Update for May 6 2018:
Yesterday I discovered something I find interesting about these 3 carriers which are about 10 Hz apart on 6993 KHz.
I was watching the three signals on the waterfall and found that two of the three carriers showed an equal and stable signal strength over perhaps 3 minutes of viewing.
However, the third carrier, in this case the lowest frequency carrier, showed obvious fading from moment to moment and at one point faded out. It soon returned to perhaps 2/3 signal strength as compared to the other two, but the signal fading continued. I recall seeing this selective fading of one carrier in the past but paid no attention to it. Keep in mind these 3 carriers are only 10 Hertz apart.
Atmospheric fading of constant carriers is easily seen on the waterfall, but two of the three carriers seemed pretty much in lockstep with signal strength: only the third showed obvious fading.
I submit these possibilities:
The signals are emmitted from one source but with different antenna patterns
The signals are of different power
The signals are eminating from different locations
A combination of the possibilites above.
The 6993 KHz signal is clearly readable during daylight hours her in Florida and with the current lack of decent HF propagation, I venture to guess the antenna is in the Southeast USA, Caribbean, Central America, or perhaps less likely South America north of the equator.
The source and function of these carriers is still a mystery. With nearly 5 years of casual monitoring of these signals, not much has come to light about their reason for being.
The 4993 KHz sister-signal is still readable here, but only at night.
Dag