Dan's Wild Wild Science Journal

Numerical weather prediction model forecast valid at 8 PM U.S. East Coast time Tuesday. This storm will spread tornadoes across the U.S. Midwest and near hurricane force winds on the Great Lakes.

When most folks think of tornadoes they imagine a warm spring afternoon suddenly turning stormy. More often than not this is true but there are glaring exceptions. Last night was one and Tuesday will be another.

A powerful storm system has been winding up  in the Plains. Last night a band of storms from Texas to Alabama brought tornadoes and large hail. Here in North Alabama, I was up for much of the night watching radar.

Winds going toward the radar right next to winds going away from the radar. This is what a tornado looks like on a Doppler radar. Image from Penn. State NEXRAD Archive. Click for larger version.

Large super-cell tornadoes are the easiest to spot and we had only one of those up in Tennessee. The ones you have to really watch for are the smaller more short lived twisters that are embedded in a squall line.

Around 4 AM CDT Monday morning, the Doppler radars indicated a strong circulation in the line over NE Alabama. The little town of Ider, near Fort Payne, was struck with a twister around 4:15.

The tornado looks to be an EF-1 or perhaps briefly an EF-2. An EF-1 tornado has winds of 32-50 meters per second (73-113 mph). That’s the equivalent of up to a category two hurricane on the Saffir-Simpson scale.

The Storm Prediction Center  (SPC) in Norman Oklahoma issues the tornado watches for all of the U.S. They’ve notified Meteorologists like myself and at the local NWS forecast offices that they believe there is a rare “high risk” of tornadoes over Indiana and Ohio on Tuesday. A high risk is rare anytime.

It’s exceedingly rare in autumn.

SPC does only one thing. Forecast severe weather. Meteorologists like myself give high credibility to their forecasts.

The Enhanced Fujita Scale- courtesy of the Storm Prediction Center (NOAA)

There’s been  an amazing technological revolution in forecasting over the last 30 years. In the 1970′s most tornado watches had no reports of tornadoes. Now it is rare for a watch not to verify.

When I was an undergrad meteorology student, I worked on a project at Okla. University in 1980 called SESAME. That stands for Severe Environmental Storms and Mesoscale Experiment. It was a fancy name for trying to correlate what severe storms were doing with what the new Doppler radars were indicating.

I remember being laughed at by people who called it a waste of money. It was anything but.

Doppler radars cover most of the nation now and make it possible for forecasters to give incredibly accurate warnings. An accuaracy I could not have imagined back in 1980.

At 4:15am Monday, the Doppler radar here in North Alabama showed a srong rotation and a tornado warning was issued. We had it on the air in less than 15 seconds from the time the NWS pushed the button. Unfortunately, most people in the little town of Ider were asleep.

The ones who has NOAA weather radios were awake because a loud alarm had gone off. If you live in area where severe weather is likely, you should have one. They only cost about 30$.

It might save your life one night.

You cannot imagine how frustrating it is to break into programming and give a warning while knowing that most people are asleep and will never hear it.

Until it’s too late that is.

Be safe,

Dan

Areas closed to fishing in the Gulf. S.A.R. images from CSTARS at the Univ. of Florida Miami show oil is now in areas still open.

There are several ways of seeing the extent of the oil slick on the water in the Gulf of Mexico. Visible light is just one way. A better way may very well be to use Synthetic Aperture Radar and look at the Gulf in high frequency radio light.

Image from CSTARS. The red line is the approximate border of the area closed to fishing by Federal Government.

The SAR image to the right is courtesy of CSTARS and the European Space Agency.  It shows the oil well.  It also shows the oil has spread to the east of the federal closed to fishing area.

The closed area will likely have to be expanded further eastward and soon. The current closed area is left of a line southward from Pensacola Bay.

There is an SAR on a Canadian Satellite as well and it too shows the oil quite well.

Something to keep in mind, we humans can only see a tiny part of the electromagnetic spectrum.

We call it light.

IR, and C band radio waves are light too though and sometimes you can see things much better in that light than the visible light our eyes have evolved to detect.

NOTE Saturday 5 June: NOAA has now expanded the area closed to fishing and the area mentioned in this post is now included.

Dan's photo.

The EF 3 tornado that hit Albertville developed very rapidly. It first touched down just west of the city. It was on the ground for about 30 minutes and lifted near Geraldine in Dekalb County. We were able to give nearly 30 minutes warning for the folks in Geraldine. Albertville had 7 to 8 minutes.

That may not sound like much, but for a town at the beginning of the track, that is actually very good. Especially in this case.

Look at the radar images.

Velocity data shows a weak and ragged circulation, at 10:07 PM, as the storm enters Marshall County.

At 10:07 PM the velocity data showed a ragged circulation in the storm as it entered Marshall County Alabama. It takes the radar about 6 minutes to do a complete scan at several elevations. This is called a volume scan.

One volume scan later at 10:14 PM a strong circulation has developed. The NWS office in Huntsville issues a Tornado Warning. We had been keeping a weary eye on this cell.  I was doing the 10 PM  weather and we immediately urged the folks in Albertville to “take cover right now!”.

One volume scan of the radar later- a strong circulation is present. Tornado Warning being issued.

At 10:22 pm the radar data shows a very intense circulation with winds over 100 knots. The circulation is very near Albertville.

10:21 PM: An intense rotation is near Albertville. The tornado struck at 10:22 PM. (Pink is high winds away from the radar, and cyan is strong wind toward the radar. Radar located at top of image.

I talked with an Albertville city councilman who is also a state trooper.  He was off duty, but looked at his watch as he saw the tornado lift a giant tree out of the ground.

It was 10:22pm.

Albertville had around 7-8 minutes of warning.

Much more than most cities right at the beginning of a track will usually get.

This is why you should have a NOAA weather radio. Even if you have the TV on, you may be in another room or preoccupied. That warning siren might give you enough time to get to shelter. If you are very close to where the tornado is developing, you may only have seconds.

When it comes to tornado warnings, 7 minutes is a lifetime to a severe weather nowcaster.

A SHORT HISTORY FIRST

Weather radar is now a common site on any TV weathercast, and radar images are all over the Internet.

It wasn’t really meant to be that way though….

Hurricane Carla in September 1961. The eye is visible on the Galveston WSR 57 radar. This was the first hurricane on radar that TV viewers had seen.

Apparently, the first weather radar image of a dangerous storm shown on TV live was back in 1961. A Houston TV station sent a young reporter to Galveston to cover the approach of Hurricane Carla. It was a mean category 4 storm and heading right toward the city. The reporter showed up at the local weather bureau office and saw an image of the hurricane on the newly installed radar.The Galveston WSR 57 radar had been installed just the year before.

It was a scary picture and he wanted to show it on air. This was unheard of at the time. It had never been done before!

This request required permission from the Weather Bureau headquarters in Washington. It was supposedly given very reluctantly, but it had the desired effect. Most people got out of town before Carla stormed ashore. That young TV news guy from Houston was named Dan Rather by the way. Betcha didn’t know that!

It was not much of a radar by today’s standards. It had no Doppler capability, and was built on World War 2 technology. They were the main weather radar used by the National Weather Service  (Successor to the Weather Bureau), until the last one was decommissioned in 1996.

The replacements are called WSR 88D. This stands for Weather Service Radar 1988 Doppler. It was, however, not up and running in most places until the mid 1990′s.

DOPPLER RADAR

Doppler Radar image of winds showing the Greensburg Kansas EF5 tornado. The blue colour near Greensburg is strong wind blowing toward the radar. The Pink is very strong wind going away from the radar. The tornado wiped out the town of Greensburg.

Doppler Radar has the ability to detect rain drops (and anything else in the beam) moving toward or away from the radar. This is based on the Doppler effect. It’s the same affect that you hear when a train passes by and the tone or frequency of the horn changes. This effect happens with ALL electromagnetic radiation as well. This includes light and radio waves (Which are basically the same thing!).

Astronomers use it to determine how fast stars that are millions of light years away are moving away from us. It’s called the red shift, because the light from objects moving away is shifted toward the red light end of the spectrum.

Is this real rain on the radar? NO! Read on! (NWS Jacksonville WSR88D)

Tornadoes are spotted on Doppler radar when rain drops are rotated around an intense low pressure in a thunderstorm. This intense low is called a meso (small scale) cyclone. These mesocyclones are the parent circulation of a tornado. Only a few mesocyclone produce tornadoes, and for this reason, false alarm rates on tornado warnings remain unacceptably high. Keep in mind the radar can only detect the motion of objects moving toward or away from it. Not side to side.

This same technology is very possibly responsible for your last speeding ticket!

THINGS YOU SHOULD REMEMBER

Radars do not see rain!

They just detect radiation that is reflected back as the narrow beam travels outward. Bugs, dust, hail stones, birds and even clouds reflect the beam as well! Weather radars operate mainly at a wave length between 3 and 10cm. This minimises reflections from objects that are much bigger, or smaller than rain drops. A chunk of ice with a coating of water is a very good reflector at these wavelengths, and many times what you may think is heavy rain is actually hail.

Rain intensity is measured in dbz which is based on the power returned to the radar. 20 dbz is a very light rain. 40 dbz is very heavy. 50 dbz is blinding rain or may be hail. Most online or on TV radars use red and yellow for the higher values. These values are vaild only if the returns on the radar are from rain drops!

Radar Beams Don’t Travel long The Ground!

Radar beams get higher as they go away from the radar dish. This can cause storms to not show up at all! Image from NOAA

The radar beam travels in a basically straight line away from the radar. It is usually pointed at 0.5 to 2 degrees above the horizon. Since the Earth is curved, a radar beam is usually nearly  25 km high when it hits an object 500 km away! This is above all but the most violent thunderstorms. Tornadoes are low level phenomenon.  A radar is unlikely to detect a circulation around one more than 150 km away. Even 100 km is really stretching it.

Radar Beams Sometimes Hit The Ground!

Temperature inversions can cause the radar beam to hit the ground (NOAA image)

On clear, calm nights, a strong temperature inversion may develop. This means the air warms with height. These inversions cause a density difference in the atmosphere that bends the beam to the ground. The radar may look like it is detecting strong storms. It’s just dirt!

Meteorologists call this Anomalous Propagation. Our top news anchor at the station keeps telling me that he had a case of that once, but the doctor gave him a cream that cleared it right up!

The Newest Technology

The TV station I work for was the first in the world to install a Dual Polarimetric Radar. The NWS is upgrading most of their Doppler radars to this technology as well. The reason: The ability to determine the shape of objects in the beam.

Image from our ARMOR radar. Click image for live view.

This has fabulous possibilities. We have even detected debris from a large tornado on our radar. We call it ARMOR for Advanced Radar for Meteorological Operations and Research. The University of Alabama at Huntsville and WHNT developed ARMOR jointly. I use it for forecasting and on air, and they are doing some incredible research science with it.

ONE LAST THING

A line of storms or an area of rain will change as it moves along. Do not assume that a storm 100 km away will be as strong, or even still exist when it reaches you. Remember also that the further away it is, the higher in the storm you are looking. The rainfall pattern may look much different near the ground. Radar is a remote sensing tool. It does great things, but it’s only detecting radio waves. No more, and no less!

Later,

Dan

Good Friday 2009 will long be remembered in Alabama and Tennessee. Huge hailstones, and violent tornadoes pounded the region. Only two deaths, and some 50 injuries, but a lot of destruction. It certainly could have been worse, and to families that lost loved ones and homes, it is a tragedy that will be remembered for a life time.

Many times, when severe storms hit, the only view of hail or tornadoes for most people, is on TV. Not this time. The news room was inundated with calls from viewers, watching large hail stones dent their car, and ruin their roofs. Local car dealers here have over 300 damaged cars.

Modern technology is still not good enough to give reliable advance warning of small EF 0 and EF 1 tornadoes. (For those reading from other countries, the U.S. rates tornadoes on the Enhanced Fujita scale. This scale runs from zero to five. Five is very bad.) The large killer tornadoes are forecasted much better now than in the past. I warned viewers Thursday night that we faced such a threat, and showed on a map, where the threat was highest. We made the warning the lead story in our newscast. This kind of detailed forecast was impossible 25 years ago. Science has learned a lot.

Everyone in the path of the tornadoes had several minutes of warning Friday. I would like to think that this explains the low numbers of deaths, and injuries.

Looks like "Cricket ball" size to me.

We have had hundreds of images of the hail and the storms, along with a bunch of questions about the event directed to the Meteorologist. Since that’s me, I spent the day refreshing my memory on the science of hail, and wind. What follows are some answers. I have also written to one of the top experts in the field of hail research, so another post is possible soon.

A lot of the hail on Friday was the size of golf balls, but many reports of tennis ball, and even softball sized hail were reported. Viewers sent me pictures to back up the reports. They sent a LOT of pictures. These images allowed forecasters like me to better predict the size of the hail, in areas further East.

The super-cell storms are constantly being measured, as the tracked across Alabama and Tennessee. One of these measurements is called VIL. It stands for Vertically Integrated Liquid. The VIL is an estimate of the total liquid water in column through the storm.

If you divide the VIL of a storm, by the height of the storm, you get a number called the VIL density. This number tends to be a good measure of hail size. It’s not perfect, and there are other factors that can affect how large the hailstones that reach the ground are. One number we look at is called the “Wet Bulb Zero” height. It tells us how close to the ground the hail stones will start to melt. Detailed info on VIL, and VIL density is available for my fellow nerds here.

When viewers start sending me pictures of golf ball sized hail, and I know the storm has a VIL density of say 4, then I can be quite certain that other storms that day with a vil density of 4 will have larger than golf ball sized hail. These images are very helpful indeed, but I worry that people will get hurt in their attempt to take photos of the hail. There was an injury Friday. A woman was hit in the head with a large hail stone, while running to a car, to get out of the storm.

Coins are a good way of communicating hail size. avoid using the term "marble size". I don't know how big your marbles are! You may have even lost them!

Some of the hail was clear ice, instead of opaque. This clear ice is likely related to the freezing rate, and the amount of supercooled water in the highest reaches of the super-cell thunderstorm. Slower freezing, and slower hail growth, gives clear ice. The amount of supercooled water around the hail stone seems to play a big roll in the determination of clear vs. opaque.

Another viewer asked about the jagged hailstones. These are likely hail stones that are made up of several chunks of ice that froze together. A large stone like this, can only be thrown upward, and held aloft, by an updraft of around 160 km/hr! Stones this size are rarely seen and when they are, a violent tornado is usually close by. The exception is Colorado, where stones this large can form with no tornado. So, that rule is mainly for areas East of the High Plains.

The Hail Friday was Quite Varied In Shape/Size

Several images of hail that was “egg” shaped were sent in. I am not certain how this happens. It may very well be related to the intense low pressure circulation inside the tornadic thunderstorm. Centrifugal force perhaps? You can read more about hail on the UCAR Fact Sheet.

Meteorologists call this the mesocyclone, and Doppler weather radars can detect the strength of the mesocyclone. The mesocyclone near the largest hail on Friday was very intense. So intense, that I was quite certain, a tornado was on the ground, and doing damage. Storms like these we tracked on radar Friday, are seen rarely. In my 30 years of forecasting, I have tracked storms like that on maybe 6 or 7 days.

Wall clouds are very popular with the public. Storm spotter courses show lots of pictures of them, and when we have strong storms, I get many photos from viewers of what looks to them like a wall cloud. I always tell our reporters in the field, that a wall cloud will look just like a wall cloud. If there is doubt, it probably is not one. Wall clouds are the parent cloud of a tornado, and a rotating wall cloud is a sign that you may be about to see a tornado. Best not be in a location ahead of one!

Wall clouds will form under the "rain free base" of a supercell.

Sometimes a pendant will slope down, from the wall cloud toward the rain shaft. This is called a tail cloud. Usually, rotation in the wall cloud is quite apparent.

The tornado that tracked through Marshall, Dekalb, and Jackson counties on Friday has been rated as an EF-3 on the enhanced Fujta scale. This means winds of over 150 mph. The maximum width of the path of damage is nearly 1KM!

The estimate of the tornado’s size, and strength is based on several factors, The width of the path, and it’s length. The maximum winds can be estimated by the damage to trees, and structures. It is important to realize that if a large tornado does no damage, then it cannot be given a rating.

Many times, people will insist that a tornado has been under rated because there are many destroyed homes. If these homes are mobile homes/trailers, then an EF2 tornado will completely destroy them. If there is no major damage to wooden homes, then the tornado rating is likely correct. An EF2 will also usually destroy a garage or out buildings as well.

The EF3 tornado is much stronger, and does much more damage to wooden/ brick structures, on a foundation. An EF2 tornado has winds up to 115 knots or round 219 km/hr. An EF3 has winds from 219- 265 km/hr. That is nearly 160 knots. A category 5 hurricane is another example.

Typical EF3 Damage (From NOAA-NWS)

Heavy cars will be thrown a good distance by an EF3 and walls of well built structures will collapse. Damage in the Sylvania area, of Dekalb County, indicates that the tornado Friday was at EF3 strength. Late word from the NWS survey team tonight, of one home that was picked up, and slammed 4 feet into the dirt in Sylvania. Read the preliminary storm survey here.

The methods of rating a tornado is explained in this slide show. If everyone who lived in a mobile home, left for safer shelter during a tornado watch, the death toll from tornadoes would drop by half in this country. Don’t think that tie downs make any difference. They almost certainly do not.

Thanks to all of our viewers here in Alabama and Tennessee whose pictures and reports helped us track the storms on Friday. If you see your picture in this post-please email me and I will add a credit. We have so many- the names have been separated!

UPDATE SUNDAY EVENING: Donnie Charles of Albertville emailed us this picture. It’s a GIANT wall cloud. Taken in 2006, but a good example.

Huge Wall Cloud Near Albertville, Al 2006.

Notice the striations. Wall clouds like that almost always accompany killer tornadoes.

Now the bad news. More storms Monday.

Keep safe,

Dan