In theory, straight line between starship and rift, times energy flash, should give you direction plus threeish light year path they could have dropped out of.
MOST starships jump at only a single possible displacement, so any variations in jump flash signature are probably due to the number of parsecs the starship is jumping.
A Scout/Courier jumping one parsec has a jump flash "brightness" of 1 ... and jumping two parsecs has a jump flash "brightness" of 2 ... for example.
In most cases, so long as you know the displacement of the starship making the jump, you can start making educated guesses as to how many parsecs the starship was probably jumping based on the sensor signature of the jump flash.
As soon as you get into
VARIABLE DISPLACEMENTS for starships, though ... that kind of educated guess gets a LOT harder to make, because behavioral spoofing becomes an option (intentionally or not, due to external loading).
This is a lot easier to do with LBB2.81 standard drives, since they work on a straight multiply/divide paradigm.
LBB2.81 Drive-A (TL=9) yields code: 2 @ 100 tons and code: 1 @ 200 tons.
So a 100 ton Scout/Courier with A/A/A drives will produce a jump flash for 2 parsecs that is broadly similar to that of a 200 ton Free Trader with A/A/A drives producing a jump flash for 1 parsec. There will be some subtle differences in details, but you'll need some pretty sensitive sensors to pick up on those differences. For the purposes of our discussion here though, let's just say that A/A/A drives produce the same "brightness magnitude" of jump flash when operating at 100% power (J2 @ 100 tons or J1 @ 200 tons).
If you have details of the starship (class) making the jump (it's a Free Trader, not a Scout/Courier) you can start eliminating possibilities.
"A jump flash of that brightness from a starship of that displacement means they must have jumped
{ insert computational answer } this many parsecs."
And obviously, the longer the jump "the bigger the haystack" of possible destination (hexes) based on the jump point location.
But if you allow external loading of your starships (reducing drive performance) the same way that L-Hyd Drop Tanks and External Demountable Tanks will do ... suddenly ... starship displacements stop being "fixed" and can instead become variable. And what if a starship adds external loading that isn't just fuel tanks (drop or demountable)?
What if you've got a "container ship" that can act as a jump/maneuver tug that can be docked to and towing external loads?
LBB2.81 Drive-H (TL=10) yields code: 3 @ 533 tons, code: 2 @ 800 tons and code: 1 @ 1600 tons.
If the jump drive needs to be run "at full power" in order to jump, the range of the "full power" jump is limited by the total displacement when jumping ... so knowing how much of an external load a starship is encumbered with becomes important.
It's possible to determine the "how much tonnage?" answer from detailed sensor scans ... but depending on the orientation of the sensor target relative to the sensor detectors, this might be a difficult question to answer from sensor readings alone if you can't resolve a visual or a silhouette (turn the target ship a different way, get a different sensor return/silhouette).
It's also possible to make educated guesses about "how much tonnage?" from the maneuver drive performance (1G? 2G? 3G?) ... but even that can potentially be "spoofed" (kinda sorta) by the sensor target starship declining to run their maneuver drive at full power. The ship is currently loaded to 800 tons and is capable of J2/2G, but is maneuvering to a jump point at only 1G ... implying that it is loaded up to 1600 tons and is only capable of J1/1G instead of just 800 tons and capable of J2/2G. By deliberately maneuvering at a lower acceleration to a jump point, the sensor target starship can "camouflage" its true condition and instead operate at J2/1G to mislead an observer into thinking that the starship was only capable of J1/1G. That error in assumption can then be used to lay a false impression of potential destinations based on the brightness of the jump flash.
"A jump flash of that brightness from a starship of up to 1600 tons combined displacement means they must have jumped
{ runs computation to reach the WRONG conclusion with confidence } only 1 parsec."
When in fact, what was actually observed was a starship of up to 800 tons combined displacement jumping 2 parsecs.
Mind you, this is almost certainly going to be a greater level of depth and detail (as an answer to your question) than you probably need or want ... but I'm giving it to you anyway (pro bono) because that greater level of depth and detail can yield a richer and more textured set of possibilities that can highlight the challenges of deducing
where a starship went based on a sensor record of its jump flash. If all you're paying attention to is a single parameter (let's call it "brightness" for simplicity) then you could easily miss a wealth of other important details that would change your conclusions.
Kind of like how nearby dim stars can appear "brighter" to an observer than far away stars that are actually much brighter. Brightness is a function of distance (the old 1/r
2 deal) such that if you don't know how far away a star is, you're going to have a difficult time figuring out what that star's actual brightness really is. You can look at other data than just the brightness (such as redshift and spectroscopic analysis, parallax, etc.) to get more details, but that takes WORK and more specialized sensors to obtain those additional details.
My point being that sensor signatures (like transponder codes) can be "spoofed" or otherwise camouflaged when you've got the added complication of starships capable of
towing external loads. This means that you can't just "know" a particular class of starship is Z many tons (exclusively) and therefore ...
{ insert jumping to the WRONG conclusions with confidence here } 
