Remember that email I sent to AccuWeather during Rita?

[DoctorHurricane2003]

It was in reference to them calling an NHC product 'virtually useless'

Well I got a typed-out response:

Dear Mr. Adam,

We appreciate your comments, and I wanted to get back to you and explain the situation you referred to in your recent comments on AccuWeather.com. We apologize for saying that “AccuWeather feels this forecast is virtually useless.” It was not our intent to offend anyone with this statement, but rather to bring to the attention of the public what we believe to be a serious deficiency with a group of National Weather Service products, that could lead government officials and the public to make decisions that might lead to increased risk to life and property. As weather scientists and a weather news provider, we believe we have an obligation to point out such issues. At the same time, we recognize that we could have chosen better words to express this.

If you check out the NWS web site (http://www.nhc.noaa.gov/archive/2005/RITA.shtml?), you will find their Strike Probabilities forecasts, which are labeled by the NWS as “PROBABILITIES FOR GUIDANCE IN HURRICANE PROTECTION PLANNING BY GOVERNMENT AND DISASTER OFFICIALS”, and is not an “experimental product”. If you look at these probabilities, which are part of the product suite that includes the NWS graphic that was referred to on our web-page, you will see that they are extremely low relative to the actual threat level.

For example, with Hurricane Katrina, while Max Mayfield, the Director of the National Hurricane Center, was reportedly calling the mayor of New Orleans around dinner time Saturday August 27 to warn him of the impending disaster, the NWS Strike Probability was for only a 21% chance that the center of Hurricane Katrina would even pass within 65 nautical miles of New Orleans (http://www.nhc.noaa.gov/archive/2005/pr ... .018.shtml?).

That means this product, prepared for hurricane protection planning for government officials was predicting a 79% chance that the hurricane would not even come within about 75 miles of New Orleans. Even as late as 4pm Sunday, when virtually all meteorologists (including those with the government) felt a hit on New Orleans was a virtual certainty, this NWS product showed only a 47% chance.

Our concern, which was expressed on our web site, was that people and government officials might see this NWS forecast and its associated graphics, and might take the wrong action based on it, when at 4 pm a day and a half before Rita was going to make landfall it showed the chance of Hurricane Rita passing within 65 nm of Galveston to be only 26%, and by 10pm that day had only increased that to 29%. Put another way, the NWS was issuing some products showing a close pass to Galveston and others showing only a low probability that it would pass near Galveston. Would you like to be the person who makes emergency decisions based on such conflicting “official” information?

We recognized that the way we expressed this concern could be done in a different manner, which is why it was quickly updated on our web site.

But we continue to believe that the concern we expressed is a legitimate one, and one which needs to be expressed.

We do appreciate your comments and hope you will visit AccuWeather.com in the future. We make great efforts to provide the best and most accurate weather content, and to provide information that will help to save lives and property and provide real value.

Thank you and best regards,

X

Email: x

Direct phone: 814-xxx-xxxx
Direct fax: 814-xxx-xxxx




I do give him props for typing up a personal response...but hmm....that's definitely a thinker.

[jasons2k]

I'm not a fan of the strike prob. guidance myself, but I wonder how they would respond if the tables were turned with the now famous "AccuWx vs. NHC" graphic they did with Ophelia. They were using that to make the same point essentially but definately got caught with their pants down that time.

[MWatkins]

There are...of course...a couple of glaring issues with that response.

A hurricane is not a point. Up until landfall, Katrina could have easily missed New Orleans by 1 degree of longitude…which is what is covered in a strike prob.

However…as we saw with Katrina, hurricane impacts can be felt 180 miles from the center…which is the reason the NHC is offering new products designed to cover probability of sustained winds in ranges.

The old strike probs were useful in the days where computers couldn’t do all of the heavy lifting…but by and large this product is not used by the general public to make decisions and is...antiquated. In fact I have been on record since Charley last year stating the product should be revised or discontinued. I feel they are making progress with the new products this season. These were designed for emergency managers. The public generally does not look at them…

I don’t have enough time to get into this further now…but I will say this. This letter is a PERFECT example of how little that company appears to understand (or communicate) about tropical cyclone forecasting.

Just curious…what were they telling their clients when they were forecasting Ophelia to go towards New Orleans?

MW

[DoctorHurricane2003]

I actually typed in something about Ophelia in my initial email, but he apparently avoided all notes on that.

*sigh lol

[brunota2003]

Go here: http://www.nhc.noaa.gov/feedback-pws-graphics.shtml
Last time I checked, the graphic they slammed said this:

These displays show experimental probabilities of surface wind speeds equal to or exceeding 34 kt...39 mph (tropical storm force), 50 kt...58 mph, or 64 kt...74 mph (hurricane force). These wind speed probabilities are based on the official National Hurricane Center (NHC) track, intensity, and wind structure forecast, and on NHC error statistics for those forecast variables during recent years.

Oh, would you look at that, it still says it...

[brunota2003]

Ok, I sent an email to accute wx, it was nasty though, i was ticked off, so, cant wait to see what they have to say to it, ought to be interesting to say the least...

[mtm4319]

For as much as they accuse NHC of errors, Accuweather didn't do so hot on its forecasts either.




That was Accuweather's forecast of hurricane-force winds. Note the 90% chance they give of hurricane winds on the west side of Galveston Bay, and the 10-29% chance they give of hurricane winds just north of the Sabine River's northern boundary between Texas and Louisiana.



Here is the actual swath of hurricane force winds (in red). Hurricane winds only extended to the far eastern edge of Galveston Bay, and extended eastward to just west of Morgan City, where the probability given by Accuweather was just over 30%. Matagorda barely experienced tropical-storm force winds, while Accuweather still gave that area a 50% chance of hurricane winds.

[hicksta]

I support what accuweather's point was. Thought i wish they would of said it in a diffrent manner.

[brunota2003]

Got an email back from Accuwx, things that make you go hmmmmm...:

Your comments about our comments about an experimental NHC product were forwarded to me. This may come as a surprise, but there were people here who questioned the remarks that appeared on our site about the experimental product you mentioned. I'm sure you have been in classes at your high school where not everyone agreed with things other people said.

You also made references to a Senate Bill (S786). I think the purpose of the bill was to make sure the NWS concentrates on its core mission to help protect life and property. I know that some people have read into the language things about privatizing all the weather information and making everyone pay AccuWeather or another company for it. That is actually not the intent of the bill at all. We want all the data released to everyone at the same time. However, we believe that if a private sector company is providing a service professionally, the National Weather Service should not deliberately come in, spend their money (your taxes and mine) to make a competing product, and then give it away for free. I am sure you will agree that the National Weather Service has a very challenging and important role to play in warning the population about any and all severe weather threats. I believe they should receive all the money they need to make sure that the equipment and people they have are able to do the best job in the world.

If you are ever in central Pennsylvania, I would invite you to take a tour of our facilities and see some of the things we do. When the head of the National Weather Service visited, he appreciated the tour and we had discussions about the relative roles of the private and public sectors in this field.