Hawaii Extended Forecast Discussion
NWS Weather Prediction Center College Park MD
825 AM EDT Mon Mar 19 2018

Valid 00Z Tue Mar 20 2018 - 00Z Tue Mar 27 2018


...A heavy rain/flash flood threat commences early Friday into the
weekend across Hawai`i...


To start the week, the Hawai`ian island chain will remain under
the influence of a mean upper ridge while being accompanied by a
sprawling surface high extending across much of the central
Pacific. This will ensure general east-northeasterly flow which
should turn more easterly as the parent anticyclone gradually
shifts toward the east. Such winds could pick up in intensity at
times given the modest pressure gradient setting up with speeds
likely in the 20 to 25 knot range. This regime will maintain a
fairly healthy plume of moisture as forecast precipitable water
values range from 1.50 to 1.75 inches for much of the week. Widely
scattered showers are possible each day although these should pick
up in coverage and intensity by later in the week.

By the Thursday/Friday timeframe, a low-amplitude shortwave just
east of the International Dateline will change the pattern as the
mean upper ridge slides away from the longitude of Hawai`i. A
low-level convergence axis is forecast to approach the
northwestern islands by early Friday with a marked increase in the
chances for heavy rainfall. This should continue into the weekend
given the slow progression of the longwave trough the region. The
06Z GFS appears to be particularly heavy with expected rainfall
amounts as a slow-moving, compact surface low moves across the
state. This would definitely enhance the threat for flash
flooding, especially over the mountainous terrain where rapid
runoff would be an issue. Impressive precipitable water values are
depicted in the 00Z ECMWF with numbers exceeding 2 inches in some
locations. Overall, while the heavy rainfall/flash flood threat
exists into the weekend and perhaps next Monday, timing is an
issue as noted in operational model and ensemble spaghetti plot
comparisons. The 06Z GFS/GEFS mean appear to be quicker than the
rest of the pack relative to the slower 00Z ECMWF/CMC and their
respective means. An average of the 06Z GFS/GEFS mean and 00Z
ECMWF/ECMWF ensemble mean would be a good starting point late in
the period given the building spread.


Rubin-Oster