Should Scientists Be Sued?

Here is a thought exercise:

Dr. Gassious T. Jones, head of Titanic Drug Company, writes an op-ed in the Wall Street Journal that says, “Widgetex improves 100% of colds within 24 hours.” 

The same day the op-ed appears, Dr. Jones writes a private email to a colleague that says, “Widgetex isn’t curing colds. It is a travesty.” 

A year later, the private email becomes public. What do you think would occur next?

I think we know what would happen: There would be a blizzard of articles condemning drug company greed and a race to the courthouse to file class action lawsuits where lawyers get millions and the ‘victims’ get coupons for 25¢ discounts.

Of course, this really happened in climate science. As revealed in the first release of “Climategate” emails, a Colorado scientist wrote an article for a local publication that said global warming was “incontrovertible” while privately writing — the same week — “we can’t account for the lack of warming…and it is a travesty that we cannot.”

In Climategate, of course, the media came to the scientist’s aid with “the science is sound” articles. The climate scientist is still gainfully employed and, last week, wrote a pro-global warming op-ed in The Wall Street Journal.

Why do I bring this up?

Meteorologist and statistician Matt Briggs has a thought-provoking article over at his blog asking, “Should Scientists Be Held Legally Responsible for Their Results?” Before you form an opinion, please go over there and read both his posting and the comments. I think you’ll find both are excellent.

My thoughts: If we are talking about results (as opposed to predictions, theories, hypothesis, forecasts) then the answer should probably be “yes” if there is fraud involved. In other words, there should be penalties for  publishing or promoting test results known to be false, especially if taxpayer dollars are involved.

But, science cannot progress if hypothesis or forecasts are held to a legal standard of accuracy.

Right now, science cannot forecast earthquakes. We know it would be a tremendous good if earthquakes could be forecast — thousands of lives could be saved and millions of property loss averted. But, if the first tentative forecasts were snuffed out by lawsuits due to inaccuracy, we’ll never have earthquake forecasts. Science progresses through trial and error.

Those are my thoughts. Go over to Matt’s blog and see what you think.

And, to see how this might work in real life, there is an major investigation due to flooding in Australia when a dam was allegedly not properly managed during a forecast of heavy rain. Roger Pielke, Jr. has more info over at his blog.

AMS Annual Meeting, Part 3: Jobs

Here is an essay from the American Meteorological Society meeting from one young meteorologist concerned whether there will still be jobs for human forecasters or whether computers will do all the work.

My estimation is that there will be a role for human forecasters for at least the next quarter century.

After that, who knows? The advance of both weather science and computer science (“Moore’s Law”) is so rapid, I won’t hazard a guess beyond that period time.

Gearing Up for Tornado Season 2012, Part 6: All Tornadoes Are Not Equal

Fifty years ago, meteorologists thought that all tornadoes, more or less, were created equal.  We now know that is not the case.

Without going into all of the details, we know there are at least five different ways that tornadoes can form (supercell, landspout, gustnado, squall line, and hurricane/tropical storm-related) and that we can, roughly, anticipate some limits as to tornado intensity.

The least intense type of tornado is the gustnado (photo below).

Like the whirl that forms on the edge of a canoe paddle as it is being drawn through the water, a gustnado forms as very cold air pushes through very warm air. This type of tornado might knock over your trash can or patio umbrella but it does not pose a significant threat to life or property with untied mobile homes a possible exception. Since its duration is tens of seconds, there is a growing number of meteorologists that believes we shouldn’t issue warnings on them because, in addition to the low level of the threat, the gustnado will likely be gone by the time the warning gets out.

At the other end of the spectrum is the violent (F-4 or F-5) tornado such as the storms that struck Joplin and Tuscaloosa. We are beginning to have the skill to differentiate when the atmosphere might produce this type of storm.

The supercell-created tornado can produce a tornado over tens of miles or more. Even though they are less than 2% of all tornadoes, they account for the majority of deaths.

Below is a composite image of the supercell that produced the Tuscaloosa-Birmingham tornado and others.

For example, there are now two levels of tornado watches: Regular and Particularly Dangerous Situation. If the NOAA weather radio or the radio or TV announcer uses the PDS words, there is a chance of major tornadoes.

For a regular tornado watch, I suggest you start monitoring the weather when the sky darkens or thunder is first heard. No need to alter your normal routine.

With a PDS watch, I recommend you periodically check the weather for the duration of the watch. If storms (tornadic or not) are within roughly 30 minutes, I’d suggest gathering up the kids from soccer practice, bringing in lawn furniture or anything that can be blown about, and check on elderly or infirm friends or relatives. Once everyone is accounted for, I’d make sure you are 5 minutes or less from shelter.

The PDS forecasts are not perfect. One was in effect for Tuscaloosa but Joplin was under a regular tornado watch.

This spring, keep your eye out for PDS watches. It is another step forward for the users of weather science.

How Good, or Bad, Were the Blizzard Forecasts? Part 1.

One way meteorologists get better at what they do is by holding themselves accountable for their forecasts by validating them after the fact.

Here is the National Weather Service’s total snowfall map as of midnight CST this morning.

click to enlarge

And, here is a detailed map of the High Plains where the heaviest snows fell.

The dark blue area in southeast Colorado is more than twenty inches.

This blog provided its first alert of a major winter storm in the Plains at 7:52pm Friday evening by linking to Mike Umschied’s blog. I followed up with my own analysis of the potential storm at 9:18pm. It included this graphic for up to 11″ that you can compare it to the NWS graphic (above) of actual snowfall.

This initial forecast, about 48 hours before the snow started falling in New Mexico, isn’t too bad but is too far southeast. Because of the holiday travel period, I saw this posting mostly as a “heads up.” My sense is it served its purpose.

The next update was 9:11am Saturday. Because of the uncertainty over the path of the storm, I presented probability maps so readers could gauge their risk as I thought it was too early to plot out an exact path.

I posted an update at 3:19pm Saturday, here is a reproduction of the most important part (click to enlarge):

I presented the northwesternmost and southeasternmost models and provided a list of roads (immediately above) with advice. This turned out to be exactly correct: All of the listed roads were closed.

Sunday evening, I posted the following:

As Interstate 70 was indeed closed between Salina and Colby (actually, WaKeeney to Colby), this was a very good forecast as was the forecast map. I was late catching on to the heaviest snow actually falling in southeast Colorado and the final map over forecast the amount of snow in parts of western Kansas by several inches.

The snow started in New Mexico shortly after the above “storm total” map was posted. So, from this point on, I was updating and nowcasting (short term forecasting) the storm. The above was the final map depicting the amount of snow. You can compare it to what actually fell. I believe they compare quite well but please form your own opinion.

In Part 2, I’ll discuss the response to the forecasts.

6 Natural Disasters that Ruined Nuptuals

The Huffington Post recently ran a piece about six weddings disrupted by natural disasters.

Disaster #1 is Tornado Destroys Wedding Reception. While it was more likely a severe thunderstorm’s gust front, the reception was ruined as the storm hit just after the couple said their vows and completed their wedding ceremony.

As you listen to the audio, you can clearly hear the words, “nobody forecast this!”

So, I decided to see if that was true. Turns out it was well forecast. The location of the reception was at the tip of the red arrow. The blue outlined counties, including the location of the reception, were under a severe thunderstorm watch (≥60 mph winds). The thunderstorms that intensified and ruined the reception are west of Chicago (orange arrow).  

The radar echo of the storm, four hours and five minutes after the watch was issued, at about the time the storm struck the reception.

Another of the “ruined” events was a wedding in Vermont washed out by the floods associated with the remains of Hurricane Irene. We have talked previously on this blog about how well forecast that was. Nevertheless, members of the wedding party had to be helicoptered out.

So, I wish to use these unfortunate outcomes to highlight two points:

  • When planning an outdoor event, take a look at the climatological records for the area in question (i.e., your home or the “destination” for a destination wedding). We can research it for you (for a reasonable fee) at AccuWeather or you can do the research yourself at NOAA’s National Climatic Data Center. Example: An evening outdoor wedding in early June in Wichita has a six times higher chance of being washed out than one in early October. Temperatures are more likely to be favorable in early October (less humid).  Early June weddings here are high risk. Even when playing the odds, make sure there is a “Plan B” to move the event indoors and there is adequate shelter if something serious develops. 
  • Take a look at the forecasts as the event approaches. This is a good source from AccuWeather. 
  • Designate someone to watch the weather if it appears the slightest bit threatening before and during the event. Make sure there is shelter adequate for all of the guests. 

Bottom line:  individuals have to take responsibility for monitoring their own weather in critical situations.  

So How Did that Drought Forecast Work Out?

As a meteorologist, I have to be concerned about the quality of my forecasts. I always want to know whether they are correct.

On November 11, I posted a piece called Dagger in the Drought? where I forecast significant drought relief to the southern Plains between then and Thanksgiving. Since no rain is forecast in the region of interest tomorrow, we can go ahead the validate that forecast.

Here is the map I presented on the eleventh:

And, below here is the actual rainfall from the 11th to today:

click to enlarge, data from NWS

The forecast rainfall between the DFW Metroplex south to I-10 between Houston and San Antonio is too high. Elsewhere, the forecast is quite good (considering it is a 13-day forecast) right down to the dry area from western Kansas into northeast New Mexico.

The purple on the lower map represents areas where 10 inches or more fell. Those areas were entirely within the European model’s forecast area of 5 inches or more. Remarkable!

When I made the forecast, I was expecting a third storm on Thanksgiving day. Turns out the models were about 24-30 hours too fast. It will move across the Plains during the calendar day Friday and into the pre-dawn hours Saturday. More helpful rain expected.

Considering how far into the future this forecast stretched, I’d give it an A-, with the minus for the over forecast in central Texas where rains were lighter than forecast.

What grade would you give it? Feel free to post your thoughts in the comments.

Amazingly Accurate Forecasts

While the media keeps whining about inaccurate weather forecasts – regardless of whether the forecast was correct or not — the science of meteorology keeps making tremendous strides forward.

Saturday night, after the record earthquake in Oklahoma, I wrote this posting at 1:12am Sunday morning:

It explicitly forecasts a chance of strong tornadoes in the hatched area of southwest Oklahoma more than 36 hours before the first tornado occurred.

Between that first posting about the tornado threat and the development of thunderstorms in Oklahoma, I did three others highlighting the potential for strong tornadoes this afternoon.

Then, at 10:40am, I did this post which depicted the National Weather Service’s experimental radar forecast system for late this afternoon and evening. I highlighted the area where the big tornado-producing thunderstorms would be most likely with the red square.

While the timing was off (it is an experimental product), it did a fine job of highlighting the area at greatest risk. The first tornado warning in southwest Oklahoma was issued at 2:14pm, 3.5 hours after the posting. Fifty-seven minutes before the first tornado warning, the tornado watch was posted for the area where the tornadoes occurred.

This is just one of the strong tornadoes that occurred in southwest Oklahoma:

As far as we know at this point, there have been no fatalities or serious injuries from these storms.

So, more than 36 hours in advance, weather science was able to explicitly forecast that strong tornadoes would occur in southwest Oklahoma. Think about that for a moment.

Weather science has made amazing progress the last ten years. My congratulations to the National Weather Service, my colleagues at AccuWeather, and to all meteorologists. 

Addition: More video of three strong tornadoes in southwest Oklahoma yesterday:

How Did the Plains Rain Storm Turn Out?

Rainfall totals since Thursday:

If you would like to compare what actually happened to what was forecast, you can do so.

Here is what I posted ten days before the rains began.

Seven days out.

Five days before the rain began.

Just over three days before the rain started.

And, finally, two days out.

I created all of these links because I want to give our readers the opportunity to compare weather forecasts to what actually occurs so that they can get a feel for the level of accuracy that is reasonable to expect. This situation was more “forecastable” than most. For reasons atmospheric science does not understand, there are times when the weather is easier to predict than others.

Good Work, Senator Cantwell

This one slipped past me when it was first announced. I’ll let her take the deserved credit:

SEATTLE, WA –U.S. Senator Maria Cantwell (D-WA) announced that Washington state’s first coastal Doppler radar, located in Grays Harbor county, has been plugged in for the first time and is now sending test weather data to the National Weather Service’s Weather Forecast Offices located in Seattle and Portland. The state-of-the-art Doppler radar is undergoing testing this month and then will be one of the first in the nation to be upgraded with the latest enhancement to radar technology in civilian weather forecasting, called dual polarization. 
“With Washington state’s first coastal Doppler radar now online for testing, we are on the final home stretch to improved detection and monitoring of storms over southwest Washington,” said Senator Cantwell. “Too often in the past, our weather radar coverage gap meant that forecasters didn’t have the most complete data set possible to help Pacific Northwest communities prepare for big storms. This new, state-of-the-art radar technology will enable Washingtonians to better prepare for the impact of the big Pacific storms on businesses and homes.”
Western Washington’s only other Doppler radar is located on Camano Island, but the radar’s reach is largely blocked by the Olympic Mountains, causing large gaps in weather data of storms approaching the Washington coast. The new coastal radar will help close this data gap, enabling forecasters to better determine wind speed and rainfall of incoming storms to give more accurate and timely warnings to residents in harm’s way and help prevent loss of life and billions of dollars in property damage.
New western Washington Doppler radar
Here is data from the new radar:

Sunday Funnies: One Hit and One Miss

Two of the Sunday comics I saw today are worthy of comment:

As a father who happens to be a scientist, I certainly identify with the frustrations of trying explain science to my children.

click to enlarge

The Wizard whiffs with this effort

I’ll put the track record of weather forecasting up against financial forecasting any day. Hmmmm. Someone should write a book about that. 

Wrap-Up on Irene

I believe it is now clear to everyone that Irene was not “overhyped” by weather forecasters,

Associated Press photo of what used to be Route 4 in Vermont

with approximately 6,000,000 still without power 4-5 days (depending on location) after the storm and communities still cut-off from the outside world, I want to close out Meteorological Musing’s coverage of Irene with a remarkable time-lapse of Irene with  the National Weather Service’s path forecast superimposed.

Hat tip: Andrea Bleistein

While they were not perfect, I’m very proud of my colleagues at AccuWeather and the National Weather Service for superb forecasts of this dangerous storm that unquestionably saved lives and dollars. The meteorological profession came through again.

I’d now like to talk just a moment about preparedness. The Wall Street Journal has an online story just posted about “slow pace” of the recovery.

Political leaders encountered frustrated residents in the northeast Wednesday, angered by days without power, continued flooding and what they perceived to be a slow government response to Hurricane Irene’s devastation.
More than 1.8 million homes and businesses from Virginia to Vermont remained in the dark—with some people told they may not have power for days…

Standing with her school-age son and daughter beside her, Andrea Trout said she had Type 2 diabetes and was struggling to keep her insulin at the proper temperature because her refrigerator lost power. “I’m feeling afraid,” she said. She has been in the dark since 7 a.m. Sunday.

I absolutely, totally feel for these people. But, rather than complain to politicians (which is the topic of the story), I would like to suggest to readers that it is wise to prepare for future disasters and plan to be self-sufficient for at least a week whether it is an earthquake, hurricane, or ice storm.

Here is a set of links for planning for a disaster.

Do it. Tomorrow.

This is Worth a Separate Posting

For those that believe there was too much “hype” about Irene (I do not have an opinion, my head was deep in the data and I didn’t watch any coverage until landfall Sunday morning), consider these preliminary numbers from the New York Times. These statistics are from 1980 (when full-time high resolution satellite data was first becoming available) to present:

  • Damage?  First estimates are $14 billion; ranking 8th
  • Fatalities? Tied with Fran for 10th with 21, but Irene’s are expected to rise. UPDATE: According to the WSJ, 38 fatalities are now reported as of 10pm CDT Monday. As of 9am Tuesday, the death toll stands at 40. Irene is now #4 in terms of fatalities since 1980.
Average the two and you get the 9th 6th worst hurricane in 31 years.  
And, there is this when you are thinking about forecasts of hurricane intensity,
Imagine, for instance, if Irene had been about 20 percent stronger when it hit New York — that it had wind speeds of about 90 miles an hour instead of 75 miles an hour. That doesn’t sound like a huge difference and from a meteorological perspective, and it isn’t.
But from an economic perspective, that may have mattered quite a lot. Some of the scholarly literature suggests that the economic damage resulting from hurricanes is a function of wind speeds raised to the eighth power. I’ll spare you the math: what that means is that hurricane with wind speeds of 90 miles an hour might be as much a 4 or 5 times more destructive as one with wind speeds of 75 miles per hour. So if Irene had been just a bit stronger, we might be talking about economic losses on the order of $55 billion to $70 billion, rather than a “mere” $14 billion.
Because wind power (the wind’s destructive force) is not linear, the forecasts showing winds of 95 to 110 mph near the storm’s center (that we were forecasting would move inland farther west than shown in this computer model forecast) had the potential to cause major, major damage.

Given that Irene is already the 10th 4th worst in terms of fatalities at Cat. 1 intensity, those stronger winds would have threatened lives to a far greater extent. This was in the back of our minds as we were making those forecasts of Cat. 2 intensity on Thursday. 

Irene Recap: How Good Were the Forecasts?

So, how did the meteorological profession do with Irene, a rare hurricane that made landfall in the Northeast United States? Meteorologists almost always do postmortems on major storms because we need to learn from our mistakes so we can do it better next time.

By definition, it is difficult to forecast rare events because of the lack of analogues. For example, if there has never been a full-fledged hurricane in Manhattan in the era of skyscrapers, how will we know the level of damage a hurricane might cause?
So, with that in mind, let’s go over some of the forecasts made on this blog and compare them to reality. My colleagues at AccuWeather and I made some excellent proprietary forecasts that were sometimes ahead of those available from government sources, which is what our business clients pay for. However, because I do not post those forecasts on this blog, I’m doing an analysis of the public forecasts I posted here.
The early forecasts of Irene, made last Sunday (21st) were not very good. They took the storm toward Florida and had the intensity too low. On the other hand, the title of that posting was “Why We Can’t Get Too Confident on Hurricane Paths 5-Days (or Longer) Out.”
The first forecast in which I expressed confidence was titled, “This Looks Correct to Me: Irene’s Path.” I wrote, The eye (which is what is being forecast, not the geographic extent of the storm) could make landfall anywhere in the stippled area.
Arrow indicates the point of landfall as determined by the
National Hurricane Center.

The intensity was overforecast at this point. We were forecasting a category 3 and it was a cat 2 when it made landfall.

Tuesday at 7:19am (roughly four days before landfall), the point of landfall in North Carolina, while slightly far west, was extraordinarily good for four days out. There was still an over forecast by one category but that isn’t bad at all given it is four days out.

Why did the forecast improve? Because the National Weather Service started launching weather balloons at six-hour intervals and NWS and USAF hurricane hunter and high-level aircraft began gathering special data to load into the computer forecast models.

Shortly after 10pm, I forecast posted the first of a series of genuinely amazing path forecasts.

Things didn’t change much on Wednesday. I wouldn’t change very much of what I posted on Tuesday and Wednesday.

Thursday at noon I posted the two sets of forecasts I’m not proud of. Titled, This Looks Serious, and,  Irene: My Concern is Rising,  I walked readers through some of the newer computer models that were taking the intensity of Irene up to, perhaps, a category 4. If this had occurred, the results would have been devastating. Irene did intensify aloft but for scientifically unknown reasons the intensification never made it to the ground. Meteorologists just do not have a solid understanding of the processes which cause tropical systems to intensify or weaken.

Throughout the week, I warned of extensive power failures due to toppled trees from wind and wet ground. I posted the map below of forecast rainfall which I predicted would lead to “major flooding.”

Below is the rainfall map from Irene. Unfortunately, the colors are not the same but I believe you’ll agree that 3-4 days out (depending on location), the correlation between the forecast and actual is remarkable.

From this point on, the path, wind, and rainfall forecasts underwent little revision as confidence went up (i.e., the boundary of the “confidence interval” narrowed).

So, how would I rate the quality of the forecasts presented on this blog?

  • Path, 5 days to landfall and later:  A+
  • Intensity: C- 
  • Forecast of Widespread Power Failures: A
  • Forecast of “Major Flooding:” A+
  • Forecast of Window Breakage in Manhattan High-Rises: F
Please take a look at this posting and feel free to offer comments.  
Tropical depression Twelve may give us some problems next week. There are areas of disturbed weather from Florida to South America that bear watching. We are only about halfway through hurricane season. I suspect Irene will not be the only hurricane we will deal with this year. 
Thanks for reading.

ADDITION: Here is another take on the accuracy of the forecast. And, the New York Times has a good article about the weakening that took Irene from the 2 we forecast just southwest of NYC down to the borderline 1 it actually was.

Interactive map of Irene’s path and damage. Total “customers” (homes and businesses) that lost power are 5.8 million. The electric utility industry uses 3.5 people per “customer” so the total number of people without power is just over 20,000,000.

UPDATE 6:50PM: My friend, meteorologist Dr. Cliff Mass, has a completely different take. He does not believe Irene was even a hurricane north of the North Carolina border. As I understand NHC’s reasoning (and I may be incorrect), there were very high winds aloft and over water and they were concerned those would sustain the hurricane.

I present all of this because I want non-meteorologists to understand the complexities of what we are dealing with when it comes to hurricanes and tropical storms.