Spatial Interpolation-Applicability of interpolation methods

Spatial deterministic interpolation is based on the similarity or smoothness of the area, and the surface is created by known samples.

1. IDW

Similarity similarity principle, inverse distance weighting.

The sample distribution should be as even as possible, and be filled with the whole interpolation area.

For samples with irregular distribution, the samples used during interpolation are often unevenly distributed in different surrounding directions, which has different effects on the interpolation results in each direction, the accuracy is also reduced.

2. Global Polynomial Interpolation

A polynomial is used to calculate the predicted value, that is, a plane or a surface is used for full-area feature fitting. The non-precise interpolation method does not completely overlap with the actual sample.

It is required that the properties of the sample simulation be gentle on the surface of the study area. It can also be used to test the influence of long-term changes on global trends (that is, Trend Surface Analysis. Select three-way fitting.

3. Local Polynomial Interpolation

There are multiple polynomials, each of which is in a specific overlapping adjacent area. Not precise interpolation.

It can be used when a small variation of a smooth surface needs to be established and the variable is determined. Describes short-range variations contained in a dataset. It is more used to explain local variations.

4. Radial Basis Functions

Five types of functions are provided: plane spline function, tension spline function, rule spline function, high-order curved surface function, and anti-High-Order Curved Surface spline function.

Suitable for surfaces with gentle surface changes.

It is not applicable when the surface changes greatly within a short horizontal distance, the accuracy of the sample data cannot be determined, or the sample data has great uncertainty.