This is a question I get asked often. How far in the future can we predict the weather?
Listen to my podcast and I’ll give you a surprising and complex answer.
But first a few suggestions….
The National Weather Service has a website that shows how weather forecasting skills decline over time (see below). This figure indicates that forecasting skill at around 18,000 ft (500-hPa) declines over time for the US global model (NCEP), the Canadian model (CMC) and the gold standard Central Europe. (ECMWF). 1 indicates a perfect forecast. When this measure (abnormal correlation) falls to 0.6, the forecast becomes only marginally useful.
Very good forecast for 3 days. Pretty good though five days, but then fell quickly after that. Some skills continued even after two weeks, but it was very small.
Another way to see skill decline is to look at a series of weather forecast maps to predict growing forecasts.
Below is a breakdown (facts) of sea level pressure (solid line) and low temperature (colors) at 4 a.m. on February 24.
The 144-hour (6-day) forecast has some notable issues, such as a low-pressure area off the California coast that is being thoroughly weakened. The greatness is in both.
Curiously, the low is back in the 180-hour (7-day) forecast, but too far offshore.
In contrast, the 240-hour report (10-day forecast) has a high pressure area heavily misaligned and shifted northward, and cold air (blue) overwhelms the northeastern Pacific. Not good.
I generally don’t trust 7-8 day forecasts unless multiple forecasts are gathered on the same page.
To listen to my podcast, use the link below or access it through your favorite podcast service.
Some of the main podcast hosts:
Like podcasts? Support on Patreon
PS: Here is the total snow for last night. There is some snow everywhere but the amount of snow that can be measured is mainly north of Seattle. In most locations, it melts quickly after falling.
