As many of you know, MemphisWeather.net is the official weather partner of the Memphis Redbirds. So it goes without saying that I communicate with them (the head groundskeeper in particular) regularly on details of the forecast and, even more often, short-term trends during inclement weather. Last night, as the Birds were chewing up and spitting out the Nashville Sounds on a warm night with fair skies and a 0% chance of rain, I was sent the following screenshot of the hourly forecast on a national weather app by the groundskeeper:
Being a prudent and thorough groundskeeper, he is apt to check multiple sources of weather information, even on calm nights. When he came across the forecast above, with its elevated rain chances, he checked radar and noticed an echo over downtown. With a forecast of rain from that app (I don't mind calling it out - The Weather Channel) and the radar image with an "echo" on it, plus the knowledge that it wasn't supposed to rain, he reached out to me. I then grabbed the screenshot below from RadarScope.
|Radar reflectivity showing a region of enhanced returns over downtown, which were persistent for perhaps an hour Friday evening prior to sunset.|
I was obviously cynical, but also curious, so I did some additional radar interrogation, including looking at one of the new dual-polarization products (correlation coefficient, or CC) shown in the right panel below. It shows blue colors co-located with the reflectivity returns on the left indicating SOMETHING present in the air. The bright blue colors are low CC values, meaning it was NOT precipitation being detected. The objects being detected were non-uniform in shape/size.
|Radar reflectivity (left) and correlation coefficient (CC - right). CC values indicate that the returns over downtown were not precipitation, but objects of irregular size and shape.|
I returned a message to the groundskeeper, assuring him there was still a 0% chance of rain and that his national app was wrong (surprise, surprise). To me, it appeared that the radar was picking up on waterfowl returning to roost for the night on the banks of the river near downtown. I have seen this before (on a repeating basis in fact) just before sunset in roughly the same place. I believe Mud Island is a favorite "home" overnight for birds and have had some corroboration of that fact from downtown residents in the past.
|What is the chance that these birds are detected by Doppler radar? About 59% apparently.|
So what to make of the crap app with the 59% chance of rain? Most national apps use a single computer model for their forecast, which results in not only a POOR forecast when the model isn't performing well, but one that is also very inconsistent, often changing several times a day.
In the case of The Weather Channel app, I understand that it is a little more sophisticated and actually performs slightly better than the typical national app. So what do I THINK happened? Though I can't prove it, I believe that TWC's app picked up on the radar echoes around downtown and "smartly" adjusted the rain chances for the first couple of hours of the forecast! If this was indeed birds that tripped the app up, it gives a whole new meaning to the term "crap app."
By the way, as opposed to the crap app, I never changed my forecast from the original 0% chance of rain. That's the difference between using a national app and shopping local. MemphisWeather.net's app and the forecast contained therein are HUMAN-POWERED. I write the forecast, using all available model data, plus intuition, local knowledge and experience. Most of the time, that will result in a much more accurate product and one that is definitely more consistent. I think that's worth far more than the $0.99 I charge for it. Check it out for yourself.
|Screenshot of the forecast from the human-powered MWN app|
Follow MWN on Facebook, Twitter, and Google+
Visit MemphisWeather.net on the web or m.memphisweather.net on your mobile phone.
Download our iPhone or Android apps, featuring StormWatch+ severe weather alerts!