Hawaii Extended Forecast Discussion NWS Weather Prediction Center College Park MD 827 AM EDT Fri Sep 28 2018 Valid 00Z Sat Sep 29 2018 - 00Z Sat Oct 06 2018 Relative agreement among guidance provides fairly good confidence in the forecast into about Tuesday. After that time there is rapidly increasing uncertainty in specifics as indicated by the model/ensemble spread that develops. Late this week/early weekend expect some pockets of enhanced rainfall, over the northwestern portion of the main islands in association with a surface boundary/moisture axis as well as the Big Island which may be within the extreme northern periphery of the moisture shield connected with a disturbance initially southeast of the state. At the time of this writing, the Central Pacific Hurricane Center indicates a 70 percent chance of this feature becoming a tropical system within the next 48 hours. This system should continue tracking well south/southwest of the state into early next week. From later in the weekend into early next week the trades will become better established, with a gradual trend toward southeasterly low level winds as high pressure passing to the north reaches the east-central Pacific. This trade flow will bring in drier air so anticipate relatively light and scattered windward-focused showers from Sunday into early next week. From midweek onward the forecast will involve the combined uncertainties of what happens with an upper trough amplifying to the northwest of the state early in the week along with the position of the likely-to-develop tropical system. To varying degrees the 00Z GFS/ECMWF/CMC show enough amplitude and progression of the upper trough to push a surface boundary close to or into the islands. The 00Z FV3 GFS is actually the most extreme with upper trough amplitude versus other models and the ensemble envelope. On the other hand the ensemble means are sufficiently weak with the upper trough that the state remains within southeasterly low level flow around eastern Pacific high pressure. At the same time solutions diverge widely for the track of the likely tropical system. The 06Z GFS/00Z GEFS mean are farthest east (about 167W as of day 7 Friday) while the 00Z ECMWF reaches nearly 10 degrees longitude farther west and the 00Z FV3 GFS another couple degrees westward. The easternmost solutions for the tropical feature would bring a threat of an axis of considerable moisture setting up over the northwestern islands while the farther west scenario would be much drier. For the purpose of a deterministic forecast, the overall complexity/sensitivity of the evolution aloft favors an intermediate approach for the mid-lower latitude central Pacific upper trough (between the models and flatter means) and likely tropical system. Rausch