In cross-section, if we have observed values at 1000 and 2000 and we need to make an estimate for intermediate locations, a simple proportion can be used. The gradient in this case can be thought of as a type of transfer function:
z=f(x)
| Most interpolation procedures assume that a smooth
or continuous gradient exists between control points. This gradient may
be linear or non-linear.
In cross-section, if we have observed values at 1000 and 2000 and we need to make an estimate for intermediate locations, a simple proportion can be used. The gradient in this case can be thought of as a type of transfer function: z=f(x) |
|
| Of course, if this assumption changes and a non-linear gradient is used, the estimated value of z will change as well. |
|
There are two types of interpolation modules that assume continuous gradients--inverse distance weighted interpolation and kriging. These two methods are explained through the links that follow.