天气雷达回波衰减订正算法的研究(Ⅱ):数值模拟与个例实验

通过数值模拟研究雷达反射率因子呈均匀分布情况下的“可订正厚度”，计算了“无误差”情况下的订正效果以及“观测误差”、“k的误差”、库分辨等因素对订正效果的影响，还用实际观测个例进行衰减订正实验，并与解析算法和迭代算法的订正结果进行比较。结果表明，逐库算法，尤其是逐库近似算法，比迭代法计算效率高，比解析法有更大的稳定范围。

A Study on the Algorithm Making Attenuation Correction to Observations of Radar Reflectivity Factor(II): Numerical Simulations and a Case Study

Uniformly distributed radar reflectivity is assumed for numerical simulation to study the effectiveness of the attenuation correction algorithms and criteria for computational stability, and the influence of the errors in radar observation and attenuation coefficient and the bin-resolution on the effectiveness. Radar observations are also used in a case study of attenuation correction. The results have been compared with those from HitschfeldMBordan algorithm and those from Meneghini's iteration. The results have shown that correctable thickness in radial direction decreases as radar reflectivity factor and attenuation increase, that high resolution of data bin is beneficial to improve the stability during correction computation and correctable thickness can be increased, and that the random error in discrete radar data and the error in attenuation coefficient make the correction more difficult to carry out. The binMbyMbin algorithm is as good as HB when attenuation is weak and errors negligible. Otherwise the approximated binMbyMbin algorithm can be used to obtain a computing efficiency higher than the iterative scheme. The criteria determined by the coefficients in k-Z relationship and bin resolution can be used to avoid efficiently the computation overflow caused by unstable over-correction.