Weather For Kossacks
Weather For Kossacks is a new regular series posted Wednesday evenings aimed at helping readers better understand the processes that make up the weather. Topics planned include Doppler radar (this diary), formation/types of thunderstorms, severe weather, fronts, low and high pressure, hurricanes, droughts, and temperature extremes.
If you have any suggestions, comments, or angry hate mail, you can also email me at firstname.lastname@example.org.
One of the most widely used tools in meteorology is the Doppler weather radar. We'd be lost without it. Imagine having to live through a tornado outbreak without Doppler radar telling you where the tornadoes are, right down to the neighborhoods that they're tearing up?
The implementation and study of weather radar technology exploded in the 1950s, and the study of severe weather grew as well. The hook heard 'round the world was measured in Illinois on April 9, 1953, when a research radar antenna detected something called a "hook echo" on radar. This was the first time a tornadic thunderstorm was ever detected on weather radar.
This research led to the development of the first generation of weather radar used by the National Weather Service. It was called WSR-57, short for Weather Service Radar 1957. They were installed in 31 major cities across the United States, and were responsible for saving countless lives with the advanced warning they were able to provide.
The current radar system is called NEXRAD, or Next Generation Radar. These radars are advanced in that they are of a higher resolution than WSR-57, and they are Doppler radars (allowing the radar to detect velocities/wind). The technical name is WSR-88, and 161 of these radar systems were installed across the United States between 1992 and 1997. A 162nd weather radar was installed on the Washington Coast in 2011 after an intense lobbying campaign by Dr. Cliff Mass and Senator Maria Cantwell.
If you've ever flown out of a reasonably sized airport in the United States, you've probably seen what looks like a big golfball sitting on a tall metal platform on one of the outer peripheries of the airport. That's the weather radar. The actual weather radar is a microwave antenna looks like a giant satellite dish, and it sits inside of a large dome that serves to protect the antenna from the elements.
The actual detailed workings of Doppler weather radar are intensely complicated, involving a ton of math and physics. Thankfully, the basic reasoning is pretty easy to understand.
The antenna inside the radome sends out numerous slightly angled beams 360° around the site to measure precipitation as it falls. When the beam strikes an object (rain, snow, hail, even birds and planes), it bounces back to the radar, and the computer measures how strong the beam was when it came back. The stronger the return beam, the heavier the precipitation. This is called "reflectivity." Whenever you hear reflectivity, think precipitation.
The radar has different levels at which it sends out the beam. These are called "tilts." The tilts range anywhere from 0.5° up to 19.5°. These may not seem like big angles, but the radar beam goes out almost 250 miles away from the radar site. Given the curvature of the Earth, the lowest radar tilt (0.5°) can wind up 53,000 feet above ground level by the time it gets a few hundred miles away from the radar site.
The number of radar tilts a particular NWS office uses is based on what their current weather is. If it's calm, they'll only have the radar slowly scan a few tilts (mainly to conserve power). If there are massive storms in the area, they'll have all the tilts turn on and have the radar sweep them as fast as it can.
There are two levels of NEXRAD radar data -- Level II and Level III. Level III has more data included in it, but it has less resolution. Level II has less data in it, but it's of a higher resolution. Level II data is what's called "SuperRes." Each radar pixel in Level III data covers about half a mile. Each radar pixel in Level II data covers about a tenth of a mile. Since Level II data has almost 5x more resolution than Level III data, it's ideal for accurately viewing individual features in a storm, such as tornadic rotation.
See the difference?
For the casual user, there are 5 major radar products: base reflectivity, base velocity, storm relative velocity, composite reflectivity, and radar estimated rainfall.
Base reflectivity is the most common radar product. "Base" refers to the lowest radar beam sweep, which is usually only a 0.5° tilt. This tilt gives the lowest parts of any thunderstorms in the vicinity of the radar site, and gives you the best idea of what is happening closest to the ground. Base reflectivity shows any number of objects -- rain, sleet, snow, hail, birds, airplanes, buildings, mountain tops, and during destructive tornadoes you can even see the debris rolling around in the storm.
Generally, reflectivity goes from blue to white on the scale. For the most part, blue and green indicate light rain. Yellow and orange indicate moderate rain. Red indicates heavy rain, and purple/white is either extremely heavy rain, large hail, or debris from a tornado.
Here's an example of base reflectivity data from the March 2, 2012 tornado outbreak across the Ohio Valley. At the time of this image, an EF-3 tornado was on the ground about to decimate parts of Salyersville, KY. I'll explain this image in more detail in a future Weather For Kossacks diary.
Base velocity is the second most common radar product. Velocity data measures how fast precipitation is moving around within a storm -- in other words, the winds. Velocity data is crucial in detecting tornadoes and severe winds within a thunderstorm. Velocities are displayed in two colors -- usually red and green. Green generally indicates winds moving towards the radar site, and red generally indicates winds moving away from the radar site. Darker greens/reds generally show lighter winds, whereas really bright greens/reds show strong winds. When the bright green and bright red come in close contact with each other, and it's within a thunderstorm, it indicates strong rotation and the possibility of a tornado.
There's also something called "range folding" which usually shows up as purple shading in the base velocity image. The way I understand range folding is that it occurs when the radar detects a precipitation in the right direction (say, southeast), but it can't resolve how far away it is for some reason. This discrepancy causes an error in the radar's computer, and it shows up as "range folding."
The following image is a split screen between base reflectivity (left) and base velocity (right) from when Hurricane Dennis was approaching landfall in July of 2005. Intense hurricanes (like Dennis) are a great way to visualize how radar velocities work.
STORM RELATIVE VELOCITY
Base velocity measures the speed of the winds inside of a thunderstorm. However, this product measures all winds -- accounting for just the speed of the winds. Storm relative velocity takes into account the speed and direction a storm is moving.
If a storm is moving east at 30 MPH, and it has 20 MPH winds, base velocity will show the storm having 50 MPH winds. However, if you tell the radar program that the storm is moving east at 30 MPH, the storm relative velocity will only show the winds moving 20 MPH.
It's useful if you want to remove the effects of the storm's motion from the velocity/wind speed output.
The composite reflectivity image shows all the radar tilts combined in one image. Say, the radar scans the atmosphere at 0.5°, 0.9°, 1.2°, 2.0°, 3.5°, and 5.0°. The radar will produce 6 individual images showing each slice of the atmosphere that it scanned. The composite reflectivity image combines all 6 images into 1 image.
It's useful for detecting virga, or precipitation that falls from a storm but evaporates before it reaches the ground. Some sleazy weather geeks also use it to impress people because it makes storms look more intense than they really are.
RADAR ESTIMATED RAINFALL
In Level III radar data, there are several products related to radar estimated rainfall: One Hour Rainfall, Three Hour Rainfall, and Storm Total Rainfall. They're all estimated based on complex algorithms built into the computers at the NWS that fairly accurately estimate how much rain has fallen in a given time span based on how strong the reflectivity was over a certain area.
This image is the radar estimated storm total rainfall from the Dothan, AL area, which received upwards of 5-10+ inches of rain in the 3 days between August 5 and August 7, 2012.
The great thing about having multiple radar tilts is that, when you combine them all together, you can get a comprehensive look inside of the structure of a thunderstorm. You can get a 2D or 3D look at it depending on what kind of software you use.
A cross-section is a slice through a thunderstorm on the radar, and it shows you the inside of the thunderstorm's structure in ways that you wouldn't have been able to see. It's extremely useful in pinpointing hail cores (hail suspended in the storm), updrafts, downdrafts, severe winds, and debris in tornadoes.
To show an example of a cross-section, we'll look at the devastating EF-5 tornado that destroyed much of southern Joplin, MO on May 22, 2011. This is what the base reflectivity image looks like, the white line depicting the cross-section we'll look at:
And here's the aforementioned cross-section:
The tornado was so powerful and destroyed so much stuff that it was a mile wide and 18,000 feet deep within the thunderstorm. Without even seeing reports from the ground, you know that such dense debris being thrown 3.5 miles into the atmosphere has to be doing incredible amounts of damage.
Another great thing about having many radar tilts is that specially designed software can make a 3D rendering of the data, and show you a pretty accurate model of what a thunderstorm looks like at the time of the radar scan. It combines all the radar sweeps into one image and turns it into a 3D model that shows almost every little bump and detail of the storm. It's awesome, and again, really useful for spotting tornadoes and hail cores within severe storms.
Here's an example from the April 27, 2011 tornado that tore through Tuscaloosa and Birmingham. Keep in mind that you're not seeing the tornado itself, but rather the massive amount of debris swirling around inside of it:
Here's another example from the thunderstorm near Dallas, TX that I mentioned in the "Level II vs. Level III" section. It shows the storm in beautiful detail, right down to the overshooting top and the anvil.
The solution that engineers came up with is to add a vertical beam to the radar. This upgrade gives meteorologists a whole new way of being able to see inside a thunderstorm. One of the big uses for the dual polarization upgrade is the "hydrometeor classification." The added dimension to the radar beam allows the radar's computer to detect what kind of objects the radar is detecting.
The upgrade allows the radar to detect the difference between the following objects:
-Biological (flocks of birds and/or bugs)
This will help meteorologists tremendously in determining what's going on inside of an area of precipitation or a thunderstorm, so they can get out more accurate and timely warnings.
There are some other new products in the dual polarization upgrade, but they're way beyond my pay grade. If you'd like to learn more about dual polarization, the National Weather Service has a great outreach website devoted to education folks about this new technology.
In the next decade or two, the WSR-88 "NEXRAD" radar system we have in the United States will begin to be replaced by something called phased array. The National Severe Storms Laboratory (NSSL) explains phased array much better than I could:
Current WSR-88D radars transmit one beam of energy at a time, listen for the returned energy, then mechanically tilts up a little higher, and samples another small section of the atmosphere. When it has sampled the entire volume of atmosphere, from bottom to top at a particular location, the radar goes back down, moves over a little, and starts the process over again. This continues until the radar has scanned the entire atmosphere, which takes around six or seven minutes. Phased arrays use multiple beams, sent out at one time, so the antennas never need to tilt. Scanning takes only 30 seconds, and it already has dual-polarization capabilities.As the blockquote states, we have to wait 6 to 7 minutes for all the radar sweeps to end and the fresh images to be uploaded to the internet. With phased array, we'd have to wait just a minute or two tops. That's mind boggling. It would cut down warning times and give us about as close to a realtime view of storms as we could get.
If you have money to blow and you're a hardcore weather geek, you can get one of the many, many weather programs that are available online for a fee. My favorite programs are the GRLevelX series of radar programs. GRLevel2Analyst is the best one if you want an in-depth look at the weather (with the super resolution data, cross-sections, and 3D rendering).
If you live on your cell phone (or have a Mac), RadarScope is the best way to go if you've got 10 bucks laying around. For the price of a medium pizza you can have a solid radar program that could save your life one day.
If you don't want to spend money on a radar program (hey, I don't blame you), Wunderground offers a pretty good site to view radar data.
Hope this helps. If you have any questions, please ask. Next week's Weather For Kossacks will cover heat waves, cold snaps, and why the weather changes from season to season (as simple as it may seem).