Performance analysis on terrain-adaptive clutter map algorithm for ground clutter mitigation in Korea

Weather radar system is a system that transmits pulses and receives backscattered weather signals. It can provide information including location, moving velocity, and power of precipitation echo. These products are used as important data for forecasting and monitoring of mesoscale precipitation. The system can detect wide range of an area over the 200~500 km with a high range and time resolution. However, an accurate precipitation prediction using a weather radar system is a challenging task in Korea where more than 70% of its total territory is covered with mountainous area. The main problem is that it is highly likely that the returned radar signals consist of both ground clutters and precipitation echoes and cause wrong predictions. Thus, it is required to remove any non-precipitation signals, such as mountains, buildings and waves from the returned signals in order to improve the accuracy of precipitation prediction and make high-quality radar data. Currently, the signal processing system in use is developed in foreign countries and does not reflect the topographic characteristics of Korea. Therefore, it is necessary to introduce a new suitable clutter mitigation technique which will be adequate in Korea. Generally, ground clutter is stationary or moves slowly causing near zero Doppler velocity. Traditional clutter filtering methods are PPP(Pulse Pair Processing) in time-domain, and fixed width clutter filter, variable width clutter filter, GMAP(Gaussian Model Adaptive Processing) in frequency-domain, and etc. PPP is a method that uses phase difference between pulses in time-domain signal, and that removes any signals with no phase-difference considering as ground clutter. However, this method can also remove precipitation signals if signals are indistinguishable. Clutter filtering in frequency domain has become popular because it can interpolate removed signal. However, it can derive weather signal distortions in no clutter area because clutter filter is performed in all observing area. In this paper, C-Map algorithm is applied to raw signals acquired from Sobaeksan rain radar for a case study. The flow chart of C-Map algorithm is shown in Fig 1 and it removes the ground clutter from the returned signals in frequency domain using the Doppler difference. The ground clutter is generated in advance with the data in clear days. The result shows that the performance of ground clutter mitigation in frequency-domain was about 92.36%. It also suggests that an adequate threshold constant is required to preserve precipitation echo from ground clutter mitigation. We will apply moment data and verify performances at each data field. The technique will also be applied to Young-In Testbed (YIT) radar of Korea Meteorological Administration (KMA), which was installed in 2014. Acknowledgement: This research is supported by “Development and application of Cross governmental dual-pol radar harmonization (WRC-2013-A-1)” project of the Weather Radar Center, Korea Meteorological Administration.