Fumihiko Mizutani

MMSE Beam Forming on Fast-Scanning Phased Array Weather Radar

A fast-scanning phased array weather radar (PAWR) with a digital beam forming receiver is under development. It is important in beam forming for weather radar observation with temporally high resolution to form a stable and robust main lobe and adaptively suppress sidelobes with a small number of pulses in order to accurately estimate precipitation profiles (reflectivity, mean Doppler velocity, and spectral width). A minimum mean square error (MMSE) formulation with a power constraint, proposed in this paper, gives us adaptively formed beams that satisfy these demands. The MMSE beam-forming method is compared in various precipitation radar signal simulations with traditional beam-forming methods, Fourier and Capon methods, which have been applied in atmospheric research to observe distributed targets such as precipitation, and it is shown that the MMSE method is appropriate to this fast-scanning PAWR concept.

Development of solid-state weather radar

Today, amid mounting sense of crisis of concentrated heavy rains, weather radars are of growing importance in weather observation and prediction by virtue of their capability of rain-fall measurement over the wide area in a short period of time. Meanwhile, because of the limitation of bandwidth availability with increasing use of radio wave frequency resources, steady demand for weather radar is also expected. Moreover, there is strong demand for decreasing the running cost of weather radar. Toshiba has developed 5GHz and 9GHz solid-state weather radar, complying with the demand the best use of our world-leading microwave transistor technology and radar data processing technology. The radar has attained the specified level of spurious radiation suppression without sacrificing any of radar system performances. Toshiba intends to promote the sales of the solid-state weather radar.

Application of Adaptive Digital Beamforming to Osaka University Phased Array Weather Radar

The X-band phased array weather radar (PAWR) at Osaka University has a rapid scanning rate and is capable of high-density observations in elevation; it produces approximately 100 plan position indicator radar images with a 60-km range, at different elevation angles, in less than 30 s. The PAWR uses a fan-shaped beam with a narrow beamwidth (1.2°) in azimuth and a wider beamwidth (from 5° to 10°) in elevation. With digital beamforming (DBF), the elevation beamwidth can be reduced to 1.2°, using 128 antenna elements arranged in tandem. Although the fan-shaped beam is useful for rapid scanning, the received signals tend to be affected by ground clutter. In this paper, we investigate the clutter suppression capability of common DBF methods: Fourier, Capon, and minimum mean-square error (MMSE) beamforming. Furthermore, to improve performance when the PAWR data contain errors—such as lacking data caused by mechanical problems—a correction method is proposed. The effect of clutter suppression using MMSE is shown to be greatly improved if used together with the proposed correction method. The resulting method is shown to sufficiently suppress clutter in all elevation angles above a few degrees, even in the presence of strong clutter; its clutter reduction performance is compared and found to be superior to the ones of the analyzed conventional DBF methods.

Performance of Minimum Mean-Square Error Beam Forming for Polarimetric Phased Array Weather Radar

In this paper, the development of a polarimetric phased array weather radar, which consists of a dual-polarized antenna with 2-D circular planar phase-array elements, is discussed. The radar is capable of measuring the 3-D rainfall distribution in less than several tens of seconds. Digital beamforming (DBF) is an important component in the development process of the phased array radar. In this paper, precipitation radar signal simulations are performed taking into consideration radar concepts in order to discuss the estimation accuracy of polarimetric precipitation profiles (differential reflectivity, specific differential phase, and copolar correlation coefficient) with two DBF methods that are based on Fourier and minimum mean-square error (MMSE) methods. A comparison of the performance of the two methods indicates that MMSE is superior in accuracy because of the effect of a stable and a robust main lobe and adaptively suppressed side lobes. MMSE also provides precipitation measurements eliminating the directional dependence of a beam pattern for improving the accuracy of measurements. It is also shown that the estimated accuracies of the precipitation profiles are almost independent of the number of pulses.

Development and observation of the phased array radar at X band

A new Phased Array Radar (PAR) system for meteorological application has been developed by Toshiba Corporation and Osaka University under a grant of NICT, and installed in Osaka University, Japan in 2012. The phased array radar system developed has the unique capability of scanning the whole sky with 100m and 10 to 30 second resolution up to 60 km. The system adopts the digital beam forming technique for elevation scanning and mechanically rotates the array antenna in azimuth direction within 10 to 30 seconds. The radar transmits a broad beam of several degrees with 24 antenna elements and receives the back scattered signal with 128 elements digitizing at each elements. Then by digitally forming the beam in the signal processor, the fast scanning is realized. After the installation of the PAR system in Osaka University, the initial observation campaign was conducted in Osaka urban area with Ku-band Broad Band Radar (BBR) network, C-band weather radar, and lightning location system. The initial comparison with C band radar system shows that the developed PAR system can observe the behavior of the thunderstorm structure in much more detail than any other radar system. The observed high temporal resolution images of the severe thunderstorm are introduced, showing the potential capabilities of the PAR system. The correlation coefficient of the reflectivity in PAR with C band radar ranges from 0.6 to 0.9 as a function of the distance from the PAR.

Osaka urban phased array radar network experiment

Osaka University, Toshiba and the Osaka Local Government started a new project to develop the Osaka Urban Demonstration Network. The main sensor of the Osaka Network is a 2-node Phased Array Radar Network and lightning location system. Data products that are created both in local high performance computer and Toshiba Computer Cloud, include single and multi-radar data, vector wind, quantitative precipitation estimation, VIL, nowcasting, lightning location and analysis. These new products are transferred to Osaka Local Government in operational mode and evaluated by several section in Osaka Prefecture.

Adaptive Pulse Compression Technique for X-Band Phased Array Weather Radar

Weather radar commonly uses a matched filter (MF) method to improve the range resolution and signal-to-noise ratio. A X-band phased array weather radar (PAWR), which is capable of 3-D precipitation observations in less than 30 s, is in operation at the Osaka University. The PAWR uses the MF method. In weather radar systems, the magnitude of the range sidelobes is an important topic because it can cause overestimation of the received power from a target, such as precipitation or ground clutter echoes. We propose a minimum mean square error (MMSE)-based pulse compression method to reduce the range sidelobes of the PAWR. We evaluated an MF, an MF with a raised-cosine window, and MMSE methods using numerical simulations and actual measurement data obtained from the PAWR. The results show that the MMSE method is clearly superior to the MF and MF with a raised-cosine filter methods when considering the reduction in the range sidelobes.

Simulation of digital beam forming method on 2-D phased array weather radar

A 2-D phased array weather radar (PAR) is under consideration. This radar system consists of 6992 antenna elements aligned to form a symmetrical round shape. Adaptive digital beam forming is used to sharpen the fan beam of scattered signals in order to observe precipitation at different elevation and azimuth angles simultaneously. In this paper, two digital beam forming methods including Fourier beam forming (FR) method and minimum mean square error (MMSE) beam forming method are used to simulate the performance of the 2-D PAR and compared with each other. Both point targets and distributed targets are simulated. It is found that the MMSE beam forming method can significantly reduce the sidelobes and is an effective beam forming method for 2-D PAR.