As you hopefully know, tomorrow will be the day we start showing you our winter forecasts for how much snow we’re expecting and how cold it will get.
All the forecasts are verified for KCI…which I know is a sore point for some of you and I understand that. You may live 30 miles away, perhaps farther south or somewhere else…but yet we use KCI as as verification point.
We do that because they take weather observations 24 hours a day with snow measurements as well. We need to have accurate data to verify with and KCI is the official weather station for KC. It’s a statistical thing as well.
All data collected there will live on in prosperity to be referenced time and time again over the course of decades to come. It’s been that way since they opened the airport in the early 70s.
I get just as frustrated as you do sometimes with it…but it is what it is when it comes to this.
Today: Cloudy with perhaps a few thin spots. seasonable with highs in the upper 40s
Tonight: Showers arrive towards daybreak with lows in the upper 30s
Tomorrow: Off and on rainy, especially in the morning through early lunch. Highs in the mid 40s
Friday: Mostly cloudy and chilly with highs in the mid 40s
As a lead up to the winter forecast coming tomorrow, and then there will be a winter forecast recap blog on Friday…Jacob did some number crunching concerning the verification of our previous winter forecasts over the years.
I don’t think there is any TV station in the country that is as transparent with their forecasts day in and day out than we are here at FOX 4. Each night I show you if we were right or wrong from the previous days forecast. We verify temperatures and sky conditions, a combination that is unique in the country. Does you no good IF I say it’s a sunny forecast and we get rain for half the day, but yet nail the temperatures.
So Jacob wanted to see how our winter forecasts have done over the years. It was an interesting exercise…some successes and some failures. Trying to accurately pinpoint a snow total for the following 4 months or so at one point in the Metro can be fools gold…but we try every year. I’ve personally had some great winter forecasts blow up in my face because of a 25 mile distance.
Last winter my forecast of around 13″ blew up because of a snow storm in early February that gave KCI 7″ of snow…the winter total ended up at around 20″ last winter. That SAME storm gave St Joseph, 25 miles north…FLURRIES. A shift farther south of the storm and I was golden last winter…but alas it doesn’t work that way…a point is a point and that’s what we go with.
So without further ado…here is what Jacob put together and wrote out for you.
How accurate are the FOX4 Winter Forecasts?
Our FOX4 weather team takes great pride in being certified by WeatheRate with the most accurate forecasts in Kansas City. But our winter season predictions of snowfall and coldest temperature aren’t checked by WeatheRate. So we wanted to verify how accurate the FOX4 seasonal predictions are to show you how we compare to the infamous almanacs and if you can put weight in our snowfall and temperature predictions this go around.
Every December, the FOX4 weather team comes up with their own predictions for how much snow will fall and how cold the temperature will get for the winter season at Kansas City International Airport. To verify these forecasts, we will look at each variable individually — snowfall predictions and temperature predictions. Those are already the two most standard ways to predict a winter season and are also used by the Old Farmers’ Almanac. We can compare those two predictions made for every year to the actual outcome of the winter season.
The baseline for every winter forecast is “normal” snowfall and temperatures in Kansas City. We forecast if snow and temperatures will be above or below average. Typically, we like to use a 30-year moving average, that can account for variations but not put too much weight on outliers. From 1991 to 2021, the average snowfall is 18.2″ with an average coldest temperature of -5 degrees.
Now that we’ve defined the variables and time period we are predicting & verifying for, let’s look at the data. Over the last ten years, FOX4 has predicted above normal snow for five winters, exactly 50% of the time. 2 of those 5 years resulted in snowier than normal winters. 3 were below normal.
Similarly, over last 10 years, FOX4 has also predicted below normal snow for five winters. 3 of those 5 winters resulted in below average snowfall. So when it comes to snowfall, our accuracy rating is 50% — we’re only right about half of the time.
When it comes to the coldest temperature, FOX4 has predicted warmer than -5° six times over the last decade. That has verified three of those six times. And again for colder than -5°, we have a verification two of four times. So only a 50% accuracy rating for coldest temperature forecast.
Our accuracy finally improves when we look at the average winter temperatures. Over the last ten years, FOX4 has predicted above normal temperatures for six years. 5 of those 6 years did indeed have warmer than normal temperatures.
When it comes to below normal temperatures, that was our forecast for four of the last ten years. All 4 of those 4 years verified with colder than normal temperatures. So our winter average temperature prediction accuracy is actually 90%! Not bad for a seasonal forecast record.
When we add up those three accuracy ratings, the overall FOX4 winter season predictions have been right 63% of the time over the last decade. While that number is certainly lower than our day to day forecast accuracy of 88%, it is higher than other publications seasonal predictions. Take the Old Farmer’s Almanac as an example. Their snowfall and temperature predictions have been right only 35% of the time over the last decade.
You can look back at our last ten winter predictions & observations below. Verified predictions are in green, missed predictions are in red.
Thanks for that Jacob.
I don’t beat myself up too much about these things…it’s a fun exercise and you sort of try to learn each year a better way of coming up with a forecast. One of our former meteorologists, Garry Frank, has taken the “Golden Snowflake” Award the last couple of years for accuracy.
As a team we all come up with our individual ideas…then average them al lout for the team forecast…sort of like ensemble forecasting I guess…get 5 additional data points and then average them out for a new point.
There are lots of different things I personally look at…some dive deep…sometimes with merit…other times without. Predicting the weather on a day to day time frame is rough enough…go out 4 months for precision for two variables…snow total and coldest temperatures is sort of nutty…but the viewers enjoy it and it’s sort of fun to push the science to some extent.