Hawaii Extended Forecast Discussion
NWS Weather Prediction Center College Park MD
359 AM EDT Thu Apr 11 2024

Valid 00Z Fri Apr 12 2024 - 00Z Fri Apr 19 2024

...Heavy rainfall threat over the western islands late this week...

Today's guidance is consistent with the deep upper low and surface
system forecast to be positioned northwest of the state late this
week and then open up as it lifts northeastward this weekend.
Along the front/convergence zone extending to the south of the
surface low, abundant deep moisture will lift northward to bring
precipitable water values to or beyond three standard deviations
above normal over the western half of the state for a time.  After
multiple days of showing moderate though locally important
east-west spread for the associated axis of heavy rainfall,
clustering of latest solutions would suggest increasing confidence
in a heavy rainfall threat for Kauai and vicinity during
Thursday-Friday.  Most guidance shows a break around Saturday as
the front drifts back to the west a bit, and then a return of
enhanced rainfall to the western and possibly central parts of the
main islands as progression of the remaining upper trough pushes
the front eastward. The front may stall upon reaching either side
of the Big Island after early next week but rainfall should not be
as heavy given less extreme precipitable water values.

From Sunday onward there is still a fair amount of spread for how
upstream energy may reinforce the upper trough over/north of the
state.  The 00Z GFS/UKMET details yield an upper low tracking
farther south than most other dynamical guidance around
Monday-Tuesday, with the 00Z GFS holding onto a deeper upper
trough over the state through next Thursday.  Remaining guidance
that holds the upper low farther north would correspond to low
level winds turning more northerly behind the front crossing the
state, versus staying more westerly in the GFS.  The majority of
dynamical guidance would suggest tilting away from the GFS, but
some of the latest ECMWF-initialized machine learning models offer
potential for some of the GFS ideas aloft.

Rausch