important that they are described conceptually.

This representation of the chain of events means that the full causal pathways can be usefully divided in two:

  • the causal pathway from the coal resource development to the hydrological changes (represented by the hydrological response variables)
  • the causal pathway from the hydrological changes to the impacts (represented by the receptor impact variables, which are linked to the landscape classes and assets).

The first half of the full causal pathway is reported in product 2.3 (conceptual modelling; see Section 4. 1 in this submethodology) and the second half is reported in product 2.7 (receptor impact modelling; see companion submethodology M08 (as listed in Table 1) for receptor impact modelling). The conceptual model of causal pathways is informed and revised by results from Component 3: Impact analysis and Component 4: Risk analysis.

2.5 Landscape classes and construction of causal pathways

The companion submethodology M03 (as listed in Table 1) for assigning receptors to water-dependent assets (O'Grady et al., 2016) explains the rationale and methods for producing a BA landscape classification. Landscape classes are ecosystems with characteristics that are expected to respond similarly to changes in groundwater and/or surface water due to coal resource development. Given that only a subset of the landscape classes will be hydrologically connected to these hydrological changes (represented by hydrological response variables), causal pathways will be extended to this subset. The causal pathway conceptual models will identify the effect on specific ecosystem characteristics (represented by receptor impact variables). Each landscape class has a characteristic set of hydrological response variables and receptor impact variables. In product 2.3 (conceptual modelling), the landscape classes and causal pathways are listed. Those landscape classes that potentially experience hydrological changes are linked to causal pathways and described in more detail in product 2.7 (receptor impact modelling) and product 3-4 (impact and risk analysis).

2.6 Existing conceptual modelling resources

Conceptual models - and the process of building them - feature prominently in the literature, and play a key role in science synthesis, communication and informed decision making. Examples include the ecological conceptual models of the Healthy Waterways Program for Southeast Queensland (EHMP, 2010) and the Queensland Wetlands Program (Department of Environment and Heritage Protection, 2012). While these represent cause-effect linkages, the aspiration within BAs is to be more explicit about these potential causal pathways.

There are several guides in the literature that discuss various steps, processes and methods for building conceptual models (Shoemaker, 1977; Maddox et al., 2001; Jackson et al., 2000; Gross, 2003; White, 2012).

The Queensland Wetlands Program has produced a guide to pictorial conceptual modelling (Department of Environment and Heritage Protection, 2012) that is widely relevant and describes a step-by-step approach to developing and applying pictorial conceptual models. The Bureau of Meteorology has identified conceptual modelling and documentation of an evidence base as a central process in developing ecosystem and environmental accounts in its Guide to environmental accounting in Australia (Bureau of Meteorology, 2013).

Gross (2003) suggests that conceptual models can be divided into the two categories of ‘control’ and ‘stressor’ conceptual models. Control conceptual models are graphical representations of how the ecosystem is thought to work. They synthesise the current understanding of the key components, processes, interactions and feedbacks in a system. Stressor conceptual models are graphical representations of the activities that may impact on the ecosystem and how the components and processes of that system are likely to respond.

The steps to building a conceptual model outlined by Gross (2003) remain one of the strongest articulations of the process and feature prominently in allied work such as Modelling water-related ecological responses to coal seam gas extraction and coal mining (Commonwealth of Australia, 2014). This particular work has investigated an approach to ecological conceptual modelling to improve the assessment of water-related ecological impacts from coal seam gas (CSG) extraction and coal mining. While the focus is on individual developments and environmental impact statements rather than regional assessments, it shares some similarities with conceptual modelling in BAs: the steps used to build conceptual models; the emphasis on documenting the evidence base that supports conceptual models; and the use of narrative tables for recording hypothesised dependencies and responses. The contrast between the control (natural) and stressor (with coal mining) conceptual models for the Purga Creek Nature Reserve in the case studies is useful, as is the landscape setting (see Figure 12 in Commonwealth of Australia (2014) for a control conceptual model at a subregion-scale resolution).

The development of conceptual models is also described for specific system components relevant to BAs. For example, Barnett et al. (2012), in their guidelines for the best practice development, application and review of groundwater models, discuss the hydrogeological conceptualisation of the groundwater system and identify some key principles that are relevant beyond the groundwater modelling:

  • balancing simplicity of the conceptual model with meeting the objectives
  • balancing the availability of data, and the knowledge base and complexity of the groundwater system of interest
  • considering different views and alternative conceptual models as part of exploring model uncertainty
  • supporting the conceptual model development with observation, measurement and interpretation wherever possible.
Last updated:
26 October 2018