Interesting take from Mount Holly on the potential, and the current differences in the models.
This sets the stage for a rather complex scenario for the weekend.
Upstream, highly amplified ridging will occur in the western U.S.,
with models generally depicting a northern-stream system lifting well
north into Canada late in the week while a southern-stream component
struggles to move through the ridge in the Southwest. Meanwhile, a
downstream northern perturbation will be digging into the eastern
North America trough. However, models are at odds as to how strong
this perturbation is and where this digging will occur. The GFS
occasionally hints at this process occurring in the Great Lakes
region Friday night and Saturday, which would be sufficiently far
west to allow for phasing with the advancing southern-stream
perturbation before it moves offshore. The ECMWF is considerably
farther east (generally east of the Great Lakes), so the phasing
with the southern-stream system generally occurs after moving
offshore.
The ECMWF has been reasonably consistent the past few cycles,
whereas the GFS has been quite variable. However, given the
complexities of the setup, the volatility of the GFS is
statistically meaningful. The ensemble means more closely match the
ECMWF than the GFS, but the deterministic GFS is also a more extreme
solution (which would ultimately make it more likely to deviate from
the ensemble mean). This makes me question if the ECMWF is a
substantially more likely solution than the GFS. Given the phasing
the ECMWF eventually develops, my suspicion is the damped ensemble
solutions are more a result of model variability than higher
probability.