dont believe the sea ice back pressure in Petermann Fjord play a major role, on the other hand I think melt water do, it is cutting through the glacier ice in many places
It's important here to keep in mind the 7 distinct components of the Petermann system: the glacier proper, its attached floating ice shelf, the ice from the four east side tributaries, ice along the western wall, and floating melange in front of it all.
The glacier is pushing the whole system forward all year long at about 3 m per day (with mild seasonal variations). The ice shelf moves at this speed too as it is rigidly connected and there is little internal dynamics (thinning or thickening by stretching or compression).
The tributary ice is not quite keeping up and an area of rotating blocks (ball bearings) takes up the difference though there is shearing too. At its edge, this ice is essentially frozen to the eastern wall. The western wall has no tributary ice but is deformed (wrinkled longitudinally) by drag between the rock walls and the moving central ice.
The melange goes from nothing most summers to a continuous extension of the floating Nares Strait ice. If that ice is stuck, which it often is, the melange has nowhere to go even as the ice shelf keeps pushing on its rear. This compression should result in ridging and rafting by early spring but we may not have enough resolution to see it.
If you accept Newton's 3rd law, the rigidly frozen melange exerts a force to the south, equal and opposite to the force exerted on it by the ice sheet. The magnitude of this force varies seasonally.
After the irresistible force meets an immovable object business gets sorted out, this melange buttressing force still may effectively resist the opening and widening of rifts (especially those near the calving front) because these represent parts of the ice sheet moving faster than parts somewhat hung up on the tributary sides.
Rift origination in Petermann Glacier has nothing to do with melt pond draining or hydrofracturing of melt water extensional crevasses. Rifts always originate on the sides on flat ice. On the west, that means off the rock walls; on the east, on the tributary ice boundary.
No rift has ever originated in the central 2/3 of the ice sheet, hinge line or anywhere else during the satellite, aerial photographic or explorer-map eras as far as I can tell.
Rift development and tip propagation initially precede obliquely backwards towards the center for the reason given. Exposed sea water in the rift can freeze over without really healing the rift however.
In 2010 and 2012, I recall an overwintering rift developed a rift within the rift that extended through this briny thin ice which eventually came to fail fairly abruptly across the entire width as leveraged motion became prodigious. Landsat did not provide high resolution imagery in these years.
Water from any active meltwater channels or intercepted melt lakes could accrue in the rift bottom but is not material since that ice is already so thin/weak/warm/briny compared to the ~200 m ice that must be broken to extend the rift. The water has little access to the growing rift tip and lacks sufficient volume, depth and time for tip melt or hydrofracture. It will simply find a melt channel through the brine ice and exit to the fjord water below.
Rifts do originate and extend mostly in the summer season, at a time when the upstream glacier is moving at its fastest, meaning the side drag is at its highest. The ice up to the grounding line is at interior ice sheet temperatures but over the ~50 years it spends in the fjord adjust to ocean temperatures from below and solar and air heat inputs from above, as dirichlet boundary conditions of Fourier's heat equation. The issue here is not ductile vs brittle but rather tensile failure.
None of the 2015 rifts developed liquid water this year. I don't recall any liquid water in any of the 0.5 m resolution WorldView rifts for August 2014 either. Rifting is mainly just a carpenter snagging his shirt on a nail.
Now Rignot one time suggested -- because of maximal erosive cavitation just past the grounding line -- that the ice sheet could calve off everything in one giant event due to tidal flexure. That hasn't happened to date but the big news is that Andreas M is setting out to actually measure tides within the fjord, notably under the ice shelf, using differential GPS.
It is possible that large tides and resonances over bedrock sills have some responsibility for rift origination hot spots or tip propagation, in a manner reminiscent of a very slow swell (here seiche or soliton) breaking up pack ice. However that would crack the ice straight across the channel which is not the reverse herringbone pattern of the contemporary rift set.