Hawaii Extended Forecast Discussion
NWS Weather Prediction Center College Park MD
259 AM EST Mon Dec 21 2020

Valid 00Z Tue Dec 22 2020 - 00Z Tue Dec 29 2020

Expect trade flow to weaken during the first half of the week,
with departure of surface ridging into the East Pacific and a
front approaching from the northwest leading to light and variable
winds by midweek.  This should transition shower focus from
windward locations to sea/land breeze boundaries.  Rainfall will
likely be in the lighter half of the spectrum given moderate
moisture levels and an east-west upper ridge building overhead. 
Arrival of the front late this week will bring an increase of
rainfall.  Then strong high pressure passing by to the north will
promote stronger trades that will be northeasterly right behind
the front and then turn more easterly and eventually
east-southeasterly from the weekend into early next week as the
high continues along to the east.  Lingering frontal moisture may
enhance windward-focused rainfall into the weekend.

Full comparison of guidance reveals at least a couple points of
uncertainty in the forecast.  The first is late in the week when
the 00Z GFS brings higher precipitable water values into the Big
Island and vicinity from the southeast versus other guidance.  An
average of the 00Z GEFS mean/00Z ECMWF appears more reasonable
while the 12Z ECMWF mean was on the southern side of the spread
with this moisture.  Then by the weekend there is lingering spread
in some guidance for upper flow specifics which could have a
significant influence on moisture/rainfall.  At first glance, the
recent GFS trend to pull energy from the southern fringe of the
westerlies farther east away from the islands would serve to
increase confidence in the previously favored moderate
precipitation forecast.  However the 00Z UKMET actually drops a
compact upper low to just northwest of the main islands by the end
of its run on Sat and 00Z GEFS-CMC/12Z ECMWF ensemble members
still show a small percentage with a trough far enough westward to
have an effect on the state.  For a specific forecast, would
recommend an average among the GFS/ECMWF and their means which
become fairly comparable by late in the period--but with the
understanding that a lower-probability alternate scenario still
exists.

Rausch