Step 9: Divide long fault traces into several fault-trains?

Does your study-area include any really long faults, such as “master” plate-boundary faults?
Do they have intersections with other second-order faults, in triple-junctions (or more complex tectonic knots)?
In either case (especially the second), it is reasonable to expect that slip-rates will vary along the trace of the long fault.
To allow NeoKinema (and NeoKineMap) to determine (and display) this varying slip-rate, you will need to divide the long fault into several “trains.”

The phrase “fault train” was formally defined in Bird [2007] as:
“a contiguous piece of the trace of a fault system along which our knowledge of fault geometry permits the null hypothesis of uniform long-term offset rate.”
In other words, if we have any reasonable doubt about the uniformity of slip rate along a long fault, it is best to divide it.

In the sketch-map below [first seen in Step 3], note that the San Andreas dextral fault was not represented as a single trace:
F0001R  San Andreas dextral fault, CA
but instead is represented by a set of 5 fault trains:
F0001R  San Andreas (Big Bend segment) dextral fault, CA
F0002R  San Andreas (Mojave segment) dextral fault, CA
F0003R  San Andreas (San Gorgonio segment) dextral fault, CA
F0004R  San Andreas (Coachella segment) dextral fault, CA
F0005DR San Andreas (Brawley Seismic Zone segment) spreading/dextral fault, CA
However, it is NOT necessary for the fault-trace numbers FnnnnV of adjacent fault trains to be in order (or even adjacent).

Another important consideration is that many faults have stick-slip (seismogenic) sliding behavior, but certain faults have steady-creeping (aseismic behavior).
In a later step of this process (Step 12), you will be asked to identify the sliding behavior of each of your numbered fault traces.
Therefore, it is better not to combine sections with stick-slip and steady-creeping behavior into a single numbered fault trace.
For example, in central California the “creeping section” of the long San Andreas fault train would always be numbered and digitized separately.

A third possible reason for dividing one fault train into multiple numbered fault traces (sections) would be if there were strong evidence for variations in dip.
In nature, we expect that most faults have dip varying continuously, and that dip discontinuities are fairly rare (and only at triple-junctions). 
However, the NeoKinema approximation for fault dip is that it is constant along any numbered fault trace.

It is actually quite easy to divide a long digitized fault trace into several fault trains, or to divide one fault train into several sections:

1. Note the critical latitudes (or longitudes) at which you want to divide the trace.

2. Use a plain-ASCII text editor (NotePad or EditPad Pro) to open your fault-traces .dig file and find this critical value along the long fault.

3. Duplicate the digitized point closest to the desired critical latitude or longitude.

4. Insert a polyline-terminating record “*** end of polyline ***” in-between these two copies of this common end-point.

5. Insert a new FnnnnV tag (with previously-unused number nnnn) and header line(s) for the new fault train/section after the terminator, giving it a (slightly) different name.

6. Adjust the header for the first polyline, so that it only identifies a single fault train (or fault section).

7. Remember to update your manually-created fault-trace map (drawn with colored pencils on tracing paper).