This past weekend we got a good reminder of why weather apps on your phone can lead you astray from a reliable forecast. Listen we all love the convenience of an app that tells us the weather. The problem is most people have no idea where that data comes from or if it’s even a real forecast. I’ve railed against the “Crap Apps” before as most people end up blaming me for the failing of their stock weather app. The problem lies in automation of these apps with a single source of data usually from just one model. That model may be doing poorly or may even just be a bad model, or just might be getting bad data No one model ever does a perfect job. All my forecasts use data from numerous models, their ensembles and local knowledge and experience. There is always a nuance to a forecast that a single automated computer model just can’t give you. Think of when you call customer support for anything. Sometimes the automated computer can give you basic information, but if you really need help, you need to talk to a person, not a machine.
This doesn’t mean apps can’t be useful they are great sources for current conditions and a general idea of what might happen. They are a lot like using WebMD a great place to start but then one should always seek professional human expertise to make any decision.
This Past Weekend:
Leading up to the holiday weekend everyone was concerned about the threat of rain from a weak tropical system. I must have been asked a few 100 times about canceling plans to go to the beach. Each and every time I said NO! please don’t change your plans. The reason was simple for me. Yes, there was going to be some rain around, but it would not wash out the entire weekend. Most saw their “crap app” say 80-100% chance of rain all 3 days and panicked. Well, I put my money where my mouth was, and I actually spent the weekend in Charleston right where Bonnie came ashore. We had a geat weekend, and I likely had it worse than just about any location on the coast.
If you look at the 3 day weekend by hours it was 72 hours long, so how much of the weekend was actually impacted by rain? Here’s a rough estimate based on hourly rainfall reports from the official reporting stations along the coast.
MOOREHEAD CITY/ATLANTIC BEACH/EMERALD ISLE AREA:
Only 8% of the weekend or 6 hrs.
WILMINGTON/OAK ISLAND/CAROLINA BEACH/WRIGHTSVILLE BEACH AERA:
Only 12% of the weekend or 9 hrs.
MYRTLE BEACH/GRAND STRAND AREA:
Only 10% of the weekend got wet or 7 hrs.
CHARLESTON/FOLLY BEACH/ ISLE OF PALMS/KIAWAH AREA:
This was the worst 24 hrs or 33% of the weekend with rain from about noon Saturday to Noon Sunday. We still had 2 solid days at the beach! 🙂
Stick local and stick human for your weather:
I’m never going to be right about the forecast 100% of the time; I am the 1st to admit that. The goal is to try to reduce my errors to as small an possible 10% or less. Remember this is predciting the future. The difference between me and a computer I can see when my forecast is going wrong and start changing it. Apps need to wait until the next model run to change and thats usually every 6-12 hrs. Some apps are better than others but they are just tools that should alwasy be used with caution and if you really have a qustion about the forecast ask me, I will answer back ASAP.