Have you ever wondered how we come up with such specific temperatures 7 days into the future? How accurate can that be, really? Let’s explore how we come up with these numbers 7 days into the future and how accurate or inaccurate they can be.
Let’s be honest, predicting the EXACT high temperature more than a day or two into the future can be challenging.
Technically, when a long-range temperature forecast map with all the different colors is displayed, it is not a forecast from Your Weather Authority. It is a forecast from the Climate Prediction Center, a branch of the National Weather Service, so if it is not technically “our” forecast, then why do we show it?
First of all, myself, Tom D. and Tom A. have to decide whether or not we agree with it. The colors on the map indicate areas where above normal temperatures are likely, and by contrast, areas of colder than normal temperatures.
That’s the easy answer, but there is more to these maps. What they are ACTUALLY showing are percentages of above or below normal temperatures. The lighter shades of blue or orange show at least a 33% chance of above or below normal temperatures. The deeper the colors the higher the chance of above or below normal temperatures.
Let us take it a step further to put it into some context. These maps indicate a comparison to AVERAGE TEMPERATURES for a given time of the year, which in our case is early Spring or late-March.
The average high temperature for next week is in the low 50s and looking at this map, temperatures will likely be above average, but there is a catch here. This map does NOT say how much above or below average it is going to be. However, you can make a conclusion by looking at the different shades of color, the darker the color, the higher the confidence level of how much above or below average it will be and this can give you an idea of the magnitude of the warmth or cold.
There is just one more thing we as meteorologists have to take into consideration: SKILL. In other words, how good are these long-range temperature predictions?
The thing to remember is we are not talking about specific high and low temperatures and that’s where SKILL comes into play. If these outlooks are performing better than other computer models then they have a good SKILL SCORE and there is way to put their SKILL to the test.
It is called the HEIDKE SKILL TEST. Any score above ZERO is considered to have a good SKILL SCORE, 100 is perfect. This is the 30-Day average scores roughly about 50…it is not PERFECT but it is not terrible either.
The bottom line is we don’t just throw random numbers on our 7-day forecast. There is much more to it and it is our job to analyze the finer details and hopefully at least come close to what the temperature will ACTUALLY be when it comes to the long-range forecast.