ABSTRACT
In Chapter 2, we saw that using a raster GIS we could store a set of spatial data in the form of a grid of pixels. Each pixel will hold a value which relates to some feature of interest at that point in space. In fact, there are two main types of raster layer which are used in GIS:
Raster imagery. These days we are surrounded by imagery in digital format – modern televisions and computers all have screens in which the picture is composed of pixels and the photographs we all take on our cameras and phones are in a raster format. We are also used to seeing maps and satellite images used on web pages and even on phones. In all these cases, what is stored is a raster grid of colours, coded in some numerical fashion.
Raster data. We can also store data which we wish to analyse in a raster format. Probably the commonest example is the digital terrain model in which each pixel contains a value for the elevation of the land surface. In general, the values in raster data layers are normally one of three possible types:
Binary – a value which indicates the presence or absence of a feature of interest. For example, in a layer representing roads, we might use 1 for pixels that contained part of a road, and 0 for pixels that did not.
Enumeration – a value from some classification. For example, a layer representing soils might contain codes representing the different soil types – 1 for Podsols, 2 for Brown Earths and so on. Since the values are not directly related to the soil type, there would have to be a key of some sort indicating the meaning of each value.
Numerical – an integer or floating point number recording the value of a geographical phenomenon. In the soil example, we might have measurements of soil moisture content. A common example of this kind of raster layer is when the values represent the height of the land surface, in which case the layer is often referred to as a digital terrain model – these will be described in more detail in Chapter 9.
