Around 2007 I began tracking accuracy of my forecasts for a few years. Then I started posting it online in 2009. After getting tired of the “you guys are ALWAYS wrong” calls and emails, I figured it’s time to post the stats online. What follows is data from a 13 year period ending in the summer of 2022. By that time, I realized it was pretty much the same each year; most of the time our forecasts are pretty accurate! You just remember the few really big misses (forecast busts).
The criteria I used? I tracked low/high forecast temp for PDX (Portland International Airport), a central location in the metro area. I track just my forecasts which are generally Monday-Friday. The forecast must be within 3 degrees of the actual temp for the following calendar day. For rainfall it’s a bit more straightforward; if rain is forecast and it is too little to get measured at PDX I’m wrong. Same with a no rain forecast and rain actually arrives. In the comments section I give my thoughts and grade myself down if the numbers down seem to represent reality. For example: I say it’s going to rain tomorrow by late morning. Then let’s say the rain doesn’t arrive until 11pm, leaving the day “rainy” in the climate stats, but to any normal person it didn’t REALLY rain. Got to mark myself down some for that. Or there is no category for “major screwups” (2-7″ unexpected snowfall 12/2009), which don’t really fall into rain/clouds or temp columns.
Chief Meteorologist Mark Nelsen
CLICK ON ANY OF THE IMAGES FOR A LARGER VIEW…