Evaluation of Weather Radar with Pulse Compression: Performance of a Fuzzy Logic Tornado Detection Algorithm

Abstract

Trends in current weather research involve active phased-array radar systems that have several advantages over conventional radars with klystron or magnetron transmitters. However, phased-array radars generally do not have the same peak transmit power capability as conventional systems so they must transmit longer pulses to maintain an equivalent average power on target. Increasing transmits pulse duration increases range gate size but the use of pulse compression offers a means of recovering the otherwise lost resolution. To evaluate pulse compression for use in future weather radar systems, modifications to a weather radar simulator have been made to incorporate phase-coding into its functionality. Data derived from Barker-coded pulses with matched and mismatched filters were compared with data obtained from uncoded pulses to evaluate the pulse compression performance. Additionally, pulse compression was simulated using data collected from an experimental radar to validate the simulated results. The data derived from both experimental and simulated methods were then applied to a fuzzy logic tornado detection algorithm to examine the effects of the pulse compression process. It was found that the fuzzy logic process was sufficiently robust to maintain high levels of detection accuracy with low false alarm rates even though biases were observed in the pulse-compressed data.