A segment of the afd from mount holly.
Focus with the medium-range forecast remains on the evolution
of a southern-stream system traversing the southeastern U.S.
this weekend before moving offshore early next week.
Operational models are beginning to converge on a solution for
the strong surface low that develops near the Gulf Coast this
weekend. The 00z CMC has trended farther south toward the
GFS/ECMWF solutions, though it does remain on the northern
fringe of the guidance envelope. However, noticeable variability
remains with the northern-stream perturbations that will
interact with the southern- stream perturbation this weekend,
particularly the predecessor trough in eastern Canada late this
weekend. Of particular note, the 00z ECMWF is most progressive
with this trough, which acts to slow the surface low near the
Carolina coast Sunday night but also prevents much poleward
progress. The CMC/GFS are noticeably slower, which allows the
low to lift farther northward but also provides more of a kick
to the southern-stream perturbation on Monday. The GFS has
trended deeper and somewhat slower with the northern-stream
system in the past 24 hours, which suggests the ECMWF solution
may be a lower probability one.
Meanwhile, a second northern-stream perturbation digs south-
southeastward into the central plains on Sunday night, which
adds substantial complexity to the interactions of the southern-
stream surface low with the northern stream. Volatility with
this perturbation is very high run to run and model to model.
Notably, the GFS has trended slower and a little weaker with
this second northern-stream perturbation, which may be
suggestive of a slower push to the southern-stream system.
Operational guidance seems to be converging on a solution in
which a secondary low develops on the coast Monday night in
vicinity of the first low 24 hours prior. The model tendencies
are similar to the first low (CMC farthest north; ECMWF to the
south), but the GFS is slower with this second low (an obvious
result of the slower trend to the central U.S. perturbation).
What does all of this mean? My conclusion is that the models
converging on a solution may be a tempting lure, but I am not
biting quite yet. The perturbations in play here remain offshore
(prior to the 00z simulations, anyway), so I suspect some
decent initialization errors are contributing to model
volatility. Perhaps as importantly, the complex interactions of
these perturbations are extremely sensitive to
positioning/orientation/strength of the individual phenomena,
which is a telltale sign to use caution in the use of
deterministic solutions as a forecast starting point. I used a
healthy blend of continuity, ensemble guidance, and multi-model
consensus as a starting point, only subtly modifying the
forecast to account for the increased southern-track consensus.
Finally, I should note that we are still in the time window in
which models tend to be too far south and too progressive with
the southern- stream portion of these systems, so I maintained a
broad-brush approach to distribution of PoPs (and timing) as a
result of the remaining large uncertainty.