• Andreas

Wind profiler and RASS – advanced weather sensors

Updated: Oct 24, 2021


Accurate weather information has been of crucial importance throughout the history of aviation. Different systems have been developed over the years and many go far beyond a casual windsock. Today, we look at two special sensors that are often combined: The wind profiler radar and the Radio Acoustic Sounding System (RASS).

Figure 1: Conventional windsock. A wind profiler is much more sophisticated than this.


Unlike other weather radars, which detect precipitation, wind profilers work with an entirely different principle: They rely on changes in the index of refraction of the air, caused by turbulence eddies [1][2]. As these eddies travel with the mean wind, it is possible to extract the wind velocity components. The typical setup involves a phased-array antenna which generates directional beams, one zenith beam (straight up) and either two or four tilted beams as depicted in Figure 2. Analysis of the Doppler-shift permits determination of the wind velocity [3]. In combination with a RASS the system can even calculate the virtual temperature of the air [2].

Key point:

Turbulence eddies drift with the mean wind and cause changes in the index of refraction [2].

Station layout

A phased-array antenna creates the beams shown in Figure 2. These are usually quite narrow (in the order of some degrees) and tilted away from the vertical by about 15° [1].

Figure 2: Wind profiler and RASS setup [3]

The beams are usually created in a repetitive pattern in order to extract the respective velocity components. This technique is known as Doppler Beam Swinging (DBS) [7]. A typical beam dwell-time is around 40 s [9]. Close to the radar antenna, there is a set of RASS antennae (loudspeakers) to create the acoustic waves. Signal processing and transmission electronics are housed in a shelter (See Figure 3).

Figure 3: Wind profiler and RASS site near Ziegendorf (Germany) [4] (legend added)

The above picture shows a wind profiler and RASS in Germany. Notice the significant shielding around the installation to avoid interference, particularly from wind turbines [4].

Emission spectrum

Figure 4 below depicts the emission spectrum using a “max-hold” measurement with a pulse-length of 1700 ns and a center frequency of 482 MHz [9]. A little math will yield a wavelength of 62 cm for that frequency. Keep that in mind for later. The power levels shown in Figure 4 do not represent the actual transmission power.

Figure 4: Emission spectrum of the wind profiler in Ziegendorf (Germany) [9]

Signal analysis

Turbulence eddies create changes in the index of refraction in a recurring pattern [1]. The frequency of the wind profiler is chosen in a way, that the Bragg-Condition is met (see info-box). Furthermore, most wind profilers use some form of pulse-integration technique, i.e. adding up multiple reflections over time to improve detection capability [5]. As the beams include a zenith beam as well as tilted beams, the measured doppler shift can be converted to represent the wind velocity vector [1]. This permits the creation of a “vector-field vs. time graph” as shown Figure 5:

Figure 5: Typical wind vector field obtained from a wind profiler facility [4]

The graph above shows the facility in “low” mode, there is also a “high” mode available which cover altitudes up to 16 km [4].



If radio magnetic waves encounter reflective objects, which are separated by a distance comparable to the wavelength, a special form of constructive interference can occur. For objects which are positioned in the direction of propagation, that distance is equal to half of the wavelength [1][12]. This is known as the Bragg-condition.

Figure 6: Bragg-condition creates constructive interference

The transmitted wave is depicted with a blue background. The left picture shows the first reflection wave (amber). The right side shows the situation half a period later, when parts of the transmitted wave hit the second reflection plane, that creates the green wave. As the amber wave has travelled half a wavelength by then, this creates constructive interference. The reflected waves add up to build a strong coherent wave front. Notice, that the pulse-length of the reflection is of course increased considerably.


Measurement of the virtual temperature

This is a very neat application of physics: The addition of acoustic loudspeakers to a wind profiler permits the measurement of the virtual temperature of the air. The principle is ingenious: The loudspeakers generate sound waves which have an acoustic wavelength equal to half of the radar wavelength [10]. Remember the 62 cm from earlier? The acoustic frequency here is around 1100 Hz, which yields a wavelength of about 31 cm – half of 62 cm. This is perfect for Bragg scattering, as we have seen earlier. The wind profiler then tracks the sound waves as they travel upwards and measures their propagation velocity – the speed of sound. This is of course temperature-dependent and thus allows to establish the virtual temperature for the air profile. The term virtual temperature is used, as the humidity of the air is unknown (see below).

Virtual temperature

The virtual temperature is the temperature that dry air would have, if its pressure and density were equal to those of a given sample of moist air [11].

Comparison with radiosondes

Notice, that this measurement is truly stationary as opposed to a radiosonde which will travel a significant distance (up to 100 km) [6]. Experiments have shown that the accuracy of the two techniques is comparable [6]. The wind profiler is fully automated and can be operated continuously whereas the radiosonde requires considerable human interaction for launch and usually requires a new piece of equipment for every launch [10]. Across the globe, several networks of wind profilers are in use to complement weather models and improve predictions [10].



[1] Rauber & Nesbitt, Radar Meteorology, 1st edition, 2018, Wiley Blackwell

[2] Martner et al., An evaluation of wind profiler, RASS and microwave radiometer performance, Bulletin of the American Meteorological society, vol. 74 nr. 4 1993

[3] Federal coordinator for meteorological services and supporting research, U.S. Wind profilers: A review, report FCM-R14-1998, 1998

[4] Engelbart, Lehmann, Görsdorf, Wind profiler radar in the aerological network of the DWD, presentation, 2005

[5] May & Strauch, An examination of wind profiler signal processing algorithms, paper, American Meteorological Society, 1989

[6] Weber & Wuertz, Comparison of rawinsonde and wind profiler radar measurements, Journal of Atmospheric and Oceanic technology, vol. 7 1990

[7] Gandhi et al., Signal Processing and Data Acquisition for Wind Profiler Using Labview, Int. Journal of Engineering Research and Applications, vol. 4, Issue 4 (version 6), pp.87-91, April 2014

[8] May, Moran & Strauch, The accuracy of RASS temperature measurements, Journal of Applied Meteorology, vol. 28, No. 12, pp. 1329-1335, December 1989

[9] Rohde & Schwarz, Wind Profiler für den Deutschen Wetterdienst, Neues von Rohde & Schwarz, Heft 180, 2003

[10] Raghavan, Radar Meteorology, 2003, Springer

[11] Glossary of Meteorology, https://glossary.ametsoc.org/wiki/Virtual_temperature, American Meteorological Society, October 2021

[12] Warsaw University, Quantum Electronics, lecture 3, 2010