A couple of weeks ago in this column I talked about the summer monsoon season and how tricky it is to forecast weather when we are under the influence of monsoon moisture.
Sure enough, the following week the monsoon kicked in, as if on cue, bringing random showers and thunderstorms to the area, along with a tough week of weather forecasting.
That brings me to the theme of this column: Is the new technology we have making forecasting any easier or any better than it was years ago? My answer is typical of what you might expect to hear from a meteorologist, “maybe, maybe not.”
Weather forecasters rely heavily on computer models to provide guidance about how the weather is going to change.
Back in the day, when I went to college about 40 years ago, we had basically two models to work with. They ran just two times a day and went no further out than about five days into the future.
Back then you relied more heavily on current data analysis, weather pattern recognition and local knowledge of climate and terrain influences.
Today, there are literally dozens of weather models available, predicting specific weather variables for precise locations, and updating information almost hourly. Some of those models are global in coverage and some are run on a regional scale.
Weather forecasts that you pull up on your computer or smartphone are generated from these models. Some forecast services, like the National Weather Service, use a blend of several models to generate their forecasts. Others use only a single global model.
Only when all the computer models are in agreement will all weather forecasts also be in agreement. That rarely happens, as each model has a different way of calculating changes in weather.
Most weather apps, particularly the ones that come with your smartphone, are using only a single global model to generate forecasts. That is so you can use it to get a forecast for anywhere in the world.
The National Weather Service uses a blend of models to generate forecasts, for any point you choose, anywhere within the United States. These are generated from its National Digital Forecast Database, or NDFD.
I believe the data from the NWS’s NDFD is probably the best forecast information available to you, for any point within the United States. As long as you are accessing the correct forecast point for your intended location.
On MountainWeather.com I rely heavily on that NWS database, with some alterations to it for the Jackson Hole and Teton Mountains forecast.
The newer NWS model
Earlier this year the National Weather Service introduced an upgraded version of its National Digital Forecast Database, known as the NBM, or National Blend of Models.
The old NDFD used maybe two or three models to generate forecasts; the NBM is using something like eight to 12 models. Right now some regions of the country are using this newest model technology, and others will likely transition to the new NBM later this year.
The NWS’s Central Region, which includes the Riverton forecast office, has been using the NBM version since June 4. Much of the Western Region of the NWS, including the Pocatello forecast office, has not yet completely switched over from the old NDFD.
Having hour-by-hour weather forecasts available at your fingertips, from any source, is a pretty amazing advance in technology. And a huge change from what was available 40 years ago.
That level of detail is great, when it is correct. Keep in mind, though, that more detail does not always translate to more accuracy.
I expect that this transition to using the newer suite of forecast models may result in more discrepancies showing up when the weather situation is more “unsettled.” I would also expect more flip-flopping back and forth in the forecast as models and forecasters try to keep up with rapid changes, and decipher which models are performing best.
That’s not necessarily anything new to weather forecasting, but instead of having just a couple of opinions to choose from, now there are dozens.
Maybe it was better back in the day when there were fewer models to choose from to help you make your forecast. Because it’s still pretty easy to be wrong.
Something I’ve learned over the years, about the public’s perception of weather forecasts, is: You’re either nailing it or you’re blowing it. There is no in between.