Increasing magnetic field and attractive strength of Gargrave round electromagnetic uncouplers.

Probably not easily.  All of the makers uncouplers seem to have a similar design.

These electromagnets are capable of drawing 4 to 5 amps easily.  I recently worked on a time saver module that had 5 of them.  We actually powered all the accessories, mostly DZ 1000 switch motors, and the electro-un-couplers off one side of a Z-4000. When you activated an un-coupler, it instantly draws 5 amps and fluctuates the voltage down 3 to 4 volts. A fair amount of power, maybe not available from a small accessory transformer.

I experimented with replacing one with an under the track electromagnet, it's astounding the amount of power you need if the electromagnet is even 1/8" farther away.

I have not seen that particular uncoupler but as John says, they seem to have similar designs - in which case it's all about increasing the coil current and/or the number of winding turns. That is, the strength is proportional to the "Amp-turns" product. 

Obviously, the easiest thing to do is to increase the applied voltage to increase the current. But consider the power handling capability of the electromagnet as a sticky activation switch or whatever will soon melt something.  There have been threads musing about why various vendors do/don't install thermal cut-off devices in uncouplers.  Re-winding the coil with more turns is a little tricky.  If you simply double the number of turns with the same gauge magnet wire, you double the wire length (resistance) which halves the coil current and you end up with no gain since the Amp-turns product remains the same. So if you have space, you could use thicker gauge wire to keep the winding resistance down.

Depending on what you're trying to do, you could try a bias magnet. There would be a constant field and when the coil turns on, it would add to that. With an AC driven coil the polarity alternates so the idea is you're increasing the strength on alternating halves of the cycle which might do the trick. 

Yet another idea would be a circuit generating a short pulse to smack the coil at a higher voltage using stored energy and/or voltage step-up - sort of like capacitive discharge switch activation. 

Well, if you put AC on your "bias magnet", you'll demagnetize it.

One issue with these decouplers is that the magnetic field strength drops as to the square of the distance, so any additional space quickly makes it virtually impossible to activate the couplers.

I suspect the way to do this is custom make a uncoupling track section with a honkin' electromagnet.

I agree. The brute-force approach to uncoupling hasn't changed much in decades.

And another idea: I don't know what kind of coupler the OP has but if it's just a few cars/couplers causing a problem, the "bias magnet" can reside on the coupler side. Nd discs 1/32" (shown below) or 1/16" thick are available for less than a quarter. Of course I have no idea if "uncoupling" is the actual application for which the OP wants more strength!

In our modular club, we were trying to come up with an under the track uncoupler so it would not be so visible (or not visible at all).  I tinkered around with a bunch of options and could never get enough strength to uncouple the coupler with any kind of spacing.  It would take a serious magnet!  I have some neodymium magnets that will pick up 100 pounds, 1/4" away from the coupler, they would not activate it!  I've about given up the fight.

Of course too consider that maximum strength also ocurs when there is a complete  magnetic circuit,  The Lionel uncouple magnets do not complete their magnetic circuit through the rails then through the wheels, truck bolster etc assuming an all metal wheel set.  Now wheel sets are plastic that effectively breaks this circuit thus requiring a large amp-turn to compensate for weak magnetic field.  Don't recall if the Lionel/MTH uncoupler magnets have a copper piece or sleeve in them like used in AC relays to attract the armature.

Correct, the trick is to make that happen!   The rails aren't an issue with Atlas track, which is what we are using, they're non-magnetic.  However, I suspect the rest of the truck was the issue.  I know that it centered the car nicely on the track when the magnet was there!