Approaches in Spatial Analysis

 

In his doctoral thesis titled Spatial Analysis in Support of Physical Planning (2003),  Eric Koomen emphasizes the important role of spatial analysis in the formulation and evaluation of physical planning initiatives. To carry out this role, one must be knowledgeable about the application of different approaches in spatial analysis in such a way that it will correspond or will respond to the planning issues at hand.

 

Koomen enumerates these spatial analysis approaches as follows:

 

1.    Transformation

Transformation methods form the basis of data visualisation and essentially change a certain form of data representation into another to enhance specific features.

 

Two commonly applied transformation methods are: classification and filtering.

 

Classification is used to diminish the variability in data values and can emphasize a certain portion of a spatial data set. By adjusting the classification in a visual representation of a data set specific phenomena can be enhanced or obscured, indicating that the selection of the appropriate class boundaries is crucial.

 

Filtering changes the value at each location in a data set based on the original values at that location and its surroundings.

 

2.    Aggregation

 

Spatial aggregation methods reduce the individual values of a data set to a single value for a specified region or the whole study area. The latter aggregation reduced the number of spatial dimension of the data set from 2 to 0, creating a non-spatial indicator or index value. This loss of spatial information is compensated by the delivery of a clear, unequivocal summary of the original content. Aggregation can also be performed at a regional level, producing a new much coarser representation of the original data. Spatial aggregation methods either deliver spatial or non-spatial indicator values depending on the use of the spatial character of the original data. Aggregations based on general averages or total values are non-spatial as these are independent of the original spatial configuration of the data. The average size of certain types of interconnected areas (average size of all urban areas) is considered a spatial indicator value as this depends on how the urban areas are connected.

 

3.  Combination

Combination of different spatial data layers is one of the key functions of GIS and it offers a powerful tool to provide an overview on many different data sets in one new integrated representation. By overlaying different data layers it is also possible to create a new data layer instead of merely visualising a result. The overlay operation is thus a typical spatial analysis operation available in any proper GIS. A classic example of this type of analysis is to define the area of overlap of two or more separate data layers indicating, for example, the area where new developments are not permitted following a large set of zoning regulations. Overlays are well suited to compare several data layers in a structured manner. Basically three different comparison options can be distinguished (Muehrcke, 1973):

 

1. a data set with another data set that represents the truth as is common in, for example, validation exercises;

2. a data set with another data set, for example, to compare the development over time of a specific phenomenon or to study spatial patterns of related spatial phenomena;

3. a data set with a theoretical data set, to test assumed relations.

 

4.      Valuation

Valuation is an appropriate tool to help interpret the results of spatial analysis operations. By applying a normative and consequently subjective classification operation to analysis outcomes their value is better understood. In essence, this is not a spatial analysis method since it, generally, only applies to non-spatial valuation functions. The main aim of valuation is to make the content of related data sets comparable. It is a common tool in environmental impact assessments and decision support systems that aim to provide clear, easily interpreted outcomes to policy makers and stakeholders. Simple valuation exercises result in a limited number of categories distinguishing, for example, positive, negative or neutral outcomes in relation to a reference value. Monetary valuation that is common in, for example, cost benefit analyses is an example of a more elaborate valuation method.

 

5.        Proximity analysis

A classic type of GIS-assisted analysis deals with the assessment of distance, normally expressed as proximity. Buffer analyses that create zones of influence (e.g. noise contours around roads) surrounding different types of shapes are typical examples of proximity analysis. Plain distance maps that describe the Euclidian or other type of distance to a specified object (e.g. railroad, city centre) offer useful input to various forms of spatial statistical analysis that, for example, aim to explain specific spatial phenomena.

 

6.       Simulation

By describing the relevant relations of a system it is possible to simulate its future state. A common form of simulation (or modelling) is applied in impact assessments that describe the possible consequences of a specific event or policy. Such assessments follow predefined cause-effect relations that are made operational by one or more of the spatial data analysis methods described before. More complex examples of simulation are offered by the models that simulate, for example, the groundwater or land-use system.

 

In addition to these Koomen’s approaches is the Hot Spots Analysis as described below.

 

7.    Hot Spots Analysis

This tool identifies statistically significant spatial clusters of high values (hot spots) and low values (cold spots). It automatically aggregates incident data, identifies an appropriate scale of analysis, and corrects for both multiple testing and spatial dependence. This tool interrogates your data in order to determine settings that will produce optimal hot spot analysis results. If you want full control over these settings, use the Hot Spot Analysis tool instead. (http://desktop.arcgis.com/en/arcmap/10.3/tools/spatial-statistics-toolbox/optimized-hot-spot-analysis.)

Source:

Koomen, Eric Spatial Analysis in Support of Physical Planning (2003)