Exact and Smoothing Interpolators
Gridding methods included with Surfer are divided into two general categories: exact interpolators and smoothing interpolators. Some exact interpolators can incorporate a smoothing factor that causes them to become smoothing interpolators.
Exact Interpolators
Exact interpolators honor data points exactly when the point coincides with the grid node being interpolated. In other words, a coincident point carries a weight of essentially 1.0 and all other data points carry a weight of essentially zero. Even when using exact interpolators, it is possible that the grid file does not honor specific data points if the data points do not exactly coincide with the grid nodes. Refer to Weighted Averaging for more information on weights assigned during interpolation.
To increase the likelihood that your data are honored, you can increase the number of grid nodes in the X and Y direction. This increases the chance that grid nodes coincide with data points, thereby increasing the chance that the data values are applied directly to the grid file.
The following methods are exact interpolators:

Inverse Distance to a Power when you do not specify a smoothing factor

Kriging and Cokriging when you do not specify a nugget effect

Nearest Neighbor under all circumstances

Radial Basis Function when you do not specify an R2 value

Modified Shepard's Method when you do not specify a smoothing factor

Triangulation with Linear Interpolation under all circumstances

Natural Neighbor under all circumstances
Smoothing Interpolators
Smoothing interpolators or smoothing factors can be employed during gridding when you do not have strict confidence in the repeatability of your data measurements. This type of interpolation reduces the effects of smallscale variability between neighboring data points. Smoothing interpolators do not assign weights of 1.0 to any single point, even when a point is exactly coincident with the grid node. When smoothing is used, weighting factors are assigned so the map is smoother. In the extreme case, all data points are given equal weight and the surface becomes a horizontal plane at the average for all data in the data file.
The following methods are smoothing interpolators:

Inverse Distance to a Power when you specify a smoothing factor

Kriging when you specify an error nugget effect

Polynomial Regression under all circumstances

Radial Basis Function when you specify an R2 value

Modified Shepard's Method when you specify a smoothing factor

Minimum Curvature under all circumstances

Local Polynomial under all circumstances

Moving Average under all circumstances
See Also
Introduction to Gridding Methods
General Gridding Recommendation
Choosing Methods Based on the Number of XYZ Data Points