Alaska Extended Forecast Discussion
NWS Weather Prediction Center College Park MD
218 PM EST Mon Mar 05 2018

Valid 12Z Fri Mar 09 2018 - 12Z Tue Mar 13 2018

A deep upper low tracking across the northern Bering Sea in the
short range time frame will likely reach the southwestern mainland
by the start of the extended period and serve as the core of an
amplifying eastern Pacific trough aloft.  After some shifting in
the exact path of the upper low over recent days, the consensus
track has held up fairly well over the past 24 hours.  This leaves
the specifics of leading Gulf of Alaska dynamics as the more
prominent uncertainty with fairly small scale differences (likely
not to be resolved until the short range) affecting the
track/timing of potentially deep northern Gulf low pressure.  As
was the case yesterday the 00z ECMWF/UKMET runs are somewhat east
of other models as well as the ensemble means at 12z Fri.  The
eastern solutions are within the full ensemble spread but the
means have been very consistent thus far in showing an early Fri
position just east of the Kenai Peninsula.  The means are still
trending a little deeper with the low to provide support for
decent inclusion of the deeper operational guidance in the
forecast, while lack of confidence in exact details aloft seem to
recommend an intermediate track.  There is reasonable
agreement/continuity that Gulf dynamics should rapidly progress
into Canada while the lingering upper low should help to maintain
an identifiable surface low that drifts southwestward off the east
side of Kodiak Island or vicinity.  By mid-late period guidance is
fairly consistent in showing the western Canada ridge aloft
building into the mainland while a weakness persists over the
western mainland and extending into the Arctic.

The one significant change in the forecast from yesterday involves
low pressure which a majority of guidance now shows tracking
across the Bering Sea around day 5 Sat and then feeding into the
eastern Pacific mean trough.  Before the 12z/04 cycle there were
only modest signals of this system, mainly from some GFS runs and
a few ensemble members.  12z/04 models and ensembles almost
unanimously latched onto this system with continuity in principle
maintained through the 00z/06z cycles and in the 12z GFS. 
Meaningful strength/track/timing differences still exist though. 
Given the history of this system so far, a model compromise
provides a reasonable starting point with somewhat better
definition than the means but accounting for the fairly high
uncertainty in details.

By the latter half of the period over the eastern Pacific
solutions diverge to a greater extent for this Bering system, with
the GFS runs tracing a wider southeastward arc than most other
guidance.  In addition operational model runs are signaling
potential for a leading system that may track northward into the
Gulf of Alaska on the eastern side of the mean trough aloft.  The
ECMWF has shown some similarities with this leading system over
multiple runs while the more variable GFS has adjusted toward an
ECMWF-like evolution in the 12z run.  Ensembles are still not
close to being sufficiently clustered to suggest a system separate
from the one approaching from the Bering and other models still
vary considerably.  Confidence was still sufficiently low in any
deterministic scenario to recommend the ensemble means' broader
area of northeastern Pacific low pressure as the best way to
anchor the forecast while awaiting better clustering of solutions.

Most guidance shows low pressure over the extreme western Pacific
by late in the period with a leading frontal system extending into
the Aleutians/Bering Sea.  Recent GFS runs (especially the 06z
cycle) have been on the fast edge of guidance with embedded
waviness over the Bering.  Aside from some 00z CMC ensembles there
is better support for slower progression such as depicted in the
GEFS/ECMWF means.

Based on considerations early in the period, the forecast started
with input from the 00z-12z GFS/00z ECMWF/00z CMC for days 4-5
Fri-Sat.  Then day 6 Sun served as the model/mean transition
toward exclusive use of the 06z GEFS/00z ECMWF means for the rest
of the forecast through day 8 Tue.

Rausch