1. Movement Assignment modelling
By dividing the study area into smaller zones, an origin-destination matrix is created and assigned to the network, using strategic highway modelling software (e.g. Saturn, Visum). This permits a form of route-choice analysis based on algorithms used for highway traffic assignment (distance/time). This type of model can forecast movement flows and can sit on a GIS platform.
Unobserved walking volumes and missing origin destination pairs will not be included. A model’s accuracy can therefore only be guaranteed at observed locations. There will be little or no consideration of any evidentially supported correlation between land use patterns, transport arrangements and observed pedestrian counts.
Origin-destination zone sizes may be too large in places for individual street flows to be presented. The model will therefore not be wholly reactive to fine-grained changes brought about by urban redevelopment. In addition, assumptions of future origin-destination pairs need to be made. Models are time-consuming and expensive to develop for large areas.
2. Dynamic Entity modelling – micro simulation
Micro simulation modelling is typically used for the analysis of building interiors and small urban areas in “closed system”, highly programmed environments such as public transport interchanges where origins and destinations can be pre-set. It is therefore less suitable for larger network assessments where many variables affect journeys and route choices are not highly programmed.
Although visually attractive, providing animated representations of flows, micro simulation models are labour and time intensive and require a much higher degree of data collection. As with assignment methods, there will be no consideration of or correlation between land use patterns, transport arrangements and observed pedestrian counts. This type of modelling is less reactive to fine-scale changes in the public realm brought about by urban development.
3. Space Syntax Multi-Variate modelling
Space Syntax multi-variate modelling permits analysis of complex environments in a straightforward and cost-effective manner that integrates with GIS and BIM environments. The dynamic and evidence-based approach does not rely on origin destination data or assumptions but does integrate spatial layout, land use and transport attraction characteristics.
Space Syntax modelling is therefore faster, less cumbersome and less prone to incorrect assumptions than either Movement Assignment or Dynamic Entity modelling. The approach allows simultaneous analysis of multiple scales of movement from local to regional, national and international.
Space Syntax spatial analysis software transforms a street pattern into a network graph by disaggregating the network at the intersections. The distance cost between two line segments is measured using the ‘shortest’ path between the pair which is then weighted by three key cost relations: metric (least length), topological (fewest turns) and geometrical (least angle change).
The “spatial accessibility value” assigned to each segment reflects the complexity of routes from that segment to all the others within the system. First, an accessible segment is more easily reached than a segregated one because it can be arrived at by simpler routes from other segments – thus it should receive a high degree of “to” movement. Second, a more accessible segment should be more likely to be selected as part of a route between other pairs of segments: that is, it will attract more “through” movement.
It is the combination of their role in “to” and “through” movement that make spatial accessibility values powerful in estimating movement potentials.