Some thoughts about forecasting.. and the MANY busts we've had this month. We've come such a long way and at times the public hardly recognizes it.
I must say that I totally enjoyed the 6 inches or so of snow that fell today. While I was a bit gun shy because of the last 2 busted events, I tried to convey "banding" and compare it to thunderstorms. There was also a sharp cut off where Farmville had NO snow just 50 miles east and Richmond was in the upper 40's early afternoon while we had heavy snow falling.
The computer models have gotten so much better, but just like you don't notice your children growing each day, you as the common person doesn't always notice or appreciate it.
We often use Jan 25, 2000 as a benchmark for busted storms. We had 7-12 from 5-6 miles west of LYH and DAN east that was 100% not forecasted in our region. However, by 11pm, it was clear snow was coming and advisories were up for 2-4 inches.
January 25, 2000 -- With that, the models were horrible and forecasters used now casting and radar to make a forecast. Everyone was watching radar with bated breath because the storm had taken on a classic "Negative tilt" look on radar. I had just gotten into reading models and local forecast office discussions and it was the NWS office out of State College who broke the silence around 8pm saying something big was up and talked about a foot or so. The model data had an incorrect Jet stream data and that was a 100 mile difference in the "track" of the storm. Not a dry slot, early change over-- track. I can't think of any storm "track" that was off by 100 miles in recent memory.
(For the record, I came home from work around 3pm that day and saw radar and told my roomates 6+ inches over night and at 5am we all had the front door open just amazed at 2-3 inch an hour snowfall on Jan 25, 2000)
With that, some amazing things the models did this month..
1. March 5 we were in the 70's and the models correctly showed within an hour or two we'd change to snow around 7 am Monday AM. I'm not sure even 10 years ago we'd of had a clue of that drastic drop in temps. The cold air actually hurt us because it blasted so far south, it pulled the forcing to our south and a band to our north leaving us in "sinking air" known as subsidence. Hence, a forecast of 5-8 inches ended up 1-2 inches.
2. March 17th- We were sitting at 48 degrees and drizzle when my kids and I walked out of lunch and as modeled, by 4pm it was 35 and heavy snow. I doubt any model would have shown that accuracy 10 years ago.The models also showed the mix with sleet we had late evening and overnight. Had the stuff been heavier, we would have had more snow mixed in with the sleet and snow grains. So, a double whammy based on the "dry slot"
Christmas Eve 2002, a surprise front running event to the Christmas Day storm (rain here, but snow to rain to snow in NOVA and a blizzard in upstate PA to NE) was very similar in that temps were WAY warm for snow but very dry air aided in the process. At that time, NO model showed snow south of the Mason Dixon line, but with dry air we had snow in NOVA, unforecasted on Christmas Eve.
3. March 25- The models did a very good job showing where this band would set up. Most data had it hugging the BR and LYH to CHO being close to the jack pot area, with ROA getting lesser amounts. Looking at the end results, you can see why it made sense.
BUT, NO forecaster followed that data because of the bast 2 busts, late season and the myriad of issues associate with that. (Warmth, rates, etc really matter in late season events, even when it's cold and below freezing)
Now, I feel like I conveyed well that "banding" would be an issue and somewhere would end up with more snow. Now, did I expect 6-7 inches-- NOT AT ALL, but did I think a surprise 4-5 inches was possibe? Yes! and I said that. I'm a weather hobbyist with 300 readers. The NWS, WSET and everyone else who is a paid professional has to appeal to the "least common denominator" and on some levels that dumbs down forecast because they have to. Some people just want to know how much, when and how will the roads be. Some what to a little more like how and potential and a crazy few like to know even more. So, a huge challenge is created because of these different needs.
So, my challenge to you as my readers is to NOT look at this as a month of Weathermen still getting paid to do their job despite the wrong forecasts, but look at the growth in the field. If you step back from the rough month, outlooks are getting better. Think of our "big" event from Feb 12-13-- we had a week lead time on that event and the window for start time was within 3-4 hours for about 3 days out. Despite some short term set backs, we are better off with Hurricane tracks, Tornadoes and severe weather and yes, even snowfall forecasts despite this challenging month. Forecasts are saving lives and making people aware of day to day conditions and we are getting better at doing it.
