Intrapulse Modulation and Pulse Compression in Weather Radar
Intrapulse Modulation and Pulse Compression in Weather Radar
The use of intrapulse modulation and pulse compression has fundamentally changed modern weather radar technology by enabling the transition from conventional vacuum-tube transmitters (magnetrons or klystrons) to low-maintenance, phase-coherent solid-state transmitters.
While vacuum tubes can generate very short transmitted pulses with extremely high peak power in the megawatt range, solid-state amplifiers inherently deliver significantly lower peak power. To transmit sufficient pulse energy for a long detection range, the transmitted pulses must therefore be extended in time. However, because a long, unmodulated pulse would seriously degrade range resolution, the transmitted signal is modulated internally during the pulse and compressed again into a narrow pulse in the receiver by a matched filter. [1][2]
Special Considerations for Volume Targets
In meteorology, the pulse compression principle familiar from air surveillance cannot be applied without modification, because weather echoes do not originate from isolated point targets but from extended volume targets. With distributed scatterers such as raindrops or snowflakes, the apparent sensitivity gain from compression is qualified: Although the longer transmitted pulse increases the total radiated energy, subsequent compression into a narrower range cell simultaneously reduces the illuminated scattering volume and therefore the number of hydrometeors detected at the same time. This effect cancels the energy gain compared with a short pulse of equal energy. The primary benefit in weather radar is therefore purely technical: it makes it possible to achieve the necessary average transmitted power using solid-state components. Furthermore, correlation in the receiver is sensitive to the Doppler shift and spectral width associated with moving precipitation particles.[1]
The Problem of Range Sidelobes
The greatest technical challenge for meteorological volume targets is posed by so-called time sidelobes (Time Sidelobes or Range Sidelobes), which appear before and after the main peak during pulse compression. While these sidelobes generally cause little interference for a point target, such as a single aircraft in open airspace, and can often be effectively suppressed, they add up and overlap across all adjacent range gates (Range Gates) in a continuous precipitation field. Because the reflectivity of weather targets spans an extremely large dynamic range of more than 60 to 70 dB, strong echoes – for example from an intense hail core – may completely mask weaker echoes from adjacent light rain through their range sidelobes, thereby distorting meteorological measurements. In addition, the long transmitted pulse creates a large blind zone at short ranges, which is why modern systems operate in hybrid mode: A short, unmodulated pulse is transmitted for the near range before switching to a long, modulated pulse for greater distances.[3]
Selection of Suitable Waveforms and Sidelobe Suppression
To reduce interfering time sidelobes to the level below -55 to -70 dB that is considered essential for weather radar, and distribute them uniformly over time, simple phase codes such as Barker codes (at best -22.3 dB for a length of 13) are unsuitable for meteorological applications, or can be used only for very short-range radar systems. Instead, specially optimized waveforms and filtering techniques are employed:
- Nonlinear frequency modulation (NLFM): Unlike a conventional linear chirp (LFM), which without additional measures provides sidelobe suppression of only about -13.2 dB, NLFM adjusts the rate of frequency change over the pulse duration in an S-shaped pattern. This gives the transmitted signal's power spectrum a bell-shaped envelope, producing very low, uniformly distributed sidelobes in the matched filter without the signal-to-noise ratio penalty (mismatch loss) associated with subsequent amplitude weighting.
- Linear chirp (LFM) with amplitude windowing: When a conventional linear chirp is used, the received signal must be weighted in the pulse compression filter with a suitable window function (e.g., a Taylor, Hamming, or Blackman window). This reliably suppresses sidelobes and distributes them uniformly at a low level, but slightly broadens the main peak (somewhat reducing range resolution) and costs approximately 1 to 2 dB in sensitivity.
- Polyphase codes and complementary sequences: Polyphase codes such as Frank, P3, and P4 codes offer considerably better autocorrelation properties and lower Doppler sensitivity than simple binary codes. More recent systems also use complementary codes (such as Golay sequences or the Welti code) or other specially optimized waveforms, most of which are based on chirp variants. With these complementary codes, the sidelobes of two consecutive pulses can ideally cancel each other out. However, for weather targets this works only to a limited extent because raindrops rapidly decorrelate between pulses. Consequently, NLFM signals or windowed LFM signals dominate in practice.[4]
Practical Examples of Weather Radars Using Pulse Compression
Figure 1: MP-PAWR: Solid-state weather radar with a phased-array antenna
While intrapulse modulation was initially introduced primarily in compact short-range systems and phased-array systems operating in the X-band (such as the Japanese PAWR and MP-PAWR), advances in high-performance gallium nitride (GaN) semiconductors have now established this technology in long-range C- and S-band weather radars as well:
- X-band (short to medium range): Systems such as MP-PAWR (Multi-Parameter Phased Array Weather Radar) combine pulse compression with electronic beam steering[5] to achieve high-resolution 3D volume scans in just 30 seconds, despite the limited peak power of their numerous individual modules.
- C-band (medium to long range): Modern commercial systems such as the Vaisala WRS300, EEC Endurance series, or Meteopress C-band systems replace conventional magnetrons entirely with solid-state transmitters. They achieve ranges of more than 250 to 400 km by combining short, unmodulated pulses for the near range with long, NLFM-modulated pulses for greater distances in hybrid operation.
- S-band (long range / national networks): Solid-state versions of the low-attenuation S-band radars (wavelength around 10 cm, ranges exceeding 450 km) are now also available (e.g., from EEC or Meteopress). Moreover, special pulse compression waveforms have been investigated for the U.S. WSR-88D (NEXRAD) to meet extremely demanding sidelobe suppression requirements, with the longer-term prospect of replacing conventional klystron transmitters.
Sources and references:
- G. L. Austin: „Pulse Compression Systems for Use with Meteorological Radars“, Radio Science, Vol. 9, No. 1, 1974, pp. 29–33. (DOI: 10.1029/RS009i001p00029)
- Sebastián M. Torres, Christopher D. Curtis, David Schvartzman: „Requirement-Driven Design of Pulse Compression Waveforms for Weather Radars“, Journal of Atmospheric and Oceanic Technology, Vol. 34, No. 6, 2017, pp. 1351–1369. (NOAA Full Text PDF)
- Cesar M. Salazar Aquino, Boonleng Cheong, Robert D. Palmer: „Progressive Pulse Compression: A Novel Technique for Blind Range Recovery for Solid-State Radars“, Journal of Atmospheric and Oceanic Technology, Vol. 38, No. 9, 2021, pp. 1599–1611. (NOAA Full Text PDF)
- Cesar M. Salazar, Boonleng Cheong, Robert D. Palmer, David Schvartzman, Alexander V. Ryzhkov: „Improvements in the Compression Filter and Calibration Factor of the Progressive Pulse Compression Technique“, IEEE Transactions on Geoscience and Remote Sensing, Vol. 62, 2024, pp. 1–14, Article 5101814. (NOAA Full Text PDF)
- Robert Palmer, David Bodine, Pavlos Kollias, David Schvartzman, Dusan Zrnić, Pierre Kirstetter, et al., „A Primer on Phased Array Radar Technology for the Atmospheric Sciences,“ Bulletin of the American Meteorological Society, vol. 103, no. 10, pp. E2391 - E2416, 2022. (NOAA Full Text PDF)