By Niki Morock, Meteorologist
With the system that brought us rain Monday night, there were 215 filtered reports of severe winds (58+ mph) from South Carolina up to the New York/Canada border, as well as seven tornado touchdowns and three hail reports. You could say that our area dodged a bullet since all of the storm reports in North Carolina happened in the western Piedmont.
Improvements in our computer models over recent years have made forecasting severe weather days in advance possible. As with any weather forecast, the resolution improves as the time period gets closer. In other words, there is some confidence several days out about the timing and geographic area of an event. There is usually more confidence with each model run closing in on the event — two days out, one day out, within hours, etc.
If you watch the progression of the models' output, you will see slight changes in space and time with each run. It's normal, and it's why the forecast for Friday published on Tuesday afternoon may look different from the forecast for Friday published on Thursday afternoon.
Metorologists have many computer model options available to use when forecasting. The most popular three include the GFS, which is a long-range model with course resolution used to see up to 16 days out. The NAM-HIRES is used to forecast for a period within 3 days. The RAP and HRRR are used for forecasting within hours. As the time gets closer, the resolution for the models gets finer.
There is also more reliable initial data and fewer assumptions with the higher resolution models. Imagine looking 16 days into the future and trying to predict what the stock market will do. A lot can change in 16 days. Politicians can make rash decisions, gas pipelines can develop leaks, iconic corporations can announce massive layoffs or take-overs, or a natural disaster could bring transportation to a complete stop. Any one of those things can change the economic forecast.
The same is true for weather forecasting. We can look two-plus weeks out and see what we think may happen giving what is happening around the globe right now combined with mathematical theory and basic assumptions, but there are likely smaller details that we are not taking into account because we don't see them yet. At one week closer to the date, we can see more details and have more confidence in the way things are shaping up. By the time the date is tomorrow, we have a pretty good idea of what will happen, where it will happen, and when it will happen. Still, occasionally, we don't get it exactly right, but by the day of the event, we're pretty sure we know what's going on.
Going back to Monday's system:
On Friday, we could tell there would be thunderstorms on the east coast on Monday, and that it would likely be during the second half of the twenty-four hour period. By Sunday, we could see the ingredients for severe weather lining up from northern North Carolina all the way to the Canadian border with New York. By Monday morning, we could tell that the rain should hit the Triangle late in the evening and clear by early morning. By Monday afternoon we could tell that the tornado threat would remain north of the Virginia/North Carolina border, but there was still the potential for severe level winds across North Carolina.
That's pretty good considering how far the science of meteorology has come in the last century! Still, as with all technology, there is room for improvement.
For example, Monday morning's outlook for severe weather that evening showed more of the state having the potential for severe, damaging winds. By the time the storms reached the Triangle, the wind threat had greatly diminished, so narrowing down that geographic area farther in advance is a potential improvement in our forecasting.
Why would that help? Any advances in our accuracy helps meteorologists' credibility when we predict severe weather days in advance. Higher reliability gives the public more reason to prepare ahead of time to protect life and property when a threat exists. Perception being reality — if you buy into that idea — means that we have to change the public's view of how trustworthy a weather forecast is. The best way to do that is to continue to improve the technology we use every day.