Interaction of Scatter and Tilt

This is an example showing the interaction of structural tilt and "scatter". Dipmeter data will have various amounts of scatter, and this scatter can appreciably affect interpretations. It is easy to see how the interaction of structural tilt with natural scatter will tend to mask significant features.   Remember that in real dip logs that scatter may be due to measurement or calculation uncertainties, but it may also be caused by smaller features, usually stratigraphic in nature.  That is why the  word "scatter" is used here instead of the word "noise".

There are two models in this applet, a fault and an unconformity which can be toggled with the "fault" and "uncon." buttons.   Before adding scatter to the models, it is instructive to vary the structural tilt and see how the models respond.

To add random scatter to a model, enter the number of degrees of scatter that you would like and then push the "add scatter" button. (The scatter is added by rotating each dip using a randomly calculated dip and azimuth in which the dip is always lower that the input scatter number.)   Click the polar plot on the right to apply different structural tilts to the log. 

As you are changing the scatter on the fault model, occasionally go back to tilt of 0,0 and you will see that the rotated (untilted) dips will reveal features in spite of huge scatter of +-10 degrees.

Another interesting interaction of tilt and scatter is how the randomly scattered dips above and below the fault pattern sometimes attain a visual coherence when rotated.  For example, with scatter of 5 degrees and tilt of 10 degrees, the random dips with no character will suddenly all point in the same direction even though the dips may vary somewhat.  On the other hand, when the tilt is zero, the azimuths point around the compass and they appear chaotic.  Small wonder that many service companies seem reluctant to use rotation.   It makes their products look bad! 

Try clicking the "add scatter" button several times in a row with a scatter of 5 degrees.  You will note that sometimes there are chance alignments of 3 or 4 dips where you know there should be none.

In the unconformity model, the angular difference above and below the unconformity is less than ten degrees.  Even 5 degrees of scatter can make it very difficult to interpret.