Why do some meteorologists get the forecast right and some are totally off base? How do forecasters predict the weather? Which forecasts can you trust? Ask Science explores the science behind weather forecasting.
One of the complicating factors in the prediction for Juno, an issue that is actually quite common, was that several of these weather forecasting models disagreed. When faced with two models predicting different answers, the National Weather Service chose to go with the model that had proven to be very reliable in the past, but in this case, left them with over-predictions of snowfall in New York City.
The Weather Channel took a different approach, and instead essentially averaged the results together (with some weighting towards more reliable inputs). Thus, when one of the model predictions turned out to be, in essence, wrong, that effect was diluted by including the results from the other, more correct model predictions.
I make these sorts of decisions all the time with the form of chaos I am most familiar with – my toddler. Often when we go to the park, I watch her devour 4-5 sticks of cheese. On our trip last week, however, she refused to eat anything, and I ended up wasting all of the cheese that I brought.
So the next time we go, I can choose to take an average and bring just 2 cheese sticks so I’m somewhat adequately prepared for either scenario, or I can prepare for the worst and bring all 4. What’s the right answer? I won’t know until we get to the park.
Which Forecast Can You Trust?
More current forecasts (called short range forecasts) will always be more reliable because they contain the most up-to-date information on a pending storm’s initial conditions. Due to improvements in models and the ability to manage the vast amounts of weather-monitoring data being collected, meteorologists can now predict weather 5 days in advance as reliably as they used to predict weather only 2 days in advance 20 years ago. Most weather reports do not go beyond 10 days in advance and rarely are any given more than 2 weeks ahead of schedule.
My good friend and meteorologist, Morgan Miller, also points out that local forecasters often have the best insight into weather patterns in their area. “When looking at the weather in a specific region, I trust the way local meteorologists look at all of the models and combine their personal knowledge and experience,” she says.
To improve the accuracy of reported weather predictions, they could be supplied with different confidence levels.
Imagine your local weatherperson telling you that over the next 12 hours, we have an 85% chance of seeing 3 inches of snowfall, and that over the following 12 hours, the snow will accumulate to 10 inches at the 60% confidence level. Not really what you want to hear, right?
We the public are usually pretty demanding when it comes to our weather forecasting. We want to know if our commute is going to be affected or if our kids will have a snow day. We tend to like our news delivered with confidence, even though (as we have seen) a prediction without uncertainty does not necessarily mean it will be totally accurate.
In the end, weather predictions help to save lives. When you know a storm is coming, you stock up on supplies so that you’ll be warm, fed, and, most importantly, not out in the elements. So instead of criticizing meteorologists for not being able to do the impossible and predict the weather with total certainty, let’s thank them for their efforts to get it right every time!
That ends our show for today. In the meantime, you can become a fan of Ask Science on Facebook or follow me on Twitter, where I’m @QDTeinstein. If you have a question that you’d like to see on a future episode, send me an email at firstname.lastname@example.org.
Until next time, this is Dr. Sabrina Stierwalt with Ask Science’s Quick and Dirty Tips for helping you make sense of science..