Gee, you got a pay grade? Nice.
And I can assure you my meteorological insights are worth every penny. It would be nice if Punaweb had a resident meteorologist, but I guess FB is punatic Harry Durgin's platform of choice.
As for the models, if you take a shorter prediction time frame like +120h (5 days) and click through the last 8 Model runs (2 days), all the predictions are more constrained and consistent, with much less variation both in north to south spread and forward movement. Consistent enough that NWS is willing to average out the ensemble models and provide a firm 1-3, and hazy 4-5 day, path and cone report. (Tropical Tidbits has some good compilations of the differing model predictions for Kiko)
As you increase the prediction length to +144h, +168h, etc, and look again at the last 8 model runs for each, you'll see that variation quickly increases for both the spread and forward movement. As Edge mentioned, by the time you reach +240h, those extrapolations from initial data diverge immensely, with some models placing the storm far past us, and other placing it still to the east of us, creating predictions with a "Disco Stu" level of reliability.
In short, like we've all realized over the years, anything beyond 3 - 5 days is basically anyone's guess (including the meteorologists and the models!) At least Kiko is being considerate with the timing and giving us the weekend to make final preparations within that 3 - 5 day window of warning - maybe enough time for a little costumed dance if Edge will share the lion and dragon adornments?
And I can assure you my meteorological insights are worth every penny. It would be nice if Punaweb had a resident meteorologist, but I guess FB is punatic Harry Durgin's platform of choice.
As for the models, if you take a shorter prediction time frame like +120h (5 days) and click through the last 8 Model runs (2 days), all the predictions are more constrained and consistent, with much less variation both in north to south spread and forward movement. Consistent enough that NWS is willing to average out the ensemble models and provide a firm 1-3, and hazy 4-5 day, path and cone report. (Tropical Tidbits has some good compilations of the differing model predictions for Kiko)
As you increase the prediction length to +144h, +168h, etc, and look again at the last 8 model runs for each, you'll see that variation quickly increases for both the spread and forward movement. As Edge mentioned, by the time you reach +240h, those extrapolations from initial data diverge immensely, with some models placing the storm far past us, and other placing it still to the east of us, creating predictions with a "Disco Stu" level of reliability.
In short, like we've all realized over the years, anything beyond 3 - 5 days is basically anyone's guess (including the meteorologists and the models!) At least Kiko is being considerate with the timing and giving us the weekend to make final preparations within that 3 - 5 day window of warning - maybe enough time for a little costumed dance if Edge will share the lion and dragon adornments?