Welcome to this lecture on remote sensing of wind, this is a lecture where I will try to explain you how we today use remote sensing techniques in the measurement of wind for wind energy. And the objectives that I'm going to tell you about are the different principles of using remote sensing, different instrumentation, and how we use the Doppler principle and how it works and eventually, I will show you how we have implemented Wind Scanners using scanner technology, so we can scan wind fields all over the place in the real atmosphere. So welcome, and let's go ahead. Today, we have wind turbines, and they are becoming bigger and bigger, they are growing in size. If you look at the lower picture here, you'll see that back in 1979, we had very small wind turbines here, at Riso, DTU, they were only 20 - 30 meters tall. Meanwhile, in the beginning of the Millennium, they reached 100 - 120 meters, as you see in this picture. And today we have huge turbines at our test site,Hovsore, we're testing big turbines for the wind industry, and they are more than 220 meters tall. So we are reaching almost the size of the Eiffel Tower with these turbines that are running in the atmosphere and therefore, we need to measure the wind at very high heights and at very big fields. So the good old technique where we could use as you see in this picture, a meteorological mast is very difficult to apply today because a meteorological mast only measures a wind field at one point where it's sitting, and at a single vertical profile, but we need to measure the wind field over the entire rotor plane of the big turbines today to understand how they are loaded with the wind and how much energy they can extract. So these are the objectives, and to do this, we need to look at the different remote sensing techniques that we have. The challenge of using remote sensing instead of an in-situ instrument is that you need to measure the wind by not being there. If you had set up a cup anemometer that rotates, then you count how many revolutions at that point in time, but with remote sensing, you can only transmit. You need to look at the signal that you get back either from sound propagation or from light propagation. You can also look at the sky, look at the clouds and see how fast they move. This is passive. But to measure the wind field around the big turbines in an active way, we have developed a system now that relies on remote sensing, active remote sensing. So we transmit a laser beam or sound beam, then we detect the Doppler shift that we receive back from this transmission. From that, we need to extract information about the wind speed and the wind direction. And the challenges, the main challenges with remote sensing is two things: You need to have a sizable signal back that you can detect. That's something with signal to noise and we need to have a decent signal to noise ratio in our measurements. And secondly, remote sensing techniques are diffused in the sense that they don't measure just in one point, but they measure in a volume or in an area. So the challenges are good signal to noise, and the second challenge is to have finite measurement volumes, that is a small measurement volume. And this is the big challenge that we are addressing with this new technology. isn't a new technology that we use today, because in the good old days in 1875, there was already a suggestion by Tyndall to use a big horn like a foghorn and then with the unaided ear, you were able to listen if there were any echo's in the straight you could listen whether there was a ship or not even though it was foggy. This is using sound, but today, we use remote sensing with light instead. And light is much more easy to handle than sound in this sense, because it's more precise, as I will try to explain you now. To detect the wind speed, you'll need to measure a frequency shift. You know all that if there's an ambulance passing you, you can hear how the tone is shifting, because of the motion of the transmitter relative to the receiver, the Doppler effect, but it's the same also when you transmit a light beam. Light also shifts it's frequency as for instance, you might know that if you look at far galaxies, they turn red, it's called red shift. So far galaxies, when you look at them in a telescope, you see how the lines of hydrogen are shifted and you can actually detect how far away they are in this manner. This is the same with the transmission of light that we use in the liders. We detect a very small deviation, a very small change in the frequency that we transmit. We transmit a laser beam, and then we look at the signal that comes back from scattering from aerosols in the atmosphere and then we compare that signal that comes back with the signal that we sent out, and this ratio of the frequencies shift relative to the transmitted frequency is the measure of the Doppler shift. Now, it's very different from sound and light. If you look at sound, you have a sound speed of typically 340 meters per second and if you have 17 meters per second wind speed, then it's 5%. So with sound you can anticipate 1 to 5% relative change in the frequency in what you have to detect as Doppler shift. With the light, however, it's much different because the speed of light is 300,000 kilometres per second or 3 times 10 to the power of 8, meters per second and if you only have 3 meters per second wind speed, this fraction is only 10 to the power of minus 8. So, it's an extremely small relative deviation in the light frequency and the light wave lengths and that's very difficult to detect. The frequency of light,which is in our case has wavelength of about 1.5 micron, the frequency is 200 terahertz, or 200 times 10 to the power of 12 [Hertz], so, 2 to the power of 14 Hertz and the frequency shift that you get back is only a Megahertz. So, this fraction is 10 to the power of minus 8. No electronics can detect, the absolute frequency, it's way too high, so we need to mix them, we need to use this principle we call coherent detection, we mix the return signal that scatters from the aerosols with the transmitted frequency, and then we get an interference which is of the order megahertz and this we can detect which common electronics. And this is how it works, let's look at these equations. These are the relative change in frequency and they are of the order of the change in the wind speed relative to the transmission velocity. The transmission velocity can be sound speed or light speed. And from this comes the Doppler equation that you see here and now I will show you how this is derived. this is the physics, but there is a coefficient of minus 2 in front of this and this you can understand from this simple argument, I'll show you here. So let's look at the Doppler equation that was developed by Christian Doppler back in 1842. We have the dispersion relation for the Doppler shift which is a relation between the frequency and the wave number that's given and we transmit a frequency f_0 which is very stable. The very stable frequency hits an aerosol and the aerosol acts like a small mirror and transmit light back. So because the aerosol acts like a mirror, it is an elastic scatter we are talking about, so the back scattered wave number is almost the same as the incoming wave number. So you can tell that from this if it's an elastic scatter, the change in wave number is simply twice the lengths or the size of the wave number. And to calculate the change in the wave number, it's vector sum, so you can take the return signal minus the incoming wave number, and that's a minus 2 times the size of the vector itself. So from here the minus 2 comes that I showed you in the previous equation. Now you take the delta k and put it into the dispersion relation, and then you have the Doppler shift equation. And the Doppler shift equation simply says that if you have a wavelength, a laser or sound with a certain wave number, and you have a speed of the object that you are measuring, then you get a frequency shift of this order it is minus two times the Doppler shift velocity divided by the wavelengths. And if you take a laser, it's approximately 1.3 megahertz per meter per second. So if you have a one meter per second wind speed in our laser beam, then we will expect a frequency shift of 1.3 Megahertz, which is easily detectable by today's electronics. So that's the background behind the Doppler shift that we use. Then you need to focus your laser beam where you want to measure. There are two ways to do that. You can for instance, send out a pulse like a radar, so you have a finite pulse, that you transmit, and you know light travels with the speed of light and then you can detect from where your back scatter is coming. That's one way, another way is to use the so called continuous wave principle and then you need a good telescope. A good telescope is a very high quality lens system that can take your single wave number, single wave length light and keep it collimated, i.e. keep it focused at very long distances. If you have ideal telescope, for instance like the Hubble telescope, it has a diameter of more than one meter and you use this wavelength that we're using here, one micron, then you have one thousand kilometers, over where you can keep your laser beam collimated, one thousand kilometres in front of the Hubble Telescope. We don't have Hubble Telescopes. We have good telescopes but they typically are two, three or four inch in diameter only. But that can easily keep the beams collimated or focused within the range where we are measuring the wind, which is typically of the e order of a few kilometres. And now if we want to measure the wind speed at a certain point, then we focus the beam so that we get a very high intensity at the focus point. So if I want to measure the wind speed at 200 meters distance, then I focus my telescope at 200 meters and I make sure that the intensity of the laser radiation then peaks at 200 meters and then I measure the wind speed in that little volume. This is how it works and after that it disperses. So you need good optics and it need very stable lasers to achieve the remote sensing. We have used this principle in our test equipment, in this case in Hovsore in Jutland. If you go back to 2004, you'll see that there were only SODAR devices, sound devices, not a single wind lidar at that point. But soon after and just a year after 2006, we had a series of continuous wave CW lidars. We had Leosphere pulsed lidar systems and pulse lidar systems. So in that aftermath of turn of the Millennium, soon after the Millennium, the whole field changed from being mainly focused on acoustic devices to today being mainly using laser driven remote-sensing equipment. And now we can use these technologies. Here's an example of what we call a Wind Scanner. This is a continuous wave (CW) LIDAR and we can focus it in front of the telescope and we can measure the wind at different distances. These pictures here show how we instrument. We have a small island called Bolund just outside Riso, DTU. So in morning, we could drive out the Scanner, we had to cross some water, carry up the instrument, mount it on top of the hill, and then here you see the hill so we could install the lidar system here, and measure the wind flow in front of the island. And this is what the instrument could do, the instrument then has some prisms, two prisms. So it can move its beam up and down or wherever we want to point it in two directions and in this case, we made virtual Meteoroligical mast(Met masts) in front of the hill. So there's a Met mast, a virtual Met mast, 20 meters in front of the hill, and then there's one where the hill begins, and then we could install as many virtual Met masts and measure wind profiles as we wanted to do. So with this instrumentation, this single lidar, we were simply able to measure the wind in front of the hill, and as you approached the hill and came closer and closer to the lidar, you can see how the wind profile changed and you can see you get much more wind energy and much higher wind speed being on top of the hill here than being up front. So that's an example. You could also measure turbulence because the instrument is very fast, it samples 400 samples per second, so you can measure many profiles in a very short time. You can scan like we have done here in 2D the wind profile and the turbulence profile. This is one use. We can also point the lidar totally vertical and just keep it pointing vertical, and then we can measure the wind speed in this case, the vertical component at many heights and this is what we've done in this graph, it was a clear day where we start in the morning. What you see here is the time of day, and what you have up here is the height up to two kilometers. In the morning, in this case again in Jutland, there was absolutely no turbulence, but as the sun sets on, and the bottom of the atmosphere starts to "cook" because of the heat, then you can see how turbulence develop and during the day, it develops more and More, and in the afternoon, you have a very, very high level of turbulence, in this case, at 600, 700 meters meight, in the vertical wind component. And maybe you experience this when you land with an air plane, if you're going to land in the afternoon, you very often experience that you have very bumpy rides and lots of turbulence, but think of that when you get closer to the ground the turbulence disappears. So whenever the airplane comes down to 150 meters, you can see how the turbulence level decreases and that's because of the presence of the ground. Vertical turbulence cannot exist when there is a boundary. So, think of that if you get nervous and it gets bumpy. I want to tell you also, make sure that you understand that the lidar only measures the velocity component in the direction of the laser beam, it doesn't measure the two transverse components. So, if you really want to measure 3D turbulence, you need three of them to be working at the same time and we have developed some instruments that does this. This is a so called SpinnerLidar, it's an instrument that you can install on a turbine, you can put it in to the spinner itself and then it measures the wind speed in front of the turbine and you can use this for controlling the turbine or optimizing the power. It's called a SpinnerLidar. It's still a lidar like we talked about, but here you see this scan head where there are two prisms that can steer the beam in two independent directions. And here's an example where we steer the prisms, so that the laser beam is following this "nice" (rosette) pattern here and every time the colour changes, it has made one measurement. So in this "rosette" scan, which we can run through with the scanner in one second, you get 400 measurements. So what is 400 measurements per second and if you combine that colour code with the wind speed that you measure, you get these high - resolution pictures of the inflow in front of the turbine and this we can refer to as a "wind speed camera". It's a camera that measures the wind speed in front of the turbine and then you can see there can be shear, there can be wakes from other turbines. This is very useful for optimal control of a wind turbine, and also for yawing the turbine into the wind. So in short, I have shown you a short history of the remote sensing techniques. We started with sound, it's now today it's mainly based on laser technology. I explained to you that Doppler shift and the Doppler equation. There's a minus 2 in front of it because light is scattered back and it change it's wave number. And I showed you here, very lastly, the SpinnerLidar where we can steer the beam and make a picture of the wind with just a single laser. But also the WindScanner technology: Often we have three lidar systems that work together coherently and make full 3D wind field measurements of all three wind components U, V, and W.
