A concept describing the possibility and the difficulty to reach a specific location when moving throughout space
Anisotropic function:
A function that models the change of influence of a feature property with respect to its relative positional direction. This
function is used to weight the friction/force effect (see Friction, Anisotropic skewed surface)
Anisotropic skewed surface:
Space can be modelled with different degrees of complexity, such as an isotropic plane surface, or an isotropic skewed surface
or an anisotropic skewed surface. The anisotropic skewed surface is the third level of complexity to model space as an heterogeneous surface. Each location in space has a specific influence
on movements, but this influence is not constant, instead it varies with respect to the relative direction of movement
Barrier:
A barrier is a location that blocks movements due to its thematic property. It is often called "absolute barrier" and is modelled
with an infinite friction coefficient value (see Friction)
Calibration stage:
The application of a predictive model is made of 2 stages: calibration and prediction stages. Before a possible use for predicting properties in an undocumented period of time, it is necessary
to calibrate (adjust, assign appropriate values to) the parameters of the model to suit the period of time validated with
measurements
Cellular automata:
Cellular automata are dynamic systems used to model spatial changes such as spatial diffusion, movements, … Cellular automata
are dynamic agents obeying to rules that modify properties in space
Change index (global):
A change index is an indicator derived from multitemporal measurements. It expresses the amount of change within a period
of time. It can describe the change behavior of a set of features (global) or of individual features. It can result from a
difference, a ratio, ...
Contagious expansion diffusion:
The process of change in the spatial distribution of thematic properties throughout time can be approached from different
viewpoints: a contiguous or dispersed spread of presence, a densification, … The diffusion process can be spatially dependant
(spatial dependant diffusion) or independant (non-spatial dependant diffusion). A contagious expansion diffusion is spatially dependant as the phenomenon spreads continuously from a location to contiguous neighbours, in a manner of a
contagious disease
Continuous spatial distribution:
Properties of a phenomenon can be distributed throughout space in a continuous or discontinuous manner. A continuous spatial distribution is characterised with gentle changes in property values throughout space. This is due to a continuous spatial dependancy
throughout space. Thus the spatial distribution corresponds to a continuous surface rather than a set of spatial objects (see
Discontinuous spatial distribution)
Cost distance:
When modelling access or movement in space, the usual euclidian distance -called also plane or horizontal distance- is not
optimal to express the access to a destination. A most efficient concept is the cost-distance that express the amount of resources (financial, time, energy, ...) necessary to move throughout space (see Friction, Barriers)
Densification:
The process of change in the spatial distribution of thematic properties throughout time can be approached from different
viewpoints: a contiguous or dispersed spread of presence, a densification, … The densification expresses an increase in frequency
of concerned items or individuals within a specific location or spatial unit of observation. In the process of diffusion of
innovation, newly settled locations show an increase in the number of people adopting this innovation, this is known as densification
(see Spatial diffusion)
Discontinuous spatial distribution:
Properties of a phenomenon can be distributed throughout space in a continuous or discontinuous manner. A discontinuous (discrete) spatial distribution is characterised by strong changes in property values throughout space. Places with value changes delineate spatial feature
boundaries (point, linear or areal) (see Continuous spatial distribution)
Extended neighbourhood:
The neighbourhood is an area surrounding a spatial feature or a spatial unit of observation (object or cell). An extended neighbourhood generally corresponds to a set of spatial features (objects or cells) that are contiguous to the central feature
in a wider sense: they at least share one vertex with it (see spatial neighbourhood and limited neighbourhood)
Forces of movement:
When modelling movement in space, forces and frictions are factors that influence movement. Force can be internal to the moving
body, but also external (wind, …). In the computation of cost-distance, the influencing forces are modelled as a friction coefficient value attached to each location in space. The friction value
is less than 1, but greater than 0 (see friction)
Friction:
When modelling movement in space, impeding properties of space are expressed as frictions. Friction coefficient values can
range from 1 (no friction) to a maximum (infinite value) corresponding to an absolute barrier. In the computation of cost-distance the assigned friction value is combined with the geometric plane distance. The concept of a friction coefficient can be extended
in order to encompass the concept of force. In this case the friction value is less than 1, but greater than 0 (see force
of movement)
Game Of Life:
A famous computer simulation illustrating the birth, growth and death of cells distributed throughout a regular gridded space.
It demonstrates the use of a cellular automata model, using very simple rules, to simulate a spatial diffusion process
Hierarchical diffusion:
The diffusion process can be spatially dependant (spatial dependant diffusion) or independant (non-spatial dependant diffusion). In this second situation the diffusion process occurs without spatial continuity, but it is based on the hierarchical importance
of centers (places) related with the innovative activity under investigation
High-pass filter:
A technique to enhance the local distribution of property values. In the spatial context it makes use of a moving window that defines the neighbourhood.
A gradient value is computed within the moving window and assigned to the central cell (see low-pass filter)
Isotropic plane surface:
Space can be modelled with different degrees of complexity, such as an isotropic plane surface, an isotropic skewed surface,
or an anisotropic skewed surface. The isotropic plane surface is the simplest way to model space as a homogeneous plane surface controlled by euclidian geometric properties
Isotropic skewed surface:
Space can be modelled with different degrees of complexity, such as an isotropic plane surface, an isotropic skewed surface
or an anisotropic skewed surface. The isotropic skewed surface is the second level of complexity to model space as a heterogeneous surface. As each location in space has a specific influence
on movements, the euclidian distance concept is replaced with the cost-distance concept
Land Use/Cover Change (LUCC):
The cover or the use of the earth surface is described with a set of category types. Throughout time these cover or use categories
can change. The objective of a LUCC analysis is to describe, to understand or to predict such changes in the spatial distribution
of landcover or landuse
Limited neighbourhood:
The neighbourhood is an area surrounding a spatial feature or a spatial unit of observation (object or cell). A limited neighbourhood generally corresponds to a set of spatial features (objects or cells) that are contiguous to the central feature
in a limited sense: they at least share one side with it (see spatial neighbourhood and extended neighbourhood)
Linear regression function:
A regression function that relates a dependant variable Y with one or several independant variables Xi in a linear manner.
A first degree polynomial function is a linear function (see Polynomial regression function)
Low-pass filter:
A technique to smooth the distribution of property values. In the spatial context it makes use of a moving window that defines the neighbourhood.
A central tendancy value is computed within the moving window and assigned to the central cell (see High-pass filter)
Markov chain (analysis):
A technique to estimate the probability of occurrence from any original state to any final state after a specific sequence
of n time steps. It makes use of transition matrices
Movements in space:
Actions to move from one location to another through space. The moving feature can be activated by external or internal forces
and its path is influenced by local frictions of space
Neighbourhood (spatial):
The neighbourhood is an area surrounding a spatial feature or a spatial unit of observation (object or cell). Based on the
spatial dependancy assumption, it stated that the neighbourhood of a feature influences its properties. A neighbourhood is
defined by a range and a shape that encompass the feature up to a certain distance
Non-spatially dependent diffusion:
The process of change in the spatial distribution of thematic properties throughout time can be approached from different
viewpoints: a contiguous or dispersed spread of presence, a densification, … The diffusion process can be spatially dependant
(spatial dependant diffusion) or independant (non-spatial dependant diffusion). In this second situation the diffusion process occurs without spatial continuity
Path (trajectory):
The trajectory used to move from an origin to a destination in space
Polynomial function:
A non-linear function that relates a dependant variable Y with one or several independant variables Xi. This function contains
a combination of polynomes at different degrees or exponents (see Linear regression function)
Prediction stage:
The application of a predictive model is made of 2 stages: calibration and prediction stages. Once the model has been calibrated to suit the period of time validated with measurements, it can be used to predict
feature properties estimated during a period of time with no existing measurements (generally a future period of time)
Relocation diffusion:
A type of spatial diffusion process where previous locations of presence are replaced by new locations throughout time
Scenarios of development:
A scenario of development describes the possible evolution of properties of a phenomenon, based on specific rules of change
and specific characteristics (parameters) at the starting date. A set of different scenarios can be defined in order to simulate
possible alternatives of development
Sleuth (model):
The SLEUTH name is the acronym for Slope, Landuse, Exclusion, Urban, Transportation and Hillshading. The Sleuth model is
a cellular automata simulation model developed to model the development of urban areas (Clarke et al., 1998)
Spatial diffusion:
The process of change in the spatial distribution of thematic properties throughout time can be approached from different
viewpoints: a contiguous or dispersed spread of presence, a densification, … The diffusion process can be spatially dependant
(spatial dependant diffusion) or independant (non-spatial dependant diffusion)
Spatial dynamics modelling:
Procedure used to summarise the changes in the spatial distribution of phenomenon properties throughout time (see Spatial
dynamics)
Spatial expansion:
The impact of the presence of a phenomeon throughout space. In a spatial diffusion process, the phenomenon spatially expands
throughout time
Spatial filtering:
Procedures used to isolate spatial components: local and regional variations, trend. Typical filtering techniques make use
of a moving window for the smoothing of local variations (low-pass filters) or their enhancement (high-pass filters)
Spatially dependent diffusion:
The process of change in the spatial distribution of thematic properties throughout time can be approached from different
viewpoints: a contiguous or dispersed spread of presence, a densification, … The diffusion process can be spatially dependant
(spatial dependant diffusion) or independant (non-spatial dependant diffusion). In this second situation the diffusion process occurs with spatial continuity
Temporal variability:
Some properties of space or spatial features can change throughout time, as others remain constant. Their temporal variability
is then different
Trend surface (analysis):
A regression function modelling the property values (Z) based on their location (X,Y) in space: Z = f(X,Y)