Forecast Numbers Are Too High!
By Erica Grow on August 22, 2012, 12:39pm
Use your ← → (arrow) keys to browse more stories.
Forecasting high temperatures in New England has been a bit challenging lately (as if it isn't challenging enough normally)! MOS guidance, which is one of our key tools for forecasting, has been running too high lately. If we fail to recognize this as forecasters, our predictions will be way off from where they should be.
First of all, what is MOS? MOS stands for Model Output Statistics. It uses information from the forecast model supercomputers (the same ones that produce FOUS data and map-based output) and the result looks like a grid with a bunch of numbers. You can check it out for yourself by navigating the NWS MOS site.
MOS numbers are usually accurate enough for us to use them as guidance when forecasting high and low temperatures. Typically, the spring is the least accurate time for MOS numbers, and the summer is the most accurate time. But right now, the numbers have not been very helpful for us as we put together the Storm Team 8 forecast!
In July, the MOS numbers went through an approximate 2-week stretch in which the forecasts were too low. As you probably remember, July was much warmer than average, with Bridgeport's average temperature 3 degrees higher than normal and Hartford's average 2.6 degrees higher (that number includes high and low temps). And now, our high temperatures have been below average for the past 5 days, including today, and all indications are that August will continue to be a bit cooler than average.
Time and time again, MOS's accuracy has been severely challenged by long-term patterns of abnormally cool or warm weather. An experienced forecaster knows that the model data tends to favor normal climatological conditions, especially in the extended MOS data (beyond 3 days). It's likely that the current difficulty we're having with MOS is due to this bias toward climatology. I, for one, am hopeful that MOS data improves in the near future, by utilizing a larger pool of climatological data and forecast possibilities!
Use your ← → (arrow) keys to browse more stories.
