As of today, the 8-14 day extended forecast now includes the weekend preceding Inauguration Day. For the period Jan. 12-18, the computer-generated forecast shown above (human modified forecasts are made on weekdays) indicates a gradual shift eastward of the temperature pattern in earlier forecasts. The area of above-average temperatures has now moved into the Midwest from the West Coast, and the below-average area along the East Coast has shrunk. For the immediate Washington, DC region, chances are still given as 40% or higher that average temperatures for the week ending 2 days before the Inauguration will be below the climatological average.The precipitation outlook continues to show dry conditions for most of the U.S., with probabilities 50% or higher for below-average amounts in the Washington area.
Today's temperatures:
High 45
Low 27
Departure from average: +1
Click here for all Inauguration Weather posts.
Images (click to enlarge): 8-14 day temperature and precipitation forecasts from Climate Prediction Center/National Weather Service/NOAA
Posts
3 comments:
Thanks for your kind comment.
We try to be educational as well as entertaining, so please don't hesitate to ask any questions you might have about weather or climate, and we'll do our best to answer them.
Steve, you absolutely correct about limitations to reliable predictions for a given day, on average, beyond a week or so.
A couple points:
The latest point on the chart of "numerical weather prediction today" from ECMWF shows useful skill on average through day 8. But, the range extends from just under 7 days to almost 10 days. However, the scores are both yearly averages and averaged of the entire Northern Hemisphere (20NB to 80N).
There are seasonal variations with generally speaking summer being more predictable than winter. And, the average skill for any season is generally notably less for limited regions, such as North America. Finally, the variability in skill from one day to the next is much larger.
On some days skill at just 3 days can be less than the average skill at 10 days. And, some days the skill at 10 days can be as good as 3 day predictions.
Bottom Line: Average skill can be very misleading. The key is to identify those days - before the fact -that the forecasts are expected to be better or worse than expected on average! The approach now used for doing this is referred to ensemble prediction many - perhaps 50 - computer runs that differ somewhat in initial conditions and/or model formulation. The concept is that the more/less the individual runs differ from one another, the less/more confidence.
More to this upon request
Excellent points, Mr. T. Since we've been in a relatively volatile pattern recently, even the average predictability is likely to be too optimistic.
Post a Comment