[mccookgazette.com] Partly Cloudy ~ 70°F  
High: 73°F ~ Low: 53°F
Monday, Oct. 5, 2015

Garbage in, garbage out

Friday, November 18, 2011

This is a phrase used in computer science and technology to remind programmers that the information computers produce is only as good as the information programmed into the computers. If the information plugged into a computer is solid, it will produce solid results; if the information is bad, then the results will be bad.

Unfortunately, the public is constantly being exposed to a shell game when it comes to accurate weather predictions. The National Weather Service is already 0 for 3 this year in its ability to predict winter storms in this area and it's only the middle of November. This inability to accurately predict has been going on for years and I've addressed it before. You would think they would get better but they don't.

Private weather forecasting companies are much more accurate than the NWS because they have clients who pay for the forecasts they produce. If they fail too often, they get fired and getting fired is not good for business. But the public can't fire the NWS so they're not motivated by profit to do any better than they have been doing.

Our first winter storm watch was issued on Tuesday, Nov. 1 for 2 to 5 inches of snow across the forecast area, which consists of far eastern Colorado, northwest Kansas and southwest Nebraska. The watch was later upgraded to a winter weather advisory. It snowed lightly for a couple of hours with no accumulation because the temperature was too warm. Five days later on Sunday, Nov. 6, another winter storm watch was issued calling for a widespread snowfall of 5 to 10 inches across the area. This was later upgraded to a winter storm warning and predictions increased to as much as a foot of snow. It didn't snow at all. Not one single flake.

And then nine days later, on Tuesday, Nov. 15, a slight chance of flurries was introduced into the forecast with no predicted accumulation of snow at all. A heavy, wet snow was falling when I got up Wednesday morning and it continued to snow until noon, accumulating three to four inches in McCook.

So when it's supposed to snow it doesn't and when it isn't supposed to snow it does. This isn't sufficient for the general public. Many people in this area travel daily and there's nothing worse than running into an unexpected snow storm and our vehicle breaks down or we slide into a ditch. On the other hand, we often delay or postpone trips when heavy snow is predicted to fall only to be frustrated that we had to change or reschedule our plans based on something that never happened. The point is that people's safety and lives are placed at risk whenever the NWS gets it wrong and it gets it wrong far too often.

I read the scientific forecast discussions of the NWS on Weather Underground when storms are brewing and am always amazed to read that one model is too bullish on a solution, another model is too bearish and so the forecasters in Goodland attempt to solve this conflict by taking a position somewhere in the middle and that position is often wrong. According to Joe Bastardi of Accuweather.com, the models produce unreliable data because unreliable data is plugged into them and when that happens, the result is the title of this column. When the models are consistently off base, we're not getting weather predictions from the NWS; we're getting guesses that often times aren't any more reliable than our own.

But because there's no standard the NWS is held to because they don't have any clients that can fire them, they continue to be satisfied with the same sub-standard predictive abilities that have plagued their offices forever. I am absolutely convinced that when I was a kid, weather predictions were just as accurate as they are now and that was long before computer- generated forecast models. I think back in those days that weathermen actually even looked outside from time to time, rather than depending blindly on a computer program to tell them what the weather was going to be.

As I write this early on Friday morning, another winter storm is brewing to our north. Our forecast calls for temperatures in the low 40s tomorrow with no chance of precipitation. We'll see how they do on this one. A 25 percent success rate would be an improvement over 0 percent which is where they are now.

Everyone knows the story that says that everyone should be a weatherman because it's the only job you can have where you can be wrong more often than right and still keep your job.

This is one governmental agency I wouldn't mind seeing abolished because their year-end success rates just aren't good enough and people suffer because of it.

Fact Check
See inaccurate information in this story?

Note: The nature of the Internet makes it impractical for our staff to review every comment. If you feel that a comment is offensive, please Login or Create an account first, and then you will be able to flag a comment as objectionable. Please also note that those who post comments on mccookgazette.com may do so using a screen name, which may or may not reflect a website user's actual name. Readers should be careful not to assign comments to real people who may have names similar to screen names. Refrain from obscenity in your comments, and to keep discussions civil, don't say anything in a way your grandmother would be ashamed to read.


Joe Bastardi does not work for Accuweather any longer - hasn't for quite a while now.

Tell me Mike, how many private weather satellites are there? Where do private forecasters get their data from?

-- Posted by Geezer on Fri, Nov 18, 2011, at 8:03 PM

Respond to this story

Posting a comment requires free registration:

Mike Hendricks
Mike at Night