A cost surface, or cost grid, is a raster grid in which the value in each cell is the cost that a particular activity or object would be in that cell. It can also be an indexed value based on costliness. Costs could be measured monetarily or in other ways such as amount of time. A cost surface includes the cost of reaching certain cells from one or more source cells.
Cost surfaces can be useful to do things such as travel cost analysis, cost path analysis, and cost distance analysis. In these analyses, cost surfaces are created for each of the factors in the overall cost of what needs to be accomplished. Varying factors could influence the cost, depending on the analysis. These cost surfaces could include slope, land ownership, environmental effects, and land value. Each of these cost surfaces would then be added together to form an total cost.
Assigning cost values
If the cost surface is to be accurate as well as the final model derived from it, the cost surface should use a ratio scale in which the values are meaningful and have an absolute zero. In the case of a slope raster for example, a cell with a value of zero means that there is no slope, so this scale has a meaningful zero. Therefore a cell with a value of 20 would twice as steep as a slope with a value of 10. If the values in other raster or vector layers that will compose the cost surface are nominal, they must also be converted to a ratio scale. It is important that those future values are realistic and reflect meaningful costs. For example cutting down vegetation for a new development might result cheaper for brush area than for forests. Therefore forests would be assigned a high value and the brush area a low value. If the exact ratio in terms of money or other measurements is unknown, research should be performed. If nothing is found, then approximations should be use, however the result will be a model that is also an approximation.It is often the case that many different layer types are combined to create the surface cost. If the values in those layers are in different units of measure such as dollars and elevation, then new values should be assigned using a common ratio scale, such as one that runs from one to ten. It is important however, that that the new values remain constant between different layers.in the case of elevation and money, a value of seven must be equally comparable in both data layers.
- Bolstad, P. (2008). Gis fundamentals, a first text on geographic information systems. St. Paul: Eider Pr.