Interpolation is the process of taking a series of discrete points, and joining them up to form a line, or a surface.
The interpolation can be a straight line interpolation, where a line is drawn straight between the points (like a dot-to-dot picture), or the interpolation can be performed mathematically, according to a predefined formula.
Straight line interpolation is a very clumsy method, and misses out much of the more subtle information in a dataset. Mathematical interpolation exploits the full potential of the data, but if used carelessly, can create false surfaces.
How does interpolation work?
Interpolation works by creating a series of regularly spaced points (nodes) around the existing points, and drawing the surface on the regular grid. The values of the nodes are either calculated by their proximity to a known point, or through a statistical equation.
The interpolation methods used in the Great Yarmouth Archaeological Map were kriging and nearest neighbour. Kriging uses statistics to look for trends in the data, and weights the surface accordingly, while nearest neighbour remains as true to the original data as possible. All surfaces produced in the Great Yarmouth Archaeological Map were produced using both methods, and compared, to make sure that we extract the maximum information from the data without creating features mathematically.