Skip links and keyboard navigation


A scientific model is an attempt to represent observed objects, phenomena and processes in a logical and objective way. While all models present simplified versions of reality, their aim is to construct a formal system whose theoretical consequences are not contrary to what is observed in the ‘real’ world. In other words, they may be wrong in some precise detail, but they are not fundamentally wrong[13].

Models are useful for predicting outcomes in the observed world such as the existence of Higgs boson (an elementary particle, once hypothetical, now shown to be real), the future climate of the earth or the effect of increasing aquatic connectivity of wetlands on fish breeding.

'Pictures worth a thousand words' Image by Queensland Government

Quick fact

"All models are wrong, but some are useful"
- Box, 1987[2]

Models can represent the observed world in different ways, such as mathematical or numerical models, descriptive text, tables, box-and-arrow diagrams and pictorial conceptual models. Each of these kinds of models work well for some applications and not so well, or even very poorly, for others. Also, human variation in thinking and perceptions influence the effectiveness of models for communicating scientific ideas. Different people relate to particular types of models and not others. As no model can fully explain the complexity of the issue it is trying to address, comprimises need to be made in the amount of data and information models can deal with, which can lead to key information not being considered. It is sometimes useful to develop composite or multiple models to communicate the same message to different people[18]. It is also essential to critically review and understand the limitations of a model in decision making before using it.

What does modelling do?

A model is an abstraction or simplification of reality. Scientists often use models to explore systems and processes they cannot directly manipulate and to fill gaps in understanding when there is insufficient information.

Models can be more or less quantitative, deterministic, abstract, and empirical. They help define questions and concepts more precisely, generate hypotheses, assist in testing these hypotheses, and generate predictions.

Model building consists of determining system parts, choosing the relationships of interest between these parts, specifying the mechanisms by which the parts interact, identifying missing information, and exploring the behaviour of the model. The model building process can be important in itself, because it reveals what is known and what is unknown about the connections and causalities in the systems under study[8].

Models can be used to[7]:

  • support a methodology for capturing, mapping, and consolidating the various sources of knowledge and data required for understanding systems in a systematic way;
  • provide a platform for bringing stakeholders and scientists together to develop a shared understanding of the problem and possible solutions;
  • set up experiments that examine a range of scenarios;
  • perform quantitative assessments of options;
  • fill gaps in understanding when there is insufficient information.

A typical modelling and assessment process has five key phases (modified from Hamilton, 2015[7]):

  1. Scoping, including a model study plan that identifies model’s purpose, study objectives, stakeholders and issues of concern;
  2. Problem framing and formulation, including conceptualisation, linking drivers and responses;
  3. Analysis and assessment of options, including model setup, calibration, sensitivity testing and validation;
  4. Peer review of the method; and
  5. Communication of findings, including simulation and evaluation.

Best practice and robust modelling to achieve an intended outcome must[11][20]:

  • clearly identify the end-user and objectives of the modelling exercise;
  • document the nature (quantity, quality, limitations) of the data used to construct and test the model;
  • provide a strong rationale for the choice of the model and its features (including a review of alternative approaches);
  • justify the techniques used to calibrate the model and conduct detailed analysis, testing and discussion of model performance;
  • make a resultant statement of model assumptions, utility, accuracy, limitations, and scope for improvement; and
  • include a peer review of the method, and sensitivity testing (to determine which variables have most impact on the results).

Types of modelling

Models can be used to simulate and investigate various processes including:

  • Hydrological models
  • Biogeochemical process models
  • Ecohydrological modelling
  • Bayesian networks
  • Species population modelling (e.g. fisheries management models for setting catch and harvest quotas)
  • Climate models
  • Watershed management models for nutrient control strategies
  • Risk assessment models for environmental engineering
  • Mapping - model of components.

Modelling of processes

Processes can be represented in models in a number of ways.

Deterministic models link the outputs to a number of independent variables (inputs, state, initial or boundary conditions). These can be divided into three types:

  • models based on the laws of physics (in particular hydro- and thermodynamics), chemistry, biology etc., and represented by equations.
  • conceptual models reflecting physical laws in a simplified manner and involve a degree of observation and understanding of the system.
  • empirical models which do not explicitly take into account of physical laws but only the case-effect relationship of system inputs to outputs in a general and empirical manner.

Stochastic models, in contrast, do not consider the causality. They are generally probabilistic models represented by a probability distribution function of the variables, often described in terms of parameters such as averages and standard deviations, based on historical data[1].

Species Modelling

Species distribution models (or SDM's), also known as ecological niche models (ENM), or habitat suitability models (HSM), are used to predict the distribution of a species based upon known/inferred relationships to the environment[6]. The models provide a geographic estimate of a species' potential distribution area, which can be used to identify priority areas for conservation, restoration and to target areas to survey for rare and little known species. Distribution models are also used to determine areas of potential spread of introduced species, deadly diseases or pests, and predict the potential threats to a species due to habitat loss, climate change, or other causes[4][17][6]. They are now widely used across terrestrial, freshwater, and marine environments[5].

SDMs typically requires at a minimum two sets of data[16]. The first is observations of a species location, i.e. where species are present and preferably also where they are absent (the response variable). The second dataset contains information about environmental conditions that influence a species' disitribution, such as temperature, rainfall, vegetation cover and incoming radiation (the explanatory variables).

Some important considerations in developing a species distribution model include appropriate selection of: the study area; a representative sample of occurrence records; ecologically relevant environmental predictors (i.e. with respect to the taxa and given the study area); and modelling algorithm(s)/approach to be used. In addition, the availability, relevance, resolution and accuracy (attribute, temporal and spatial) of input data may impact end use.

Continental scale analyses that use course resolution bioclimatic predictors maybe suitable for assessing broad current and/or future climate envelopes for example. However, if the end use is targeted reserve design, adoption of a finer scale approach that uses high resolution bioclimatic and biophysical variables and is constrained for example to a taxa’s current known realised niche, maybe more appropriate[14][15].

Conceptual Modelling

Conceptual models are abstractions of reality expressing a general understanding of a more complex process or system. They tell the story of how a system works. Conceptual model building consists of choosing the system parts and the relationships that link these parts, specifying how the parts interact and identifying missing information[18].

The sections below provide a brief overview of different types of conceptual models.

Descriptive texts

Descriptive texts and tables provide text-based summaries of concepts and relationships. Generally pitched at more specialist audiences and often couched in technical language, such texts can be time consuming to interpret[18].

Diagrammatic conceptual models

Diagrammatic conceptual models use iconic representations such as boxes, arrows and pictures. Diagrams can be seen as a type of visual formatting, a way of showing relationships and abstract or conceptual information, rather than quantitative data


Conceptual or qualitative models show the main elements and flows of material, information, and causation that define a system. They require model builders to explain why the chosen elements are important, what assumptions are being made, and how key concepts are defined[8].

Examples of diagrammatic conceptual models are: 

Pictorial conceptual models

Pictorial conceptual models—also known as conceptual diagrams or models—use drawings and diagrams to explain how a system, such as a wetland, works. The models are powerful tools as they offer a way of visualising complex environmental processes. They can easily be used by a variety of audiences with varying levels of knowledge. See the Pictorial Conceptual Models page for information on how to develop conceptual models and a variety of examples.

Hydrological (water) modelling

Hydrological models are simplified, quantitative representations of a part of the hydrological (water) cycle. They are typically used to understand hydrological processes and generate hydrological prediction.

These models represent the processes of the water cycle processes, namely[10]:

Why use hydrological modelling?

A hydrological model can be used to estimate catchment flows (sometimes called water yields) for areas of interest in a catchment and can also be used to generate future estimates or scenario modelling. Forecasts of future seasonal stream flow records are valuable to a range of water managers and users, including water resource planning authorities, irrigators, urban and rural water supply authorities, environmental managers, and flood engineers. Such forecasts can inform planning and management decisions relating to available water resources and form the basis of water quality modelling. An example of this type of modelling includes the Paddock to Reef (P2R) modelling for the catchments of the Great Barrier Reef.

Generally, there are only a small number of stream gauging sites within a catchment, or a station may have only been established for a short period of time, and therefore not provide a long enough record to establish long term yields or trends. This makes it difficult to establish flows within a catchment. Having a long-term record allows statistical analysis to be performed over a period that is not influenced by extremes of droughts or floods[9].

Types of hydrological (water) models

Hydrological or catchment models simulate conditions over a broad catchment[20]. This may consist of a water balance or rainfall-runoff model.

Water balance models:

  • Investigate components of the hydrological cycle
  • Identify sources and sinks
  • Use mass balance, i.e. inflow vs. outflow
  • Advance conceptual model understanding
  • Quantify the importance/contribution of a source or sink.

Rainfall-runoff models can be used to:

  • generate runoff characteristics for a catchment from rainfall;
  • create timeseries of flow (wet events and baseflow);
  • assess catchment characteristics (e.g. land use, slope, area);
  • estimate the timing and volume of runoff.

Hydrodynamic models:

  • are used to predict dynamic movement of surface water within a waterbody (wetland, stream, river, lake, estuary, ocean), in terms of water flow, depth, and velocity.

Groundwater flow models:

  • predict dynamic movement of water within the ground.

Coupled water quality models:

  • Add water quality constituents and/or processes to other hydrological models.

Biogeochemical processes and how they link to hydrological modelling

Biogeochemical models describe the behaviour and cycling of water and a variety of elements within ecosystems[12].

Biogeochemical models have various degrees of complexity from simple decay equations, for example, the loss of leaf mass through decomposition, to large integrated ecosystem models capable of simulating dynamic processes such as energy flows, nutrient cycling, and hydrological flows[3].

Models vary in spatial and temporal scale. Processes, such as denitrification, can operate at a micro-scale, while others operate at landscape or catchment scale. Scaling up can be difficult due to changes in the relative importance of processes at different scales, data availability, and aggregation. There are trade-offs between scale and model complexity. The more parameters a model contains, the more difficult it is to derive and calibrate them directly from available data or even indirectly[12].

Biogeochemical models can be coupled to hydrodynamic models to create integrated models at different levels of complexity.

Outputs of modelling

Hydrological models can produce output data describing:

  • Flows – runoff and groundwater
  • Water levels
  • Flood levels.

When coupled with biogeochemical models they can generate:

  • water quality constituent concentrations;
  • aquatic biogeochemistry;
  • biotic habitat;
  • aquatic ecosystem dynamics.

Choosing the right model

It is important to choose the right kind of model for the purpose. The Queensland Water Modelling Network (QWMN) has a number of publications that can assist including the QWMN Wetland Hydrology Models Review, the Queensland Water Modelling Network Strategic Review of Models Suite, and the Queensland Water Modelling Network Water Model Catalogue.


A model is only as reliable as the calibration. A model’s robustness relies on[19]:

  • Calibration data (i.e. stream flow data, catchment characteristics)
  • Boundary data (i.e. rainfall spatial and temporal distribution in catchment)
  • The experience and skill of the modeller.

If there is no calibration data, it is important to carry out sensitivity testing to simulate the model with parameters adjusted through a range of higher and lower input values than the initial estimates. This provides a measure of model uncertainty.

A modeller can also check model output ranges by comparing the modelled catchment to a gauged catchment with similar characteristics (i.e. latitude, coastal proximity, slope, land use), and check that the model parameters, such as the runoff coefficient, are reasonable.

Model review

Model users also need to ensure that the model is fit for purpose. Many aspects of a model are relevant to its accuracy and usefulness[19]:

  1. Process knowledge – Do the modellers and end-users agree on the model outputs and does it align with their conceptual understanding of the system?
  2. Temporal sensitivity – Does the model account for temporal events and other key processes?
  3. Spatial sensitivity and hydrogeology – Is the spatial scope appropriate? Has hydrogeology been incorporated correctly? Is it too simplistic or complex? Does it include all key hydrogeological and coastal processes if relevant?
  4. Cumulative impact – Does the model incorporate the impacts of earlier and potentially later landuse changes? Is climate change included?
  5. Data – Does the model development and calibration include appropriate and sufficient data?
  6. Communication – Have stakeholders been engaged sufficiently? Is the model described adequality? Is the model well described or does the report obscure details or overload reviewers with text? Are the figures easy to understand? Is sufficient review time allowed?
  7. Impact bias – Does the model focus on limited areas or impacts? Is uncertainly addressed? Is the proposed solution suitable or does it risk worse impacts?
  8. Tools – Are the correct tools used? Is the model too simple, too complex or inappropriate for the purpose? Has uncertainly analysis and sensitivity testing been carried out? Has the model been validated?
  9. Peer review – Has the model been peer reviewed? Has there been field testing of ideas? Are the QA/QC processes adequate? Has local data been used? Is the model highly complex? Does it make sense? Do the results support the conclusion? Where in decision making process does the model fit?
  10. Implementation – Will the proposed solution work as modelled? Is maintenance addressed? Are the controls practical? What is the timing of the offsets claimed? Is the modelled result achievable?

Restoring blue carbon ecosystems: a best practice guideline for hydrologic assessments

QWMN Wetland Hydrology Models Review

Wetland Management Tools and Guides

Information sources for aquatic ecosystem rehabilitation planning

Water quality, water quantity and aquatic ecosystem monitoring

Pages under this section


  1. ^ Becker, A & Serban, P (1990), Hydrological models for water resources system design and operation, vol. 34, Secretariat of the World Meteorological Organization, Geneva, Switzerland.
  2. ^ Box, GEP & Draper, NR (1987), Empirical model-building and response surfaces, John Wiley & Sons.
  3. ^ Ecosystem Biogeochemistry, 'Biogeochemical Models', Springer.
  4. ^ Elith, J, Graham, CH & Anderson, RP (2006), 'Novel methods improve prediction of species' distributions from occurrence data', Ecography, vol. 29, no. 2, pp. 129-151.
  5. ^ Elith, J & Leathwick, JR (2009), 'Species Distribution Models: Ecological Explanation and Prediction Across Space and Time', Annual Review of Ecology, Evolution, and Systematics, vol. 40, no. 1, pp. 677-697.
  6. ^ a b Franklin, J (2009), Mapping species distributions: spatial inference and prediction, Cambridge University Press.
  7. ^ a b Hamilton, S, El Sawah, S, Guillaume, JHA & Jakeman, AJ (2015), 'Integrated assessment and modelling: a review and synthesis of salient dimensions', Environmental Modelling and Software, vol. 64, pp. 215-229.
  8. ^ a b Heemskerk, M, Wilson, K & Pavao-Zuckerman, M (2003), 'Conceptual models as tools for communication across disciplines', Conservation Ecology. [online], vol. 7, no. 3. Available at:
  9. ^ Hydrological Modelling: How and Why it is Used. [online], Tasmania, ed. B Graham. Available at:
  10. ^ Introduction to Hydrological Models (Module 1.2) (2015). [online], eds. University of Oklahoma & HyDROS. Available at:
  11. ^ Jakeman, AJ, Letcher, RA & Norton, JP (2006), 'Ten iterative steps in development and evaluation of environmental models', Environmental Modelling and Software, vol. 21, pp. 602-614.
  12. ^ a b Kros, J (2002), Evaluation of biogeochemical models at local and regional scale, p. 286, Wageningen University, Netherlands.
  13. ^ L. Apostel (1961), The Concept and the Role of the Model in Mathematics and Natural and Social Sciences. [online], vol. 3, Springer, Dordrecht. Available at:
  14. ^ Manzoor, SA, Griffiths, G & Lukac, M (2018), 'Species distribution model transferability and model grain size – finer may not always be better', Sci Rep, vol. 8, no. 7168.
  15. ^ Merow, C, Smith, JM & Silander, JA (2013), 'A practical guide to Maxent from modeling species' distributions: what is does, and why inputs and settings matter', Ecography, vol. 36, pp. 1058-1069.
  16. ^ Pearson, RG (2007), 'Species' distribution modeling for conservation educators and practitioners', Lessons in Conservation, vol. 3, pp. 54-89.
  17. ^ Sangeeta, R, Ashish, S, Santanu, R & Kumar, SS (2002), 'Use of species distribution models to study habitat suitability for sustainable management and conservation in the Indian subcontinent: A decade's retrospective', Frontiers in Sustainable Resource Management, vol. 1.
  18. ^ a b c d Tilden, J, Baskerville, H, Lammers, H, Ronan, M & Vandergragt, M (2012), Pictures worth a thousand words: a guide to pictorial conceptual modelling, Queensland Government, Brisbane.
  19. ^ a b Water Research Laboratory (2021), Assessing model applicability.
  20. ^ a b Water Research Labratory (2021), 'Wetland Hydrodynamic and Hydrologic Modelling', (Unpublished).

Last updated: 31 May 2023

This page should be cited as:

Department of Environment, Science and Innovation, Queensland (2023) Modelling, WetlandInfo website, accessed 18 March 2024. Available at:

Queensland Government
WetlandInfo   —   Department of Environment, Science and Innovation