Monday, February 21, 2005
The week ahead and forecast evaluations...
Jason Samenow @ 12:11 AM
Another week gone by, another instance where a snow threat did not materialize. With the days of remaining winter numbered, time is running short for snow lovers. But there is hope...at least, a little.
There certainly won't be any snow today, as we're left in the wake of last night's so-called snow. Expect a cloudy day, with pockets of drizzle, and highs in the mid 40s. If any sun peaks out in the afternoon, 50 is possible.
Tuesday may well be the nicest day of the week. With partial sunshine and a westerly wind, temperatures should climb to near 50.
Winds become more northwesterly on Wednesday, ushering in some colder air. Still, there should be plenty of sunshine with highs 40-45.
Thursday and Friday remain big question marks at this point. A southern storm may move northeast towards our area or may glide harmlessly off the Southeast coast. So we may see a threat of some wintry precipitation, or just dry and cold conditions on those days. High temperatures should be near 40 -- although they will be held down a bit if it is precipitating.
Post mortem evaluation of "snow" event
As I write this, a gentle rain falls whereas for a time, I thought it would be a white Presidents' Day. However, even when I was most bullish about snow, I did caution there was a 40% chance of less than 1" of snow (and a 20% chance of no snow). Furthermore, I (in consultation with others on the CapitalWeather team) made the correct decision to forecast little or no snow accumulation about 24 hours in advance of the storm. While I wish I had never jumped aboard the snow bandwagon, all in all I'm not displeased with our forecast. We easily outperformed all three TV stations (4, 7 and 9) and the National Weather Service, who called for 1-3" and 2-4" when it was clear (to us at least) there would be little or no accumulation.
CapitalWeather defeats Channel 7, 9 and NWS in mock forecasting contest
In last week's "week ahead" post, I posted 6.5 day forecasts prepared by Channel 7 (Joe Witte), Channel 9 (Tony Pann), the National Weather Service and CapitalWeather and indicated I would compare them in a week and see who fared the best. For every degree each temperature forecast was off (13 total), an error point was assigned. The graph on the right summarizes these results in average number of error points (or degrees off) per forecast. It's nice CapitalWeather's forecast was most accurate, but I recognize it's just one forecast and my 'competitors' had no clue their forecasts were being subjected to a contest. We'll do this again periodically and, over time, pool the results.
There certainly won't be any snow today, as we're left in the wake of last night's so-called snow. Expect a cloudy day, with pockets of drizzle, and highs in the mid 40s. If any sun peaks out in the afternoon, 50 is possible.
Tuesday may well be the nicest day of the week. With partial sunshine and a westerly wind, temperatures should climb to near 50.
Winds become more northwesterly on Wednesday, ushering in some colder air. Still, there should be plenty of sunshine with highs 40-45.
Thursday and Friday remain big question marks at this point. A southern storm may move northeast towards our area or may glide harmlessly off the Southeast coast. So we may see a threat of some wintry precipitation, or just dry and cold conditions on those days. High temperatures should be near 40 -- although they will be held down a bit if it is precipitating.
Post mortem evaluation of "snow" event
As I write this, a gentle rain falls whereas for a time, I thought it would be a white Presidents' Day. However, even when I was most bullish about snow, I did caution there was a 40% chance of less than 1" of snow (and a 20% chance of no snow). Furthermore, I (in consultation with others on the CapitalWeather team) made the correct decision to forecast little or no snow accumulation about 24 hours in advance of the storm. While I wish I had never jumped aboard the snow bandwagon, all in all I'm not displeased with our forecast. We easily outperformed all three TV stations (4, 7 and 9) and the National Weather Service, who called for 1-3" and 2-4" when it was clear (to us at least) there would be little or no accumulation.
CapitalWeather defeats Channel 7, 9 and NWS in mock forecasting contest
In last week's "week ahead" post, I posted 6.5 day forecasts prepared by Channel 7 (Joe Witte), Channel 9 (Tony Pann), the National Weather Service and CapitalWeather and indicated I would compare them in a week and see who fared the best. For every degree each temperature forecast was off (13 total), an error point was assigned. The graph on the right summarizes these results in average number of error points (or degrees off) per forecast. It's nice CapitalWeather's forecast was most accurate, but I recognize it's just one forecast and my 'competitors' had no clue their forecasts were being subjected to a contest. We'll do this again periodically and, over time, pool the results.
Comments are closed for this archived entry | Link | 
Email this post
Email this post
