X波段双线偏振多普勒雷达共极化差分相移资料的滤波方法研究

为了提高X波段双线偏振多普勒雷达的共极化差分相移Ψ_C质量，利用714XDP-A型车载X波段双线偏振多普勒雷达在北京顺义的探测资料，构造了一个综合滤波方法。将其应用于滤波处理，得到结果:（1）综合小波去噪滤波方法在滤波处理及差分传播相移率K_DP的估算方面相较于中值滤波、综合滑动平均、综合卡尔曼滤波及综合有限脉冲响应（FIR）滤波都是最优秀的，其能较好地保留原始Ψ_C平均趋势及细节特征（偏差基本在±2°可信范围内，部分区域由于对原始Ψ_C细节特征的保留会略有超出），避免探测误差零值对滤波产生的影响（误差零值订正值回归平均趋势水平，偏差在±2°可信范围内），对原始Ψ_C距离廓线的平滑程度达到了94%~97%平滑了超出±1.5° Gate-1（Gate表示雷达距离库）的波动，对±1.5° Gate-1内波动保留]；（2）估算的K_DP整体在±5° km-1边界阈值内，避免了探测误差零值、环境噪声和δ对K_DP估算产生的误差影响，K_DP对较强降水指示明显（在回波强度30 dBZ ≤ Z_H < 37 dBZ的层积云降水中K_DP普遍大于0.25° km-1，峰值可达到4.3° km-1；在回波强度30 dBZ ≤ Z_H < 50 dBZ的积雨云降水中K_DP普遍大于0.1° km-1，峰值可达到1.95° km-1）。本文工作对于降水估测和水成物粒子识别有着改善效果，这对雷暴天气预警预报具备重要意义。

A Study of Method for Filtering Copolar Differential Phase of X-Band Dual-Polarimetric Doppler Weather Radar

Measurements of the 714XDP-A X-band dual-polarimetric Doppler radar located at Shunyi of Beijing are used to develop an integrated filter method that can improve the quality of the copolar differential phase Ψ_C measured by X-band dual-polarimetric Doppler radar. Application of the integrated filter method shows that:(1) The integrated wavelet de-noising method is the best and better than the median, integrated mean, integrated Kalman and integrated FIR (Finite Impulse Response) filters. It retains the mean trend of raw Ψ_C range profiles and detailsdeviations are within ±2°, which indicates the results are reliable, although the values may slightly exceed the range in several gates (the numbers of radar range profile)]. Furthermore, the influence of zero value caused by measurement errors on the filtered results is avoided (the corrected value for zero value in the measurements returns to the level of mean trend with deviations within ±2°, which is regarded as credible). In addition, the fluctuation index of Ψ_C range profiles filtered by integrated wavelet de-noising can be reduced by 94%-97% compared to raw Ψ_C range profiles, and the smoothing effect is remarkable (smoothing the fluctuations exceeding ±1.5° Gate-1, and retaining the fluctuations inside ±1.5°Gate-1). (2) The estimated specific differential propagation phase (K_DP) are within the range of ±5° km-1, and influences caused by zero errors in the measurements and δ and environment noises are avoided. In addition, K_DP can well indicate strong precipitation (for precipitation in stratocumulus clouds, radar reflectivity is usually within 30 dBZ ≤ Z_H < 37 dBZ, and K_DP generally is greater than 0.25° km-1 with the maximum of 4.3° km-1; for precipitation in cumulonimbus clouds, radar reflectivity usually is within 30 dBZ ≤ Z_H < 50 dBZ, and K_DP is generally greater than 0.1° km-1 with the maximum of 1.95° km-1). This study is beneficial to rainfall estimation and hydrometeors classification, which is important for thunderstorm weather forecast.