 Testing. Can everybody hear me? Okay. Well, welcome to the beautiful Anchor-Mesa lab, and thank you for sharing your Saturday afternoon with me. So let's get to the talk. So weather radar has been used as a tool to understand and solve weather problems. So for a radar scientist were interested in these topics, and the scientific curiosity often translates into some useful numbers. For example, when will the storm start? When will it rain? How much will it rain? And how strong will the wind blows? These numbers hopefully will improve the convenience and planning of daily life and disaster prevention. If you recall last June, a few days National Weather Service predicts severe thunderstorms, and before you went out for dinner, you probably check your cell phone for the most recent radar loops to see if your lovely car will be in the damaged path or approaching hail storm. And as a radar meteorologist, I forgot to do that, and I end up with a insurance check. Okay, so I want to give you a quote to put the radar meteorology in some perspective. So this is a quote from a radar textbook, an ideal radar meteorologist, therefore should be an electrical engineer, a mathematician, a computer science meteorologist, a cloud physicist, and a hydrologist. In one person. It never happened, but really. So this quote basically put the context into radar meteorology as a interdisciplinary subject. And I hope I'll share some of these interdisciplinary diversity in this talk. Now, one of the real pioneers of U.S. radar radar, Dr. David Alice. In 1991, he received the American Meteorological Society Award for remote sensing lecture. It's actually a war called remote sensing lecture. So he opened his award speech by this paragraph. I'll read it. It transmits high frequency pulses with amplitude and frequency modulation. And variable pulse length and repetition frequency, it switches on receiver after the high energy transmitted pulses are launched and measure the elapsed time between the transmitted pulse and the echoes, thus determining the distance to the target. It's perfect pitch allows determination of the Doppler shift and the relative speed of the target. And he uses modern match filter techniques to identify the nature of the target. So everybody in the audience thought he's talking about radar. No, it's a bat. So it took a long time, about 32 million years before the human finally figured out that a bat naturally can do through evolution to apply these to a modern radar. So talking about whether actually radar, you have to go all the way back to 1865. 1865 Maxwell equation. First describe how electromagnetic wave can be formed. In 1904, a German inventor and engineer, Christian Hirschmeyer, I hope I said it right, developed the first radar on record. He actually invented a device to detect ships in a harbor in dense fog. So when a ship approaching where the device was, a bell will ring. So that's the first radar on record. In the 1920s, the heart of a radar called magnetron was invented. Simple and affordable version of magnetron. It's probably in your kitchen. It's called microwave. You eat your lunch every day. Then the World War II broke. Radar becomes a hot topic because both the allies and the Axis trying to manly develop radar techniques to detect enemies, airplane, and missiles. And some historian actually attribute radar as one of the reasons that the allies won the war. Because the US and British has a leg up on the Germans. So in 1941, February 20th, we can trace the birth of a radar meteorology back to that day because one of the rain shower was tracked by a radar for seven miles on that day. At that point, the military radar operator sees those echoes as very annoying because it masked the aircraft. But the radar meteorologist loved those. 1960s is when Doppler radar was was available for meteorological use. In the 1970s, we developed a polymetric radar. Polymetric radar means you transmit both horizontal and vertical polarized waves. And by 2000, we started to explore face array radar technology for meteorological use. On the bottom of the diagram, so 1940s, we have ground based radar radar on the ground. By 1960s, US Air Force used radar on the aircraft to study hurricanes. 1970s, radar was put on ship to do weather experiment. 1994, we first put a radar on the on the truck to drive around US Midwest to chase tornadoes. And by 1997, NASA put a first radar with a radar on satellites. So a radar, what's a radar? Radar is the acronym for radio detection and ranging. This term first coined by US Navy in 1941. So a radar transmit radiation and hits a target and receive a return. So a simple weather radar system consists of four major components. The transmitter, the antenna to focus the beam energy to a certain direction, then collect the information back. And the receiver and the signal processor extract the meteorological information and display it on an indicator. In the modern days, it's a computer screen. So the return from the radar, the simplest parameter is called radar reflectivity factor. If you watch the nightly news of the weather report, a weather man will recall forecast meteorologists will probably point to a screen and say there's a 60 dBz echo somewhere and someone residents need to watch out. So what is what's a reflected factor mean? It's a scale that's proportional to the target size. Okay, the size of the drops, rain drops. And I put a note here, six power. So it's really weighted very highly with a target size. Also a number of targets in the unit volume. So how many rain drops in a specific volume and how large of the rain drop really determined the radar reflectivity factor. So I'll give you an example. I mentioned 60 dBz echo. It can it can be a one, a single one centimeter diameter drop in a one cubic meter volume. Or it can come from one million, 0.1 centimeter drops in the same volume. So this nature created some problem to use radar reflectivity to derive rainfall rate. But this example, you live in Colorado at the beginning of a Sunday shower, you probably experience that there are just very few big drops dropping. Okay, you can probably dance between those rain drops and without getting wet. That's a 60 dBz echo. If unfortunately, you encounter a lot of smaller drops and you're in trouble, you'll be soaking wet. Let's go to tornado. The first hook echo photograph on the radar display. I believe it's on April 9, 1951 or two. I forgot the exact date. A radar operator photographed these pictures. Okay, yes, back then they're recorded on film and printed on paper. So I want to focus on these area. These are what we call ground clutter, which we all just hate. We mask our echo. So you can see that this is super south under storm and there's a feature here like a hook in your closet. And in a few minutes later, when it's moved away from the clutter, it's much clearer. There's a hook. And later on, the hook actually developed an eye just like a typhoon or hurricane. And this series of photo was captured by this radar called APS-15 right at the Shampai Indianore Airport. Later on, we found, meteorologists found that there was a tornado identified and those are photographed on that day. And when they tried to match the tornado damage with the radar images, they found that a tornado was actually occurred near this spot, the head of the hook. So this discovery paved the way for National Weather Service to use radar returns as a way to warn severe weather. Okay, the microburst, the bow echo, et cetera. The experience started from here. And speaking about tornado, I cannot ignore this gentleman, Ted Fujita. The media call him Mr. Tornado. So Dr. Fujita was a professor at the University of Chicago. And in the late 50s and 60s, he did lots, lots of damage survey after tornado occurred in the Midwest. So he came up with a scale called Fujita scale. Many of you, almost everybody should hear about that. When there's a tornado, the media always say it's F1 from 0 to 5. Okay, in 2007, a modified scale was adopted called the effective Fujita scale. But essentially, they're pretty much the same. So these scales are assigned a wind speed range, which I will not go into detail, but I want to point to that. For F5 tornado, the wind speed is roughly about 89 meters per second. And each scale is actually a damaged scale. So based on the damage of the structure, Dr. Fujita taught the structure engineers. Okay, and show them a picture and say, okay, tell me, what's the wind speed you think will create this damage? So this picture illustrates different type of damages associated with different type of different Fujita scale number. Okay, from EF0, which probably a few tiles of your roof disappeared, to F5 where only foundation of your house was there. So one last note about Fujita scale, I won't remind everybody that this is a damaged scale. Okay, it's not a wind scale. The wind speed was never scientifically demonstrated, because I'm not aware of any any mobbing ever survived a strong tornado at this point. Okay, now Doppler effect. 1842, Christian Doppler published a seminal paper. The title is translated into English. It's on the color light of the binary star and some other stars of the heavens. Basically, he described a binary star in space, why the light, the color of those lights are different, because those stars are rotating with each other. So here is a demonstration of the Doppler effect. A real-life example is when you stop at the real-row cross track. And a train is moving towards you. Then you, in boulder, you hear the engine start to, the horn start to sound. And when the train is approaching you, you hear a higher pitch. And when the engine pass you immediately, you hear the sound, the pitch dropped. So this is a Doppler effect. So meteorologists use Doppler effect to build a Doppler radar. Oh, by the way, if you ever get a traffic ticket, speeding ticket, you're probably a victim of the Doppler radar. Talking about Doppler radar, I have to mention this pioneer, Dr. Roger LeMitt. He was a French scientist and first demonstrated how Doppler effect can be applied to a Doppler radar. So basically, his work set a tone for all the modern Doppler radar in the world. He was recruited to U.S. in 1960, something. It's hard to chase it back on literature. And he actually developed many single and due Doppler wind retrieval algorithms. And he was awarded the first single Doppler wind retrieval technique called VAD Velocity Asthma Display. And the U.S. weather service still use that technique today. So, how do we peer into a storm? How do we know how much wind is blowing inside a storm? So we use weather radar to do it and use a Doppler effect. So a Doppler radar measures one component of the three-dimensional wind velocity. So, you know, you only have one component, you cannot recover three-dimensional wind speed. So in theory, we need three Doppler radars, three or more. Okay, but meteorologists are smart, so they say, okay, I can use two. So here's an example to illustrate that. So you have radar one here. Observe a point in space with this component and radar two point to this direction. So you have two vectors. If you do the magic of vector math, you can recover the horizontal vector at that point. If you do this for every single point, you will have a field of horizontal velocities. Then you can integrate a free equation in atmosphere called continuity equation. What that equation basis says is what air comes into the volume has to go out unless the density in this volume increase or decrease. And there are two hard boundaries in this equation. In this situation, one is the Earth, one is the strong top. So the air cannot go down. So if you have a convergence, the air has to go up. Near the strong top, the air doesn't go past that, so the air goes around it. In the summertime in Colorado, you see this anvil cloud flowing out from a storm. That's where that vertical velocity stops. So how do we do it in the air? NOAA P3 had two doubtful radars since early 1980s. Primary, those radars were used for tropical cyclone hurricane reconnaissance. So many of our current understanding of tropical cyclone were actually derived from the observations from those two tailed-out radars. And NCAR developed this state of our radar in the late 1980s called Eldora electro-adopter radar. And this was a picture taken in the same experiment. So what does the tailed-out radar look like? So there were flat-flat antennas called slotted waveguide antenna. And two of those mounted back-to-back at a tail of the aircraft. And this is a picture of the mounting structure without the outer shell and the two flat-flat antenna. So you can imagine that if you have these two antennas and the structure rotates, and if the aircraft is not moving, you are sweeping out two conical surfaces. One, if a beam is pointing forward, you sweep out a forward cone, and for the aft beam, you sweep out another cone. So when the aircraft is moving forward, you're actually sweeping out two helical surfaces. And any point in space will be first scanned by the forward radar, then a few minutes later, scanned by the aft radar. Okay, this is how we obtain two components of velocity without using two aircrafts with the doped radar on it. It will become a coordination nightmare. It's been done, but it's not easy. So after this is done, the rest is very similar to the ground-based radar. You integrate the continuity equation vertically, you get the third component of the wind velocity. So before we get into the storm structure, I need to mention the importance of the radar beam width and the sampling resolution. So I use a passive remote sensing camera as an example. So the left-hand picture is taken by a 10-megapixel camera. It shows two handsome gentlemen and a delicious sushi. They're ready to enjoy. And if you take the same scene using a 0.03 megapixel, which is 300 times worse resolution, you don't see the imperfection on their face, and the sushi becomes not that appealing. So this effect can be can be arrived by either using a lower-resolution camera or use the same high-resolution camera but taking the photo far, far away. Okay, then you blow that image up, you see the semi-fact. This is important because it is related to why we want to sample tornado and hurricane in the same way. So using radar as an example. So the left-picture illustrated NCAR hyperclaw radar, which is a pop-based radar on our G5. You can fly over 45,000 feet and looking downward into a storm. So this is an example taken about four years ago in the east coast. And looking downward, you can see the resolution is very high. The radar is about three to nine kilometers away. So you can clearly see the individual convective element and a lot of details. If we look at a pretty much the same slice from a WSR-80AD, our next radar located about 100 kilometers away, about the same place, you can barely see the detail. Okay, so if you are probing to a storm, which picture would you like to see? You want to see this one or this one? Okay, I'll pick this one. So the resolution is related to the radar characteristics. Okay, it can be either achieved by a bigger antenna or a smaller beam width. I'm sorry, wavelengths. Okay, smaller wavelengths. And typically for a mobile radar or airborne radar, the size of antenna is a limitation. Okay, you only have such a big fuselage. You cannot view much bigger antenna. Or you can basically sample in a storm closer. That's why we have radar on the mobile truck and put it on the aircraft. So we can chase tornado with flying to hurricanes in order to achieve these high-resolution data. All right, so chase tornado in the air. Vortex 1995. So this was a footage taken by a TV crew on the ground on May 22, 1995. Well, you see the anchor electro with anchor or door flying. Okay, so wherever I show this video, the audience often ask questions. How close were you? Is it turbulent? Is it dangerous? And the worst question or comment I received, are you guys crazy? And I want to show, before you make any judgment, I just want to show you the scientists are not crazy at all. So this is a picture, a top view, a satellite view of a supercell thunderstorm. Okay, you live in Colorado in the summertime, wherever you hear a forecast of severe thunderstorm or tornado, you go outside, you should see a supercell somewhere in the front range. So beneath that huge envelope cloud, there are this shape of precipitation. This is the hook echo that I mentioned earlier. A tornado supposed to form, if it ever form, will be around here. Okay, so this is called a forward flank gust front. This is called a rear flank gust front. Some weak tornadoes can be formed along the rear flank gust front. Okay, so for a seasoned tornado chaser, I'm teaching you how to chase tornado next season. If you look at a radar picture on your cell phone, you've got to have a cell phone when you chase tornado, otherwise you are crazy. So you look at a cell phone, you see this type of hook echo, you know you're on something and you want to stay on this side of a supercell. Okay, this is a clear side of a supercell. Everywhere else, most likely you'll be blocked by rain. Okay, so based on all these ground chasing experience, the PI of that project came up with the schematics. So we're going to fly the aircraft and the radar just on the clear side of the storm. And we try to stay away about 10 kilometers. Okay, for N-car aviation facility, all the PIs and the scientists, safety is our number one priority. Okay, we want to be alive after we collect the data. So we did not take this lightly. So the PI communicated with the pilots, the scientists, the radar scientists. We had so many discussions just to make sure everybody understand the risks and how can we fly the storm and collect the data we want. So we decided we're going to fly close, but not too close at 10 kilometers, roughly 10 kilometers away. And we're going to fly low, but not on the ground. We're flying at 1,000 feet. That's not very high to begin with. So this image occurred on June 8, 1995. You will see a F5 tornado near Caldwell, Texas. This image shows the funnel right here, rotating war cloud. You can find many similar videos on YouTube. And some of them with very exciting comments from a strong chaser. Just watch out. And I was on the N-car Electra, sampling the storm, lying in front of the storm. I did not know this footage until almost 15 years after that event. So thank you. There was no cell phone, no wireless internet at that time. Okay, so enough interesting pictures. Oops, okay. Now it was rated as F5 tornado. You can see one example of a damaged track. So this picture shows the damage track of this tornado. And these dotted lines illustrate where the aircraft was flying, sampling. As I said, the storm is here. We are basically, according to our playbook, staying in front of a storm, flying back and forth, back and forth, back and forth, sampling the storm. And I mark A, B, and C here. So about this track, we collect the data like this. Okay, this is the radar is collecting like a vertical slice. So this was interpolating to a horizontal plane to illustrate tornado. So one slice, oh, by the way, this is where the aircraft was. Okay, so we're about roughly 10 kilometer away from the storm. And on one slice, I observe a plus 89 meter per second. Remember that number 89 meter per second earlier? So they have five tornado. And the next slice, about two and a half seconds later, there was a 49 meter per second approaching. So this probably was one of the first, if not the first, direct measurement actually hit inside a tornado. It's probably not measured the highest speed of a tornado, but I mean, it's big enough. Okay, so this is from the AF radar. We also have observation phone for radar. So after some manipulation, you can see the background color is reflectivity. This is the hook echo that I repeatedly mentioned, signature of a tornado. And we can actually see a huge rotation right here. Okay, so this is, this pretty much tells me that the radar works. Okay, the philosophy and the reflectivity are matching. Next thing, at point B, a ground observer, Bruce Hyene, took a picture. If you look at the direction, it's roughly the same time as we sampled this part of storm. So you can see the funnel cloud was about one kilometer away. I only knew this later. Okay. And why there's a point C? Because the electron encounter a turbulence, we dropped from 1,000 feet to 660 feet. In a couple seconds, damage the radar and ended that experiment. Okay, and my son was two and a half years old then. I never had the courage to look at that inks issue data. My colleague told me we were that close. Okay. So being an aircraft scientist has some advantages. Because in vortex 95, on the ground crew, when we chased all the way to western Texas, we had home base in northern Oklahoma right here. Okay, so western Texas here. If you ever drive driven on I-40, you know how boring that was. And as an aircraft scientist, the aircraft, after each mission, the ferry time back to Oklahoma City was about an hour. And when we finished dinner and ready for bed, the poor ground chasing team were still approaching the home base. There's a good reason because logistically, you need to fix your equipment and get a good night's sleep and ready for the next day. But that really limits where you can chase. So pretty much, this was a circle that if we base on northern Oklahoma, this is a range that we can chase. So if we fought a clock 15 years to 2009-2010, I actually participate in the second year of the Vortex II experiment, the technology change evolved big time. So cell phone becomes available, wireless internet becomes available, satellite communication becomes available. So a group of scientists was bold enough to propose that they were going to conduct a completely mobile field experiment, which means no home base. The entire team is traveling in the Midwest through this tornado alley. The sampling strategy hasn't changed. Okay, here are a few images. If you watch the Discovery Channel or National Geographic, you should see this vehicle very often. And this vehicle actually appeared in the IMAX movie. Okay, so as I mentioned, the technology made this thing happen. So there were about 50 vehicles participating each year of the experiment. And how do we manage all these vehicles? It's by this control center called Mobile Operational and Repair Center. Remember, I mentioned that home base needs to repair our stuff. Now, this vehicle served part of that purpose. So every single vehicle has a position in this map. This is Supercenter Storm, the Hook Echo. Okay, each vehicle has a codename and a specific position. Okay, if this operation center says Supercenter is going to happen at what time, at what place, all the vehicles will position themselves to the right place at the right time to collect data. So I was associated with a team called CamerC. I'm going to play a video here. Let me explain what's this. Okay, so each vehicle has a GPS location unit on board. Then the information will be transmitted to that control van. Then the control vehicle collect every vehicle's location as superimposed, overlay them on a radar image like shown here and beam it back to each vehicle roughly every five minutes. So the movie you're going to see is a collection of all those images in a day. Okay, and as I said, you may be able to find CamC on this graph when I play it. Then this display called SASE, situational awareness of severe storm intercept. And on the left hand side is our daily routine of a typical day during Vortex 2. So I'll start to explain after when I play second time. So the first part was an ending part of a previous day operation near North Platte, Nebraska. Okay, then we move to Ogallala for the night. Then in the morning, we get out. So we take a rest in Goodland, Kansas, then converge to lunch. Then spread out. Then the storm forms. Then each vehicle was in position to sample the storm. One thing to mention that this is a very significant logistic challenge because there aren't that many towns in the Midwest that has the hotel facility to house 100 scientists with only a few hours of notice because typically the decision will be made about 3 p.m. that afternoon and decide where we're going to sleep tonight and position for chasing for tomorrow. Okay, so we probably usually arrive at a hotel about 10 p.m. and wake up at 8 and all that. So this is a non-trivial. So after I pulled out my camera, I was awarded my first tornado. You can see the rotating cloud based. Then I reposition the camera and suddenly I saw a funnel cloud start to develop. Be patient. Okay, now it's more obvious. Funnel cloud and the touchdown, unfortunately, it's gone. It's not very satisfactory. First experience. But after 21 days, 25 days in the field, that's better than nothing. Okay, so the next question you may want to ask is what am I doing? Why do I want to photograph all these tornadoes and being in the field for that long? For what? Let me explain. Oops. I need to control this. Okay, so Storm Chasers has many, many photographs of tornado since maybe 50s, 60s. So you can easily tell this is the beginning of a tornado and then a funnel cloud. In 1995, when we first put the radar on a truck, we started to collect this very, very high-resolution radar data. Okay, so this is the hook echo. This is the Doppler velocity. The radar was here. You see a receding velocity or approaching velocity. Then all these other interesting duple parameters basically tells you if a radar sees a hook echo, we can tell inside this region, it's more like a debris. We can see the strongest wind is blowing near the edge of this hook echo. However, when you look at this picture, how do you know the edge of a funnel cloud, the wind speed at the edge of a funnel cloud? You don't know that. Okay, you only see there's a funnel cloud. So we try to combine the radar information with the tornado picture. If we can match those, we'll learn a lot more. Okay, that's why I'm carrying a camera and a video camera chasing tornado for that many days. We are going to use the mathematics details. So the meteorologists use photogrammetry. Okay, if you know spherical trigonometry, you can understand this graph. But don't worry, you don't need to. So we can use a spherical trigonometry with some GPS measurement and a few targets on your picture. We can figure out exactly where the storm is located relative to the photographer. That's why I want to take picture next to a radar van because then the radar azimuth and elevation can be superimposed onto the photo. That's enough for spherical trigonometry. When you look at a radar picture, we typically just say, okay, the maximum speed of those tornadoes is this, they are approaching the next, approaching velocity is what? But look at the picture, there are lots of data points. Okay, can we extract more information from there? Yes, we can. So in 1999, I developed this velocity track display concept just in 10 seconds. Hope you can understand. So if we know where a tornado is, then I take a full circle, take all the data points along that circle because tornado is rotating. So it's basically a sine and cosine function. So if I do that to every radius, every height, I can extract the primary circulation of a tornado. Okay, without boring you with our equations. So if we combine all these information together, so we have a photo of a tornado and these dots, great dots, what the point that was determined by the spherical trigonometry for each radar pixel supposed to be and these vectors, oops, the vectors and the green colors or the tangential velocity are extracted from the velocity track display technique. Okay, so now we can see the edge of the funnel cloud was actually near the maximum rotation wind speed. Okay, so we can learn a lot from there. We can actually retrieve the central pressure from these tangential wind through a wind pressure relationship. But if you are a tornado expert, you will probably immediately tell me when child there's something wrong here. What's wrong? A tornado is supposed to suck the hero up. How come your vectors are all pointing downwards? Okay. Yes, Houston, we have a problem. The problem, in short, is that dots for radar observe particle motion, not a true air motion. And we assume the particle, the raindrop will move with the wind. So if we detect the particle speed, we know where we know the wind speed. In a tornado, it's a perfect storm to break that assumption. Okay, the physics is called centrifuging. Okay, and the centrifuging force is proportionately to a square of the velocity. Okay, in your middle school physics, there was the illustration with a string with a stone that you rotated. Then when you release it, the stone doesn't keep rotating. It goes out because there's a force to force a particle outside when it's rotating. So it's proportion to the square of the velocity and inversely proportion to the radius of the feature. All these two factors are working against our dot for radar assumptions. So once we scratch our head and figure this out with this some correction, now the strong downdraft in the middle becomes a weak updraft. We start to see much stronger updraft near the edge of the funnel cloud. That makes sense. There it comes in and goes out, goes up. I won't bad my salary on this magnitude. Okay, but at least it's qualitatively correct. It's a bad news for me because I thought I got a problem solved. The good news for the younger generation is there's still lots of things for you to do research. So some fun statistics of Vortex 2010. In 47 days, I traveled 17,000 miles on the Hertz rental car. On the right hand side, the blue lines and the symbols are where I stayed and I traveled in those 47 days. When I returned a rental car, the Hertz lady wasn't very happy. We were completely mobile experiments, so we stayed in the same hotel two consecutive nights only three times. So I learned very well how to unpack and pack my suitcase just moving every day. It's painful to watch tornado live on TV in a different state. That means we missed the forecast. We stayed in the hotel room and watched a potential target. It's happening live on TV. And too many visits to McDonald's Burger King, etc. Some fun pictures. Okay, chemistry actually stuck in mud in Texas and were rescued by AmEx film team. Fortunately, we were not on the footage in AmEx movie. It'll be embarrassing, but since every vehicle has a GPS and the control center knows exactly where you are, every other vehicle knows exactly where you are, so everybody knows chemistry will stop because we're not moving. After so many McDonald's Burger King, we finally arrived in Amarillo, Texas and found this great place to offer a $72 free steak dinner. If you can finish everything here in an hour and the steak is $72. If you want to try it, go to Amarillo, Texas. This is a camera I carried for 47 days, a lightning show almost every night. Only time we have fun for ice cream is during lunch hours because we're waiting for storms. And the IMAX film actually filmed the entire Vortex Armada wherever you see in this other than this fancy car, our Vortex team, called Tornado Alley. I don't know if anybody saw that IMAX movie. It was debuted in March for 2011 in Chicago. Now, how do we chase hurricanes on the aircraft? So, I mentioned earlier that two NOAA P3s equipped with tailed-out plurator, they were in hurricane reconnaissance business since 1977. In 2005, National Science Foundation approved the NCAR-EL-DORA to be loaded on the NRLP3 to join the two NOAA P3s to do a field experiment called RENEX. And the gentleman sitting there, Jim Morrow, was the ops director for that experiment. So, if you have more questions, ask him afterwards. So, this was the first time that we have three aircraft with three Doppler radars sampling a storm. And if you recall, Hurricane Rita reached category five once it entered a Gulf of Mexico. So, this movie loop shows three aircraft fly out of Tampa and how do we sample this storm. The ELDORA was this blue track, and in the end you will see it accidentally fly into the center of the tropical hurricane, Rita. Well, accidentally also intentionally. So, what makes this possible? It's still the communication, okay? So, this display is called Zebra display, which is a data integrator. So, it ingests the next red data, the satellite images, these are the next red reflectivity, and the NOAA P3 actually being transmitted their lower field search radar back to the control center. Then everything will be sent back to the aircraft. I, as a radar scientist onboard the NOAA P3, I would look at this image every five minutes and decide where to tell the navigator where to fly the aircraft next. So, we intentionally and accidentally entered the eye of Hurricane Rita, and this was a picture I took in the eye of Rita. So, this was the eye wall, like a stadium effect. This is looking down with some low cloud there. So, what's new about this? This was the flight track that was taken a day before when Hurricane Rita was actually a Calibri 5 storm. You see the reflectivity pattern, this is the inner eye wall, outer eye wall, this feature is called concentric eye wall, or so-called double eye wall. And these aircraft symbols, illiterate wear, the NRP3 with Eldora, where I was a flight scientist onboard. So, we circled the storm twice, and after the first circumnavigation, I found the outer rainbow wasn't very interesting. So, I sweet-talked the lead pilot and said, hey, can we do something new? Something interesting. And the pilot said, what do you want me to do? I said, well, can we get a little bit closer? I took about 20 minutes to negotiate that, and suddenly Jim Morrill in the operation center, I said, well, what the heck Wenchao is doing up there? So, we were actually in the moat region between the Indian and outer eye wall. And surprisingly, we were so surprised after that to actually see the detailed structure that we can observe by flying in this region. On the right-hand side, it illustrates three different times. One time, aircraft was outside the outer eye wall, then this time, the aircraft was in a moat, then we get out. So, it illustrates the evolution of Hurricane Rita in about two hours time frame. You see the inner eye wall, the outer eye wall, you see the radial flow, inflow outflow, and you see the intensification of the outer eye wall here. The updraft become much stronger. I want to point out a few other things that's sort of new, that's new that we found. Remember, this is a asymmetric averaging of every point around the circle. So, whatever shows up in the picture means it's very strong. On average, it exists. So, a few things worth noting. One is the downdraft in the moat region. The downdraft in the moat region is supposed to be dry. It's actually feeding to both the inner eye wall and outer eye wall. This was somewhat surprising that we found. The other things you see when we are flying very close, we started to see the air coming out of the eye, which is supposed to be very dry, entered the eye wall. So, these two features all have thermodynamic implications on Hurricane intensification. And the result was published in a science article. So, if you're interested, you can check that out. Now, some fun pictures of Rainnecks 2005. Before we fly on the NRP3, we have to pass through this very nice work called Navy MP4 Water Survival Training. It's actually a test to see if you can survive. So, you can see I struggled, tried to be afloat for five minutes before I can get to the next round of a test, which is where the full gear and have to flow for two minutes before I can pass. I passed, but I just don't like swimming. So, this was a picture taken by a French scientist inside category five, Katrina. Okay, you see an eye wall and a low cloud below. This was an image of a drop sound released by one of our technician. And this is like a radio sound upside down and measure the wind temperature in very detail, good detail. If we get out early in the morning, we'll see this nice rainbow, this is what we eat on the plane. And these pictures are happy graduate students at that point. One lesson learned is treat graduate students well because one day they may become your boss. And this gentleman now is a professor at CSU and this young lady is a professor at University of Illinois. They all write proposals to request a facility that I manage. So, I hope I treat them well. Okay, now how do we apply all these scientific findings and look at their impact on our daily life? So, this is a diagram, it's an old diagram, but get a message across. So, these are the years from 1978 to 2007. These are minutes. Basically, the tornado warning, average tornado warning time issued by the National Weather Service was about two minutes in 1978. So, you can imagine, if you hear a tornado surrounding boulder, you can hardly get to the parking lot from here in two minutes. Okay, there's almost no time for you to react. By 2007, it was about 13 minutes. The time ends here because the scientists did a statistic to write their Vortex 2 experiment proposal. In their proposal, they proposed that if Vortex 2 is a success, was a success, is a success in 2007, they hope they can increase the warning time from 13 to 30 minutes. Typically, there's about a 10 to 15 year cycle from the data was collected, the research is done, to it can be implemented. So, if you add 15 years from 2009 by 2024, you should see an update of this diagram. Another thing I want to mention is about Tropical Cyclone example. So, Hurricane Charlie in 2004, I don't know, some of you may remember that event. It was a category two storm, I played a radar picture collected by Key West and also Tampa. You'll see the picture actually swapped because the radar moved from one radar into a different radar range. So, I'll play the movie when I'm talking. So, it was a category two storm getting off the coast of Cuba and National Hurricane Center basically forecast that it will cruise just as is toward Florida. And interestingly, it rapidly intensified just before it may landfall. It was a surprise for National Hurricane Center for the coastal residents. It creates tremendous damage on the coast. So, my research partner and I found this poster child case, so we download the data from the National Weather Service and did a analysis on this case using the VTD analysis technique that I applied to the tornado analysis earlier. I'm going to show you the result. So, you can tell we have about 13 hours of data from this point when Hurricane Charlie was here until about here after the middle and fall. So, a couple of things illustrate on this diagram. This is time. The green line illustrates the pressure, central pressure retrieved from the Doppler observation. Okay. And the black diamonds indicates the drop sounds. I illustrated in the picture before. The U.S. Air Force actually released drop sound at the center of a storm. And by the way, it was mentioning that the drop sound was originally developed at NCAR, EOL about 20 years ago. Yeah, 20 years ago, GPS drop sound. It made tremendous contribution to track forecasts of the tropical cycle. Now, what I want to say is the red line was the storm maximum wind speed released by the National Hurricane Center. So, what you can see is the National Hurricane Center received the drop sound data and updated the storm's intensity three hours later, two hours later. If National Hurricane Center actually paid attention to the radar data at that point, they already knew rapid intensification, the pressure spotting off the floor in Hurricane Charlie. So, I showed this picture to National Hurricane Center and they're gladly to find me to work on this algorithm called Vortrack. Okay. And it was accepted for operational use since 2009. Okay. Future, phase array radar technology. Okay. The next red current technology is being parabolic antenna around and around and around. It's very inefficient way to sample storm. The phase array technology basically control the phase of all these little every single element. So, you can change the direction of the beam almost instantaneously by controlling the phase of each element. Okay. In the bottom panel, basically illustrate now we don't need to scan the radar as before. We can focus on sampling this nice storm, even a different storm somewhere else almost instantaneously. So, we don't waste time to scan all these nice clear where we're not interested. So, National Weather Service has a program called Metpar Meteorology Meteorological phase array radar program to evaluate how they can apply this technology to upgrade their national weather radar network. At NCAR currently we're developing a airborne phase array radar which we propose to put four panels like this on four different locations on our NCAR NSF NCAR C-130 aircraft to simulate what Eldora can do before it was retired in 2012. One big factor we decided to go to phase array radar technology because phase array radar technology enables the duperization capability where the old antenna cannot be upgraded for that. So, just in summarize so each panel consists about 37 LRU line replaceable unit which is the smallest panel that we can pull it off the radar if something goes wrong and each LRU has eight by eight elements. We really don't want to replace individual elements. So, a total of about 94, 72, 90,472 small radars. The development will be huge we're currently writing a proposal to National Science Foundation to request money to build it and how to manage this many radar has remained a challenge. Last slides from research to operations. So, the weather radar beginning in 1940s the US National Weather Service installed the first national weather radar network called WSR 57 in 1957 then they upgraded in 1974. In 1970s as I mentioned NCAR started develop with a community with developed Doppler radar trying to tackle the microburst feature and the National Weather Service was saw the great use so they upgrade the National Weather radar network in 1988 called WSR-88D Doppler and in the 80s NCAR and the community developed the polar metric radar and we demonstrate it's very useful in helping us to identify the precipitation, quantitative precipitation and what's inside a storm. So, the National Weather Service upgrade their radar to do pull completely in 2013. As I mentioned we're currently doing phase radar research hopefully 10-15 years from now you will see another update of this diagram. And I cannot do a justice to the weather radar and the science impact in 50 minutes. I hope I give you a good overview of why a radar meteorologist is such a fun job. I know hopefully my wife is listening this in Taiwan now. Hopefully she never didn't learn something that she had never learned before. Hope you have a great afternoon thank you. Eileen on the right side.