Spatial Statistics
The concepts involved in spatial statistics require a basic knowledge of GIS and statistics. It may be worth visiting our GIS information page first.
What is Spatial Statistics?
Spatial statistics is all about analysing data that has a spatial (location) characteristic to it. This type of analysis looks for patterns or correlation in recorded observations of some process that occurs across a space. There are two main forms of these data:
-
Point Processes - This is data comprises the location of particular point events. Typical analysis would concentrate on analysing the patterns produced by the points. More specifically, analysis would try to determine if there are random, regular or clustered patterns evident. An example of this could be the study of the positions of a specific type of tree in a forest.
-
Geostatistical or Continuous Processes - This data is collected at point locations but represents a continous measurement. The method can be used to statistically predict a value in a new location based on values around it. An example of this is a study is a prediction of a yearly rainfall value based on the rainfall figures available around town. There is also another type of data known as lattice data which is basically continuous data in grid form. This data can be analysed the same way as geostatistical data using the centroids of the cells.
SPAN and Spatial Statistics
Some Spatial Statistics support is available through SPAN's representative Simon McDonald at Albury/Thurgoona CSU. Enquiries about courses available at CSU in Spatial Statistics should be directed to Kylie Thomas (lecturer in statistics).
How do I get started?
Training, software and support are available from SPAN. The software used for spatial statistics is S-Plus and is available for all CSU researchers.
SS software can be borrowed from SPAN for you to install on your local computer or alternatively you can utilise SPAN's Research Computer (Tesla).