Angle Deviations (a) Sample Plot of Sample Data. (b) Same Plot After Points With Small Angle Deviations Have Been Removed.
The window in the top left corner of the dialog lists the existing boreholes, with the currently selected borehole being highlighted. Although only one borehole at a time is selected in the dialog, when the OK button is selected, scatter points will be created from all of the boreholes that have sample data. Below the list of boreholes is a button with the current data set name on it. Pressing this button brings up the Select Data Set dialog from which a new data set can be selected. If the Convert all data sets option is chosen, all of the data sets associated with the boreholes will be converted to the new 3D scatter point set. If the Convert only the selected data set option is chosen, the new 3D scatter point set will only have the currently selected data set associated with it.
The main part of the dialog is a plot of the current data set vs. depth for the selected borehole. Along the right side of this plot is a picture of the scatter points that will be created from the current borehole. In the bottom right corner of the dialog, the number of sample points and scatter points for the hole are shown. Because there may be hundreds of sample points per borehole, it is often desirable to try and reduce the data before creating scatter points. Otherwise, a unique scatter point will be created for each sample point.
Filtering is the term used to describe the process of reducing the number of sample data before the data are converted to scatter points. Three filtering methods are provided.
With this option, groups of points are averaged together to produce a new point, with the user specifying the number of points which are averaged together. As the number of points that are averaged increases, the number of scatter points that will be created decreases. A second line on the plot drawn in red shows what the plot looks like with the filtering applied.
With this option, no averaging takes place. Points are simply thrown out to reduce the number of points. The user can specify the number of points to skip at a time. For example, if the number is three, the first three points will be skipped, the fourth point will become a scatter point, and the next three points will be skipped, etc.
The final data reduction method is to Reduce using deviations. One advantage of this method over averaging or skipping is that extreme peaks can be preserved. No averaging takes place; points are skipped if they do not meet the deviation criteria. The deviation criteria can be based on angle deviations or value deviations or both.
If the Remove small angle deviations box is checked, points with small angle deviations will be thrown out. An angle deviation is the change in direction between two vectors formed from three points. Points that are nearly collinear have small angle deviations and points at peaks have large angle deviations. The user can specify the minimum angle deviation in degrees and any points that have angle deviations smaller than the minimum will be thrown out. This is illustrated in the following figure. As the specified minimum allowable angle deviation increases, the number of scatter points decreases.
Angle Deviations (a) Sample Plot of Sample Data. (b) Same Plot After Points
With Small Angle Deviations Have Been Removed.
If points are thrown out based solely on the angle deviation criteria, a lot of points might be preserved that are unnecessary. For example, a series of points might zigzag back and forth on either side of a straight line. It would be desirable to keep only the first and the last points and throw out all of the ones in-between. However, the angle deviation criteria may keep all of the points. If the Remove small value deviations box is selected, the in-between points can be eliminated. A value deviation is the length of the line segment between two points. There are two value deviations associated with each point (except for the first and last points). If both value deviations are smaller than the specified minimum deviation, the point is eliminated. If one or both of the value deviations is longer than the specified deviation, the point is preserved. Part a of the figure below shows a typical plot after points have been eliminated by the angle deviation criteria. Part b shows how this plot might look after points with small value deviations have been eliminated. If both angle and value deviation criteria are being used to filter data, the angle deviations are checked first, and the value deviations are checked second.
Value Deviations (a) Sample Plot of Sample Data. (B) Same Plot After Points
With Small Value Deviations Have Been Removed.
The value deviation criterion is based on a percentage of the maximum value deviation. For example, if the maximum value deviation for the current borehole, current data set happened to be 4 and the user specified a percentage of the maximum value deviation of 25 percent, then all points whose longest value deviation was shorter than 1 would be eliminated. A larger percentage results in fewer scatter points.
When using the Reduce using deviations filtering method, the scatter points are not spaced at regular intervals and may be bunched together. This may cause problems with interpolation depending on the interpolation method used. Also, if a different data set is selected, the scatter point spacing will change. Since the current scatter point spacing only corresponds to the current data set, the Convert only the selected data set option is usually selected when using this filtering method.
Related Links:
Converting Borehole Data to Other Data Types