Why another model?

Sustainability is one of today's major public concerns. There is evidence that transport and land use systems of cities all over the world are unsustainable. Indicators backing up this hypothesis are among others:

urban sprawl, pollution and consumption of non-renewable resources.

As urban planning has become increasingly complex, decision support tools are essential to help to achieve the overall objective of sustainability. Recent research has shown that any single policy instrument cannot achieve sustainability. Policy strategies employing several instruments are needed to be successful. The use of formal models and optimisation methods is suggested to be used to identify the best performing strategy.

Why another model?

Research of the last decades has shown that land use and transport form a closely linked dynamic system. Therefore integrated land use and transport models are needed to assess the performance of urban policy strategies. A literature review has shown that a variety of operational transport and/or land use models exists. The current trend in land use and transport modelling is characterised by an extreme disaggregation even down to individual household level. This extremely detailed modelling approach is, independently from its theoretical appeal, inappropriate for identifying sustainable policy strategies. Despite the ever-increasing computational power, model runs take too long to be able to consider a reasonable number of instruments. Additionally data needs are very high. Even synthetic data have to be produced to match the level of model disaggregation. Data requirements might be one of the reasons why the use of integrated land use and transport models is still not widespread. A different approach was therefore chosen here.

The rather high aggregated integrated, dynamic urban land use and transport model MARS was developed as the core of a sustainability assessment framework. The underlying hypothesis is that cities are self-organising systems and that the principles of synergetics can be applied to describe collective behaviour. A qualitative model was developed based on Viennese research. The method of causal loop diagrams was used for this task. From this basis a quantitative model was built and written into code. MARS was calibrated for the city of Vienna. An extensive model-testing program was carried out using observed data from 1981 to 2001. A back casting exercise and sensitivity tests have proven the usability of MARS.

Decision making process in land use and transport planning

DMP in land use and transport planning

Metropolitan Activity Relocation Simulator - MARS

MARS (Metropolitan Activity Relocation Simulator) is a strategic and dynamic Land-Use and Transport Interaction (LUTI) model. The basic underlying hypothesis of MARS is that settlements and activities within them are self-organizing systems. The development of MARS started in the year 2000.

On the highest level of aggregation MARS can be divided into the following two interconnected sub-systems:

The land use sub-system of MARS can be divided into two further sub-systems:

Cause effect digrams

Transport model part

Land use model part

MARS - MARS FS - Decision Makers Tool

MARS - MARS FS - Decision Making Tool

Detailed introduction of the MARS Flight Simulator

We are currently working to integrate the model MARS with a dynamic GIS system to allow a spatial-temporal presentation of MARS simulation results.

Observations and outlook


Within the project SPARKLE MARS FS has been used as a decision support tool in a series of workshops in Thailand and Vietnam. Here is a summary of the conclusions and comments of the workshop participants:

  • MARS FS was found easy to navigate
  • The causes-tree tool was found extremely useful to understand the behaviour of the model
  • Possibility to work interactively with the model and test each policy instrument individually
  • Explore synergetic effects of policy combinations
  • Outputs as graphs or tables, and exporting function
  • Minor criticism was the list of pre-prepared indicators