3.1 INTRODUCTION

This chapter examines the issues associated with building historical GIS databases. The process of GIS data capture, as it is known, is slow, expensive and frequently tedious. It is almost always the most expensive stage of any GIS project. Bernhardsen (1999) estimates that this stage usually accounts for 60 to 80 per cent of the total cost; other authors present similar figures. The data-capture phase of the project is also time-consuming, and during it there is little obvious short-term reward. This means that, before attempting to build a GIS database, the researcher needs to think carefully about whether the rewards of the completed GIS justify the costs of building it, and about the project management implications of creating the GIS, the complexity of which is frequently underestimated by academics. It is also important to remember that once a GIS database has been built, it will often have uses that go far beyond the original purpose that it was created for. For this reason, it is also important that the researcher considers issues such as documentation, metadata and long-term preservation to ensure that the data can be used appropriately by others.

As discussed in Chapter 2, GIS data are structured using either a vector or a raster data model. Vector data use points, lines and polygons to represent discrete features on the Earth's surface, while raster data represents a continuous surface using pixels. The two types of data are captured in very different ways.