A Botched Forecast: What Went Wrong and How We Get Better
The hate mail rolled in today, mostly stemming from folks not getting the forecast amount of snow in their backyards or that the snowfall amounts were underdone. We “try” to predict the behavior of a science that is far from perfect, and more times than not, we come up short.
Since the start of the week, we had forecast a significant winter storm that would take shape across the Southeast and Mid-Atlantic regions, while threatening the coastal Northeast regions. Forecast models are your friend, but they can be your enemy at the same time. I always point out that computer models are ONLY a tool for the forecast and NOT a forecast themselves. Looking at model parameters illustrating surface precipitation type, QPF and projected snowfall amounts will do you NO good. We have to look beyond the surface to understand why what is projected at the surface is being projected there. With a number of items that need to be taken into account into a forecast, sometimes something gets overlooked, and the parameter that was overlooked was a critical one.
As the event transpired, it was becoming noticeable that warm air aloft was beginning to build just above the surface, which led to one of two things:
- No snow at all for areas that were forecast to receive snow
- A limited amount of snow giving way to more freezing rain
Because of these two things, the forecast was missed in several locations. Warm air aloft is not something models are great at picking up days in advance. It causes a forecast to shift into a “nowcast” event.
While computer models lead to a great deal of error at times, it’s the job of the forecaster to be able to weed out the fallacies of a forecast model and produce the best possible forecast they can and make adjustments as needed. The issue is, at times, we develop a one-track mind and overlook tools that can be critical. In this case, I overlooked sounding data near the surface and while I anticipated the chance that warm air aloft could build, I didn’t anticipate the magnitude/strength of the warm air layer above the ground which would, and in fact did, limit snowfall amounts in several locations.
Across the Northeast, many coastal areas got a good blast of snow today, and many areas got much more than forecast as the precipitation shield became far more expansive than anticipated as the low was within closer proximity to the coast than expected. All told though, it was noted many times throughout the week of the threat that back edge snow could make its way into coastal Northeast regions.
Overall, we can learn a lot from this winter storm: where we went wrong and where we MUST improve, especially on the communication front. Am I apologetic for the Southeast getting as much snow as it was forecast to receive, not at all. I’m ecstatic they did not receive as much as the last time they saw a significant winter storm, a city gridlock, lockdown and overall humanitarian crisis followed.
Onward and upward…