Weather Prediction - Can't be that Difficult

[Rich Riddle]

I am in the home in Northern Kentucky at the moment. Just shoveled about four inches of snow. The drive was clear from the previous snow on Monday. The prediction was for one half inch. Because of their prediction, the cities didn't treat roads or send out crews overnight. It has shut down the city of Cincinnati. They are still predicting we might get two or possible three inches of snow. Do these forecasters ever walk outside to see how much has already fallen? The look like total incompetents. Before you defend them saying how difficult their jobs are, realize these folks are highly trained and highly paid. If they can't differentiate a storm from a minor dusting, fire them.

[Jim Matthews]

I grew up near there.
I recall Cincinnati getting glazed with ice so fast that
people would make it to work, and be unable to drive home.

That was back when weather models for the Midwest were
still on paper maps. Understand that it isn't the
weather services that decide when to plow, salt, etc.

That's down to budget constraints at a municipal level.

I'm just down the road from the National Weather service office in Taunton, MA.
We still have tremendous variation in local effects, often severe, within a 5-10 mile range.

Places like Cincinnati and Pittsburgh are at the boundary lattitudes, prone to big swings in atmospheric pressure.
This can happen really fast, and without any fixed start time.

Science is difficult, and imprecise.
It's not like the magic bread transporter,
where Bread goes in and Toast comes out.

I still dunno where the bread goes.

[Scott Shepherd]

There's a guy in this area that's been on the radio for a couple of years and his business is to predict the weather for financial markets. He has come on the radio for years now and slammed weather forecasters and their practices. He posts detailed models on his website and Facebook page (various models) and explains what they are looking at and what they are missing and then he makes his analysis of the events. I've found him to be quite accurate, more so than the weather people. He really has a hate for tv weather people and he's constantly calling them idiots, all the way up to the people at the weather channel.

[Brian Tymchak]

Hmm. The Weather Channel and our local weather (Columbus) had been saying for several days that southern OH would get 5"-8" and even more for eastern half of KY. Time to change the station you're listening to I think..

[Chris Padilla]

Scott,

We need that guy working on climate prediction.... ;)

[Dan Hintz]

In the end, true forecasters are using models created mostly before their time. My issue is why haven't these models improved beyond their current levels given the amount of data thrown at them over the years? I'm curious to know if anyone has tried neural networks as an attack path... I'm sure someone has by now, but I'd love to see the research on it. My only conclusion is the models are not taking into account the correct variables, and no one has decided what those variables are.

[Jerome Stanek]

I have heard that it is hard to predict the weather every where in the world. That is the standard answer from the weather people. My wife hates it when I look at a weather map and put it in motion and say that it is not going to do what the dumb weather people say and I am right more often than they are.

[Joe Tilson]

One of the problems is meteorologist training is done in one area of the country, then they get hired in another area, and they have this idea the weather is the same everywhere. IMO they need some geological or geography training as well. Like has been mentioned before, weather can be different within fifteen miles or less.

[Mel Fulks]

We have a local guy who once predicted an all day rain for a Tuesday,five or six days out. And every subsequent day right
up to Monday night, then he said " that rain is going to miss us". It rained all day long. Some of their goofs might even come from playing to public demand. In a winter without snow they will point to North Pole and say " there's a lot of cold
air and moisture here, if it pushes this big warm front out of the way, we could see snow".

[Steve Baumgartner]

C'mon, give the scientists a break! The basic physics were figured out a long time ago. Meteorologists continue to use these formulations because there has been no new science to indicate that they are wrong. The difficulty is that the math, like the atmosphere itself, is very intractable and sensitive to having the data exactly correct, complete, and detailed. Tiny discrepancies or data glitches rapidly grow into totally different outcomes. The mathematical theory of chaos was developed in part by meteorologists trying to explain why their simulations are so often severely wrong. But it's not a question of stuck in the mud, lack of trying, or lack of openness to new techniques, it is that the problem is really that difficult.

On a different note, the majority of TV "weatherpeople" are hired because of their looks and on-air presence, not because of their meteorological skill.

[John Sanford]

Chaos theory, manifested.

Too many variables, many of which aren't even known, and many of which can't be quantified. With overlap between the two groups.

[Phil Thien]

There's a guy in this area that's been on the radio for a couple of years and his business is to predict the weather for financial markets. He has come on the radio for years now and slammed weather forecasters and their practices. He posts detailed models on his website and Facebook page (various models) and explains what they are looking at and what they are missing and then he makes his analysis of the events. I've found him to be quite accurate, more so than the weather people. He really has a hate for tv weather people and he's constantly calling them idiots, all the way up to the people at the weather channel.

I agree with this. I have a client that has a long-haul fleet and he subscribes to a private forecasting service, and their material is quite detailed and offers a range but as the clock ticks down to the event, the range tightens up. They're really very good.

About fifteen years ago, in fact, I scheduled a visit to the home of a guy that runs a large institution for which I do work. We both spoke as I was leaving for his place and had both heard local forecasters say we were going to get 1-2" of snow. He told me his private service was saying 9-10", that we were going to get significant lake effect snow.

His service was spot-on, we got 10". And fast.

It wasn't a big deal for me because his place wasn't that far away.

[Rich Riddle]

Hmm. The Weather Channel and our local weather (Columbus) had been saying for several days that southern OH would get 5"-8" and even more for eastern half of KY. Time to change the station you're listening to I think..
It was all the stations in this area calling for a dusting to 1/2" maximum. Even in the midst of the storm they were saying that we would struggle to get 2" of snow. All they said on the radio a few minutes ago was that this was a record snowfall in addition to the record snowfall on Monday. Tomorrow is supposed to be record cold.

[Lee Schierer]

I've often shoveled 4-8 of a trace of snow from my driveway. When dealing with lake effect, there isn't a math model that actually tells where it will happen or how long it will last. Just look at the photo taken from the air over Buffalo below and tell me how they can predict snow like that? One side of that cloud line got almost nothing the other side got several feet in a few hours.
307260

[Brian Elfert]

If some forecasters are supposedly producing weather forecasts that are much more accurate than the average weather forecast then why aren't they making millions selling their forecasts to companies that need accurate weather forecasts?

[Scott Shepherd]

If some forecasters are supposedly producing weather forecasts that are much more accurate than the average weather forecast then why aren't they making millions selling their forecasts to companies that need accurate weather forecasts?

They are, but there appears to be a huge rift with those people and the actual weather people. They don't get along well. I think those that do a good job with weather work for companies that make millions from the weather (crops, shipping, etc.) and those that can't, end up on the evening news ;)

Just to show how much love there is, here's a quote from the guy's facebook page :

"Yes you can get an accurate forecast 5 days out...9 days...out 15 days. Just because your Local TV weather or TWC cant do it doesn't mean it cant be done"

It's actually been really interesting following him over the years. He explains in great detail about all the different modeling systems and why the people in the USA use one that is flawed, where the more accurate ones aren't used by them. He uses a number of different modeling types and explains why he uses which one in which circumstance.

[Larry Frank]

The best weather predictions that I have found is at theweathercentre.blogspot.com. His predictions for two weeks out are pretty good.

[Rich Riddle]

They are, but there appears to be a huge rift with those people and the actual weather people. They don't get along well. I think those that do a good job with weather work for companies that make millions from the weather (crops, shipping, etc.) and those that can't, end up on the evening news ;)

Just to show how much love there is, here's a quote from the guy's facebook page :

"Yes you can get an accurate forecast 5 days out...9 days...out 15 days. Just because your Local TV weather or TWC cant do it doesn't mean it cant be done"

It's actually been really interesting following him over the years. He explains in great detail about all the different modeling systems and why the people in the USA use one that is flawed, where the more accurate ones aren't used by them. He uses a number of different modeling types and explains why he uses which one in which circumstance.

What is the forecaster's name? I prefer accuracy over popularity and looks.

[Scott Shepherd]

His company is wxrisk.com. He used to post a lot of free stuff on his site, now I think he posts free stuff on his Facebook page which is linked at the top right of his website. He really only does major weather issues, I think. I don't think you will see him saying "it's going to rain tomorrow", that's not how he makes his money. His comments to people on Facebook are brutal some times. He says what's on his mind.

[Brian Ashton]

On a different note, the majority of TV "weatherpeople" are hired because of their looks and on-air presence, not because of their meteorological skill.

I had a meteorologist that was also the local tv weather personality teaching me meteorology at uni... He said the number one reason certain weather people get hired is because the producer is trying to get in their pants. 2nd was being in the right place when someone says the tv station needs someone to come down and do a weather report asap, any volunteers...

[Phil Thien]

If some forecasters are supposedly producing weather forecasts that are much more accurate than the average weather forecast then why aren't they making millions selling their forecasts to companies that need accurate weather forecasts?

Google Met Office and Meteogroup. There are others.

[Ken Fitzgerald]

Check out this link for the requirements to have that little certified AMS symbol after your name: https://www.ametsoc.org/amscert/cbmrequirements.html

It's a little more complicated than most people realize.

[Brian Elfert]

If some forecasters are supposedly producing weather forecasts that are much more accurate than the average weather forecast then why aren't they making millions selling their forecasts to companies that need accurate weather forecasts?

I'll rephrase my question: Why wouldn't a TV station or radio station buy forecasts from someone who is much more accurate than their own weather person? I imagine a TV station might have a lot more viewers for the news if they advertised their weather forecasts as more accurate than the other stations. Of course, the forecasts actually have to be more accurate to keep the ratings up.

[Dan Hintz]

I had a meteorologist that was also the local tv weather personality teaching me meteorology at uni... He said the number one reason certain weather people get hired is because the producer is trying to get in their pants. 2nd was being in the right place when someone says the tv station needs someone to come down and do a weather report asap, any volunteers...

The first reason sounds like bitterness at getting passed over by another... I've seen significantly more men forecasters than women (a.k.a., eye candy). So unless the majority of producers are gay...

The second reason also sounds pretty suspect. Most of the forecasters I've watched seem to know something about the weather and discuss the hows/whys, they're not just reading off of a teleprompter.

But who knows...

[Larry Edgerton]

Where I live in the tip of the mitt weather prediction is hard to predict because of the great lakes. Slight differences in temp can change weather we get precipitation or not, both air and water temp. This year we did not get a lot of Lake effect snow because the water temp never got up t its normal temp over the summer and the early winter temps were not particularly cold.

Its interesting if you catch conditions just right when crossing the Mackinaw bridge you can watch cumilus clouds forming. It only takes seconds. In the first part of the cold season if the water is warm and the air is cold you will see wisps forming on the water and within a few seconds poof, a cloud is formed. Often when driving inland in low wind conditions you can see the whole outline of the lakes around the tip of the mitt as there is a bank of clouds hanging around the shoreline. Temps within a couple miles of the lake will vary by as much as ten degrees, warmer in the fall and cooler in the summer.

Temp this morning is -26, so I am sure there is a bank of clouds around the lake in the areas where it is not frozen. If the lakes totally freeze like last year we get no lake effect. Interesting tidbit, much more water evaporates from the great lakes than goes down the St. Laurence.

[Andrew Kertesz]

It is the only job where you can be wrong 50% of the time or more and still make a great salary...

[Charlie Velasquez]

While part of a Dept of Defense program to help science teachers grow STEM participation I had the opportunity to work with some meteorologists. During discussions the topic of private weather services (pws) came up. Their explanation:
Yes, almost all private services have a better track record than the NWS. But, they have a different clientele with different needs.


All weather predictions are based on statistical models. The pws have the same data and the same models as the NWS, in fact much of the data used by the pws is from the Nws. These models don't give predictions, they give statistical distributions of events that have occurred given a specific set of data. It is up to the meteorlogists to decide which model to use and how much of the distribution to examine.
The nws main concern is safety. If an ice storm is probably coming DOT people need to be spreading salt. Spreading salt is an expensive endeavor. Spreading salt and getting no ice is a waste of money. But, not spreading salt and getting ice results in the loss of life. Which forecast to give, error on the saving life side or the saving money side? They intentionally widen their distribution of events and that tends to increase false positives because a false negative is much more of a loss.


On the other hand, many pws subscribers are businesses where profits are extremely important. How many snowblowers, snow shovels, and tons of salt should we send to a particular store, should I spray my fruit trees to help ward off the effects of a possible frost... You want the forecast that is most likely to make you a profit. So, the pws can make predictions based more closely to the actual statistical median. And they can even tailor their analysis of the models to take into account the type of industry, for example wind speed and direction might be more important in the trucking industry than say the actual temperature.
It should be noted that most pws defer to the nws when it comes to severe safety issues as most pws give verbatim the nws weather alerts and warnings.


But even then these are statistical models and there will always be outliers that occur. Stuff happens...

[Kevin Bourque]

Weather forecasting isn't about science anymore, its a commercial endeavor. It's all about injecting fear into the forecast to get people to watch your news station.

And then every now and again there a coating of snow that turns into 6 inches...SURPRISE!!!

[Phil Thien]

Weather forecasting isn't about science anymore, its a commercial endeavor. It's all about injecting fear into the forecast to get people to watch your news station.

And then every now and again there a coating of snow that turns into 6 inches...SURPRISE!!!

Yep, that is why, over the years, they've devoted more and more time to covering weather.

Weather is also cheap to produce. The purpose of the station is to sell advertising. They know people are going to watch the news, they have a guaranteed audience. Actual local news stories are costlier to produce, sports doesn't appeal to everyone, weather is the common denominator.

So the local news has become a weather sandwich where the actual news and sports are the bread, and the weather is the meat.

That is probably why we have a weather channel. Because people watch the silly thing.

[Jerome Stanek]

Where I live in the tip of the mitt weather prediction is hard to predict because of the great lakes. Slight differences in temp can change weather we get precipitation or not, both air and water temp. This year we did not get a lot of Lake effect snow because the water temp never got up t its normal temp over the summer and the early winter temps were not particularly cold.

Its interesting if you catch conditions just right when crossing the Mackinaw bridge you can watch cumilus clouds forming. It only takes seconds. In the first part of the cold season if the water is warm and the air is cold you will see wisps forming on the water and within a few seconds poof, a cloud is formed. Often when driving inland in low wind conditions you can see the whole outline of the lakes around the tip of the mitt as there is a bank of clouds hanging around the shoreline. Temps within a couple miles of the lake will vary by as much as ten degrees, warmer in the fall and cooler in the summer.

Temp this morning is -26, so I am sure there is a bank of clouds around the lake in the areas where it is not frozen. If the lakes totally freeze like last year we get no lake effect. Interesting tidbit, much more water evaporates from the great lakes than goes down the St. Laurence.


That is just an excuse all weather people say it has something to do with the part of the country they are predicting in. They called for record lows last night and it wasn't even close to being cold here barely made it below 0

[Jerome Stanek]

Check out this link for the requirements to have that little certified AMS symbol after your name: https://www.ametsoc.org/amscert/cbmrequirements.html

It's a little more complicated than most people realize.

Doesn't sound all that bad nurses and teachers need more than that.

[Curt Harms]

One of the problems is meteorologist training is done in one area of the country, then they get hired in another area, and they have this idea the weather is the same everywhere. IMO they need some geological or geography training as well. Like has been mentioned before, weather can be different within fifteen miles or less.
I think there's something to this. There are also local variables that need to be accounted for. It seems to be better accounting for location on weather web sites vs. TV forecasts. There are also weather 'tracks' due to terrain that I'm not sure are accounted for. We see this every summer in who often gets isolated showers/thunderstorms (north and south of us) and who doesn't (us)

[David Ragan]

I took a meterology course in college cause I needed something easy. Big Mistake.

It is all about the physics of gases/fluids. You have all these masses of fluids (something by definition that takes the form of the container it is in-the blanket of our atmosphere is the container) swirling around. you then factor in the various regions of dry earth/water underneath, the reflectivity of each (albido, i believe is the term), the dynamic influence of absorption of solar energy @ the equator-this is really what drives all the weather on the planet--the increase in energy makes water go gaseous, increase in energy, etc. Give is low pressure cells. Hurricanes, Typhoons, etc. Recall that all fluids/gasses move from high to low pressure. This is a prime driver of weather.

Then, factor in the spin earth (Coriolis force) that accounts for most all weather going from west to the east.

Then, you have the jet stream. My limited understanding-how I conceptualized the jet stream back then (and now) was that it served to 'hold back' the polar air masses....fenced them in, sort of. The cold polar air masses are high pressure cells. Not to parochial, but the tilt of the earth, and subsequent shift of these air massses is what gives us seasons (except for the unfortunates living in FLA)

Then, you have the temperature of the oceans.....2/3 of the planets surface-that influence all the above. That particular temperature is interdependent upon the air mass above in it's own sort of huge, poorly understood oceanic 'weather' system. Remember El Nino?

After all that, if you want to fine tune it, add in the "heat island" effect that our cities have on local weather, as well as the irregularities in the landscape such as the Great Lakes (more downwind rain, etc), and upsloping effect of mountain ranges, deserts, etc.

Then, please recall that for precipitation to actually occur, you have to have 'nucleation' or some such term that means a particle needs to be there for the moisture to form around-rain, snow. Such particles come from dust-while I was in college out @ Univ of Okla we had a real 'dust storm'--the horizon got red clayish- and in a few hours dust was so thick, it was surreal. Then, it rained mud.

Such nucleation can come from factories, deserts, pollen, etc.

You get the picture, and I have to get back to work @ my real job.

Thank you for your attention.

PS I pulled a B in the course.

[Dan Hintz]

Weather forecasting isn't about science anymore, its a commercial endeavor.

I'd tweak that to say "Weather newscasting isn't about science, it's a commercial endeavor."

But weather forecasting? It's definitely about science.

[Jerome Stanek]

I find most news channels use the weather to try and get you to watch. What I can't wrp my head around is why they try to dramatize the weather with unrealistic claims about the wind chill. They try to make it sound like it is always going to be the lowest wind chill that they say when we may get 1 gust of wind the get that chill. Right now they are saying that there may be wind chills as low as 50 below

[Pat Barry]

I took a meterology course in college cause I needed something easy. Big Mistake.

It is all about the physics of gases/fluids. You have all these masses of fluids (something by definition that takes the form of the container it is in-the blanket of our atmosphere is the container) swirling around. you then factor in the various regions of dry earth/water underneath, the reflectivity of each (albido, i believe is the term), the dynamic influence of absorption of solar energy @ the equator-this is really what drives all the weather on the planet--the increase in energy makes water go gaseous, increase in energy, etc. Give is low pressure cells. Hurricanes, Typhoons, etc. Recall that all fluids/gasses move from high to low pressure. This is a prime driver of weather.

Then, factor in the spin earth (Coriolis force) that accounts for most all weather going from west to the east.

Then, you have the jet stream. My limited understanding-how I conceptualized the jet stream back then (and now) was that it served to 'hold back' the polar air masses....fenced them in, sort of. The cold polar air masses are high pressure cells. Not to parochial, but the tilt of the earth, and subsequent shift of these air massses is what gives us seasons (except for the unfortunates living in FLA)

Then, you have the temperature of the oceans.....2/3 of the planets surface-that influence all the above. That particular temperature is interdependent upon the air mass above in it's own sort of huge, poorly understood oceanic 'weather' system. Remember El Nino?

After all that, if you want to fine tune it, add in the "heat island" effect that our cities have on local weather, as well as the irregularities in the landscape such as the Great Lakes (more downwind rain, etc), and upsloping effect of mountain ranges, deserts, etc.

Then, please recall that for precipitation to actually occur, you have to have 'nucleation' or some such term that means a particle needs to be there for the moisture to form around-rain, snow. Such particles come from dust-while I was in college out @ Univ of Okla we had a real 'dust storm'--the horizon got red clayish- and in a few hours dust was so thick, it was surreal. Then, it rained mud.

Such nucleation can come from factories, deserts, pollen, etc.

You get the picture, and I have to get back to work @ my real job.

Thank you for your attention.

PS I pulled a B in the course.
This is exactly why they can't predict the weather - no where did they have you go out, look at the sky, call your neighbor out west, etc, etc. They just tried to make scientific methods to explain what already happened.

[David Ragan]

They said the best method of weather prediction is (i forget the term)....is...."what happened today is what will happen tomorrow"
no formal education necessary

[Brian Ashton]

I'd tweak that to say "Weather newscasting isn't about science, it's a commercial endeavor."

But weather forecasting? It's definitely about science.

And to add to that. I've never had much to do with the media but my wife has... One constant she always says is they're not there to report the truth or remotely get it accurate, they're there to make money. So, why would a media outlet, that is responsible for where the vast majority get their weather information, pay for data in hopes of being accurate when the government gives them all they need to be close enough to draw viewers, for free.

And a couple observations regarding the conversation:

You can't fault the government. They produce the data not interpret other sources. Whereas those contracted to produce highly accurate predictions will utilise a number of data sources.

Regarding your favorite media presenter and their ability to interpret the weather bureau's data... In my travels around the globe one thing is fairly constant. Weather is mostly reported by young girls that fit well into tight dresses. You actually think they went and did a meteorology degree to get the job they have. You'll often also notice is they don't stay in the job all that long either.

[Tom Stenzel]

It's now below zero. Why did they have to get THAT prediction right? I would have settled for 60 degrees and wrong.

Here's a picture I took last fall, think I'll set up a weather station like this one:

307353

:p

-Tom

[Rich Riddle]

I followed advice and started looking at outside sources for weather news. The outsiders indicated we will get 5 - 8 inches more tomorrow. The locals had been calling for another "dusting," so I forwarded the link to one of them. They are now calling for significant snowfall. More record lows as well.

[Mel Fulks]

Rich,I can explain that. Sounds like there is some event scheduled for your area. Promoters buy commercial time from that station .....and tickets are still being sold.

[Rich Riddle]

Cincinnati had it's last "weather man" named Pat Berry. He reported the weather and isn't a meteorologist. People like him better than his replacements. Remember the simple "weather man" type?