WeatherWeather Blog

Actions

Forecast Accuracy

Long-Range Forecasting is Hard
Posted at 6:23 PM, Sep 13, 2021
and last updated 2021-09-13 19:34:31-04

The world of weather forecasting has seen leaps and bounds in forecast improvements over the past couple of decades, mostly thanks to computers which can process large amounts of data quickly and perform equations that would take me waaaay longer to do myself.

Short-term forecasting is quite accurate. According to NOAA, a five-day forecast can accurately predict the weather 90 percent of the time. That drops to 80 percent for a seven-day forecast and then down to 50 percent for a 10-day forecast. Granted, there are times when a front speeds up or slows down or clouds move in or out faster than expected and even a “Tomorrow” forecast gets busted. It happens (and the meteorologist whose forecast just busted is VERY AWARE of the fact and gets frustrated by it, too).

Extended PM.png

But do we let limitations with long-range forecasts stop us from trying our best? No! Thankfully, we have climate data and specific long-range forecasts to help us.

So, when you ask me at the beginning of fall what our winter is going to be like, I feel confident in saying it’ll be cold, and we’ll get some snow. We ALL know we can expect those two things during the winter in Nebraska and Iowa. We can look ahead to see if certain months will be more likely to see above or below average temperatures and precipitation, too.

What is nearly impossible to do though, is give you an accurate forecast when you ask, “Is it going to rain on my wedding next month?” I should know. I’ve been looking at the forecast for my wedding day since the beginning of this month (and a little bit before the start of this month, too).

As you might know, some websites (like Accuweather) will give you a 90-day forecast. So, how has their forecast changed over the time I’ve been checking it out? (This is not a dig at or an endorsement for them, just showing their forecast, too.)

August 14th, forecasting 80° and mostly sunny for the date.

August 17th, forecasting 69° and mostly cloudy for the date.

August 22nd, forecasting 76° and rain for the date.

During the month of September, other models (like the Euro, GFS, blend, etc) have been forecasting highs ranging from the mid 70s to mid 90s!

Forecast Table.jpg

But now, it’s the week of (!!) and the forecast is becoming a bit clearer. Highs look to be in the mid 80s, partly cloudy, and a small rain chance. Granted, this forecast currently hinges on a cold front and how far south it will drop (or not) so there could still be some adjustments over the next couple of days.

In the end, it’s not going to make or break the wedding day and contrary to popular belief, I can’t control the weather.

TL;DR - If you’re trying to plan something well in advance, check the climate data to get a general idea of what to expect and then keep an eye on the forecast once you’re a week or less out from the event.