How do I know which Grid interpolation method to use?
Users who are unsure of which interpolation method to use should consult the Grid Analysis help topics in the GeoMedia Help: Start > Hexagon GeoMedia Desktop > User Documentation > > Help.
Users may also wish to consult a spatial statistician to engage the services of Hexagon to assist in developing the best quality data set for interpolation and to maximize the interpolation results.
The following are "rules of thumb" for Grid Analysis users who are already experienced with raster GIS and who have a good knowledge of the limitations of their data set:
IDW (Inverse Distance Weighting))
Use for high density and/or regularly spaced data points, such as elevation data, temperature data, rainfall data, or data from any continuously varying surface.
Use to fill in small gaps in satellite data, aerial photo mosaics, or DEM mosaics.
Use when there is a low degree of confidence in the exactness of the original data.
Use for very sparse data (<1000 data points), such as core samples, well logs, ground water measurements, seismic recordings, clustered data, and random points.
Use when the distance at which data points become independent is known.
Use for spatially correlated data or data for which there is a trend (ground water plumes, ore or other geostratigraphic bodies, seismic measurements, etc.).
Use when there is a high degree of confidence in the exactness of the original data.
Use to predict (interpolate or extrapolate) the distribution or structure of a phenomenon.
Use to interpolate continuous surfaces from well distributed, sparse to very sparse, data. Data should be thinned if it is too dense or if there are a lot of redundant values, such as along rasterized contours.
Use to model a continuous surface from sparse data based on general trends (that is, relatively smooth results) in the surface, rather than exactly predicting specific points.