Abstract
This study investigated the melting layer detected by X band polarimetric radar (XMP). The feature of melting layer provides important information for quantitative precipitation estimation. XMP was used to observe upper areas for detecting initial symptoms of precipitation every 5 min. A 20-degree elevation was used for investigating melting hydrometeors. The analysis period was October 2010 – October 2011. Days when the daily rain amount was more than 10 mm were objects for study as analysis days. The melting layer existence was investigated at observation times(every 5 min) when the radar reflectivity index was higher than 25 dBZ near the surface on the analysis day. The polarimetric melting layer (PML), which was defined by the minimum of co-polar correlation coefficient (ρHV), was located lower than the reflectivity melting layer (RML), which was defined by the maximum of the radar reflectivity index, known as brightband. Furthermore, the PML width was less than that of RML. The different height or width between PML and RML was regarded as a reflection of the rain size distribution. Improvement of the rainfall estimation method was attempted by optimizing the coefficients in ZR relation according to the difference between the height of PML and that of RML. Rainfall estimations using variable coefficients were better than those using fixed coefficients.