Time series of microphysical structure of a thundercloud examined with hydrometeor classification method for X-band polarimetric radar

   It is important to examine the time series of microphysical structure of thundercloud in order to understand the mechanisms of severe weather phenomena, such as hail-fall, tornado, lightning and heavy rain. Polarimetric radars are useful instrument to obtain microphysical information and, therefore, they have been used for hydrometeor classification (hereafter, HC) in many researches (e.g. Liu and Chandrasekar, 2000, published in J. Atmos. Oceanic Tech.). The authors have modified HC method for S-band polarimetric radar (S-pol) described in Liu and Chandrasekar (2000) to adapt to X-band polarimetric radars (X-pols) and tried HC with X-pols of Nagoya University (Kouketsu and Uyeda, 2010, presented at ERAD2010). In this study, we targeted single thundercloud of which we observed entire life cycle. We conducted HC in a single thundercloud around Nagoya, central Japan area, observed with an X-pol of Nagoya University and examined time series of microphysical structure of the thundercloud.

   Around 2030 JST (Japan Standard Time; JST = UTC + 9) on 26 July 2010, a thundercloud was generated over Gifu Prefecture, about 50km north of Nagoya, and from 2054 JST, cloud-to-ground (CG) lightning was observed. The place where the thundercloud was generated is the north edge of plain and wet southerly wind from the Pacific Ocean blows in summer season. According to balloon sounding at 21 JST at Hamamatsu, the nearest balloon sounding point from the thundercloud, the air of low and middle level, up to 600 hPa height, was wet with relative humidity (RH) exceeding 80 %, and there was cold air in the upper level. And the CAPE was very high (more than 2000 J/kg). We observed the thundercloud with an X-pol of Nagoya University. The volume scan interval of the X-pol was 6 minutes with 15 elevations from 0.5° to 33.5°. After observation, we conducted HC with four polarimetric valuables obtained with the X-pol: radar reflectivity with horizontal polarization (Z_h), differential reflectivity (Z_dr), specific differential phase (K_dp) and correlation coefficient of horizontally and vertically polarized signals (ƒÏ_hv). And for supplementary information, we used temperature data of ground and balloon sounding observational data. The polarity of lightning is important information because it suggests the presence of graupel and/or ice crystal and, therefore, it is useful barometer of validity of HC. To obtain the information of polarity and frequency of CG from the thundercloud, we used the data of Lightning Location System (LLS) performed by Tubu Electric Power Company.

 In this study, we used HC method for X-pol described in Kouketsu and Uyeda (2010). We classified hydrometeor types into 10 categories: 1) drizzle, 2) rain, 3)wet snow, 4) dry snow, 5) ice crystal, 6) dry graupel, 7) wet graupel, 8) small hail, 9) large hail and 10)rain and hail. This HC method is basically based on that for S-pol (Liu and Chandrasekar, 2000) with some modifications. We tuned membership functions (MBFs) of K_dp in order to adapt to X-pol because the value of K_dp depends on wavelength. We also made MBFs of temperature and for a part of HC categories (drizzle, rain, wet snow, dry snow, ice crystal and dry graupel), we take account for RH at surface because the temperature that solid hydrometeors melt depends on RH (Matsuo and Sasyo, 1981, published in J. Meteor. Soc. Japan). With this method, we conducted HC for every sampling element with 150 m range and 1.2° azimuthal angular resolutions, and calculated the volume of the region where the hydrometeor type was classified as graupel or ice crystal by summating volumes of sampling elements.

   When the thundercloud was generated, there were few km³ regions classified as graupel (wet and dry graupel) and ice crystal, and no CG was observed. Then, graupel region increased and when the volume of graupel region exceeded about 100 km³ (2054 JST), negative CGs were began to be observed. From 2100 JST to 2118 JST, the volume of graupel region reached the peak (about 150 km³) and the frequency of negative CGs also reached peak. From 2112 JST to 2118 JST, when the volume of ice crystal region increased rapidly, few positive CGs were observed. After that, the volume of graupel region decreased and the last (negative) CG was observed at 2124 JST, when the volume of graupel became below about 100 km³.

As described above, negative CGs were observed only when there was large volume of graupel region. And according to HC of RHI scan at 2117 JST, the peak of the frequency of negative CGs, the graupel region existed up to 11 km height and the temperature of the main graupel region was below -10 °C. These fact is consistent the polarity of CG expected from the riming electrification process (Takahashi,1978, published in J. Atmos. Sci.).

   In the conference, we will discuss the detail of time series of microphysical structure of the thundercloud with consideration of the life stage of the thundercloud.