If you've activated the Confirm Dimensions option in the gridding options window, the program will display the following confirmation window prior to creating the grid model.
Here you can adjust any of the coordinate boundary or spacing settings. The defaults will be pulled from the Grid & Model Dimensions you've established, or they will be computed by scanning the data, as you defined in the Gridding Options.
The appropriate node density to use will vary from dataset to dataset. Increasing the node density will generally create a more detailed model, but if you create too many nodes you'll get artifacts and the gridding time will increase. It's helpful to know roughly how close your control points are. If, for example, the data being gridded is spaced an average of 100 feet apart, then setting the grid nodes at a 500-foot spacing won't be fine enough to show the detail of your observed data. By contrast, setting the grid nodes at a 2-foot spacing will introduce all sorts of spurious noise.
Success in gridding depends in large part on your subjective determination of how good a map is created. You might start with a node density that delivers an average of 2 or 3 nodes between control points, and then take a minute to evaluate the resulting map. You can re-grid and increase node density if detail is not adequate; you can re-grid and decrease node density if bull's-eyes and other artifacts are present.
