GIS Files 3: Adding real-world information

3.5: Structured GIS data is the key (4)

Spaghetti data

Large-scale topographic data is normally captured by the survey of physical objects on the ground. Before the advent of GIS the need for structured data wasn't understood, so data was captured quite randomly. Often a line along a connected series of objects was captured as a single feature, instead of a separate line for each object or house. This is especially so in housing estates, where long fences or walls surrounding a series of properties were digitised in one go. When the data is drawn on screen it looks fine; you can see a row of houses, each with its own front and back garden. However, a look at the attribute table shows that the data does not consistently correspond to discrete real-world objects like the wall of a single house or the house itself.

This kind of data is often referred to as spaghetti data (as unstructured and random as spaghetti thrown onto a plate). The lines are a tangled mixture that can be interpreted visually as a large-scale map, but they do not explicitly store each separate datum that the GIS could potentially analyse. Put simply, even though it looks good, it doesn't make much sense and it is not very useful for analysis. There is much more need to analyse information about houses, properties and land parcels than there is about the linear features which enshrine them; only a creosote manufacturer is likely to be interested in analysing the attributes of fences.

Now lets look at how improving this data structure provides significant benefits to GIS users...

< 3.5: Structured GIS data is the key (3) | 3.5: Structured GIS data is the key (5) >