The Evaluation of WSR-88D Hail Detection Algorithm over Guizhou Region
-
摘要: 通过制定一定的规则,使用贵州504个防雹炮点的冰雹观测资料及2005,2006年贵阳雷达站8次冰雹过程观测资料,使用时间窗方法间接地将风暴单体与降雹记录相联系,建立了冰雹算法校验数据库,并对降雹校验数据库的数据进行统计分析,应用探测概率、虚警率、临界成功指数来检验冰雹探测算法。强冰雹概率 (POSH) 的强冰雹探测算法的总体评估结果表明:当POSH算法的强冰雹预警阈值为30%时,在贵州地区获得最高的临界成功指数评分,但这个阈值在每次强冰雹预警时并不是都获得最佳结果。强冰雹预警阈值选择模式 (WTSM,即根据冻结层高度动态选择每天强冰雹指数的预警阈值) 在不同地区气候状况下的差异,是导致缺省的POSH算法在贵州地区应用不佳的最主要原因,这也说明对冰雹探测算法的局地性适用评估非常必要。通过对WTSM的调整,改进了原来的POSH算法。
-
关键词:
- 冰雹探测算法;
- WSR-88D风暴算法的评估;
- 强冰雹概率;
- 强冰雹预警阈值选择模式
Abstract: The evaluation databases for the WSR-88D hail detection algorithm have been built by using hail observation data of 504 hail prevention spots in Guizhou and Doppler radar data of Guiyang during 8 of severe hail cases from 2005 to 2006, filtered by specific conditions including definition of severe hail, observation range along the cell track, and time-window methodology, etc. These conditions make the databases provide a more accurate picture of algorithm performance. The algorithms are evaluated using the probability of detection (DPO), false alarm ratio (RFA), and critical success index (ICS) statistics.It shows that POH (probability of hail) threshold of 50% get the highest RFA, and different POH thresholds get similar ICS, suggesting that it is unreliable to use POH as the only parameter for hail detection in Guizhou region. The difference of climatology between Guizhou area and central Switzerland where the initial POH curve is derived is the crucial cause why POH algorithm becomes unreliable and gets higher RFA.Assigning POSH (probability of severe hail) 30% leads to the highest ICS score in Guizhou region, but this threshold does not always get the best performance in these 8 severe hail cases. The difference of WTSM (Warning Threshold Selection Model) in different climatic region is the main cause why the default POSH algorithm gets a bad performance. An improved WTSM will predict the optimum ISH threshold for each day more accurately. It will help ensure that the POSH threshold of 50% always corresponds to the largest possible ICS every day. The re-evaluation of the improved POSH algorithm shows that it has decreased the hail detection RFA, and gets a higher performance of severe hail detection in Guizhou.-
Key words:
- hail detection algorithm;
- evaluation of WSR-88D algorithm;
- POSH;
- WTSM
-
图 4 2006年4月9日15:15贵阳雷达1.5°仰角风暴识别图 (相邻距离圈之间距离为30 km)(a) 及风暴1在l线上的垂直剖面图 (b)
Fig. 4 1.5° elevation PPI reflectivity and storm identification displays of Guiyang radar at 15:15 9 April 2006
(the distance between odjacent circles is 30 km) (a) and cross section of storm 1 as line l in Fig. 4a(b)
表 1 8次冰雹过程的POH算法总体评估结果
Table 1 Performance results of POH algorithm for 8 hail cases
冰雹
概率/%记录
总数n成功 n虚警 n漏报 DPO RFA ICS 10 562 341 196 25 0.93 0.37 0.61 20 534 323 186 25 0.93 0.37 0.60 30 510 310 174 26 0.92 0.36 0.61 40 484 295 161 28 0.91 0.35 0.61 50 463 285 150 28 0.91 0.34 0.62 60 434 265 138 31 0.90 0.34 0.61 70 393 232 121 40 0.85 0.34 0.59 80 337 194 98 45 0.81 0.34 0.58 90 250 139 64 56 0.71 0.32 0.54 100 176 90 29 66 0.58 0.24 0.49 注:记录总数表示以对应冰雹概率作为冰雹探测预警阈值时进入降雹校验数据库的记录数。 表 2 8次冰雹过程的POSH算法总体评估结果
Table 2 Performance results of POSH algorithm for 8 hail cases
强冰雹
概率/%记录
总数n成功 n虚警 n漏报 DPO RFA ICS 0 362 192 163 7 0.96 0.46 0.53 10 194 115 69 10 0.92 0.38 0.59 20 157 99 47 11 0.90 0.32 0.63 30 119 82 25 13 0.86 0.23 0.68 40 86 50 16 21 0.70 0.24 0.57 50 63 27 9 28 0.49 0.25 0.42 60 39 9 5 35 0.20 0.36 0.18 70 9 2 3 41 0.05 0.60 0.04 表 3 每个冰雹检验日POSH算法评估结果
Table 3 POSH performance for individual days
日期 统计量 强冰雹概率/% 0 10 20 30 40 50 60 2005-05-01 记录数 60 31 21 17 9 2 1 n成功 46 24 21 17 9 2 1 n虚警 13 6 5 1 0 0 0 n漏报 1 1 1 1 2 6 6 DPO 0.78 0.80 0.81 0.94 1 1 1 RFA 0.22 0.20 0.19 0.06 0 0 0 ICS 0.77 0.77 0.78 0.89 0.82 0.25 0.14 2005-05-02 记录数 52 33 32 29 25 14 6 n成功 24 22 21 21 19 12 1 n虚警 28 9 9 6 4 0 0 n漏报 0 2 2 2 2 2 5 DPO 0.46 0.71 0.70 0.78 0.83 1 1 RFA 0.54 0.29 0.30 0.22 0.17 0 0 ICS 0.46 0.67 0.66 0.72 0.76 0.85 0.17 2006-04-09 记录数 62 19 14 9 7 4 0 n成功 24 7 6 6 4 0 0 n虚警 37 11 7 2 2 1 0 n漏报 1 1 1 1 1 3 3 DPO 0.39 0.39 0.46 0.75 0.67 0 0 RFA 0.61 0.61 0.54 0.25 0.33 1 0 ICS 0.39 0.37 0.43 0.67 0.57 0 0 2006-04-24 记录数 63 43 37 23 14 12 9 n成功 40 29 28 16 7 6 2 n虚警 22 13 8 4 1 0 0 n漏报 1 1 1 3 6 6 7 DPO 0.65 0.69 0.78 0.80 0.88 1 1 RFA 0.35 0.31 0.22 0.20 0.12 0 0 ICS 0.63 0.67 0.76 0.70 0.50 0.50 0.22 -
[1] Forsyth D E, Bjerkacs C, Petrocchi P. Modular Radar Analysis Software System (MRASS)//Preprints, 20th National Conventional Conference on Radar Meteorology. 1981: 696-699. [2] Lemon L R. On the Use of Storm Structure for Hail Identification//Preprints, 18th Conf on Radar Meteorology. 1978: 203-206. [3] Witt A, Eilts M D, Stumpf G J, et al. An enhanced hail detection algorithm for the WSR-88D. Wea Forecasting, 1998, 13: 286-303. doi: 10.1175/1520-0434(1998)013<0286:AEHDAF>2.0.CO;2 [4] Witt A. Comparison of the Performance of Two Hail Detection Algorithms Using WSR-88D Data//Preprints, 26th Int Conf on Radar Meteorology. Amer Meteor Soc, 1993: 154-156. [5] Witt A, Johnson J T. An Enhanced Storm Cell Identification and Tracking Algorithm//Preprints, 26th Int Conf on Radar Meteorology. Amer Meteor Soc, 1993:141-143. [6] Witt A, Mitchell E D, Johnson J T, et al. Evaluating the performance of WSR-88D severe storm detection algorithms. Wea Forecasting, 1998, 13: 513-518. doi: 10.1175/1520-0434(1998)013<0513:ETPOWS>2.0.CO;2 [7] Winston H A, Ruthi L J. Evaluation of RADAP Ⅱ severe storm detection algorithms. Bull Amer Meteor Soc, 1986, 67: 145-150. doi: 10.1175/1520-0477(1986)067<0145:EORISS>2.0.CO;2 [8] Eric L, Henry E F. An evaluation of WSR-88D severe hail algorithms along the northeastern Gulf Coast. Wea Forecasting, 1998, 13: 1029-1044. doi: 10.1175/1520-0434(1998)013<1029:AEOWSH>2.0.CO;2 [9] Mather G K, Treddenick D, Parsons R. An observed relationship between the height of the 45-dBZ contours in storm profiles and surface hail reports. J Appl Meteor, 1976, 15 : 1336-1340. doi: 10.1175/1520-0450(1976)015<1336:AORBTH>2.0.CO;2 [10] Waldvogel A, Schmid W, Federer B. The kinetic energy of hailfalls. Part Ⅰ: Hailstone spectra. J Appl Meteor, 1978, 17: 515-520. doi: 10.1175/1520-0450(1978)017<0515:TKEOHP>2.0.CO;2 [11] Waldvogel A, Federer B, Schmid W, et al. The kinetic energy of hailfalls. Part Ⅱ: Radar and hailpads. J Appl Meteor, 1978, 17: 1680-1693. doi: 10.1175/1520-0450(1978)017<1680:TKEOHP>2.0.CO;2 [12] Waldvogel A, Grimm P. Criteria for the detection of hail cells. J Appl Meteor, 1979, 18: 1521-1525. doi: 10.1175/1520-0450(1979)018<1521:CFTDOH>2.0.CO;2 [13] Foster D S, Bates F C. A hail size forecasting technique. Bull Amer Meteor Soc, 1956, 37: 135-141. [14] Miller R C. Notes on Analysis and Severe-storm Forecasting Procedures of the Air Force Global Weather Central. Tech Rep 200 (Rev), Air Weather Service (MAC), US Air Force, 1972: 1-190. [15] 王瑾, 刘黎平.基于GIS的贵州省冰雹分布与地形因子关系分析.应用气象学报, 2008, 19(5):627-634. http://qikan.camscma.cn/jams/ch/reader/view_abstract.aspx?file_no=20080515&flag=1 [16] Browning K A, Foote G B. Airflow and hail growth in supercell storms and some implications for hail suppression. Quart J Roy Meteor Soc, 1976, 102: 499-534. doi: 10.1002/(ISSN)1477-870X [17] 葛润生, 姜海燕, 彭红.北京地区雹暴气流结构的研究.应用气象学报, 1998, 9(1):1-7. http://qikan.camscma.cn/jams/ch/reader/view_abstract.aspx?file_no=19980101&flag=1 [18] 朱君鉴, 刁秀广, 黄秀韶.一次冰雹风暴的CINRAD/SA产品分析.应用气象学报, 2004, 15(5):579-589. http://qikan.camscma.cn/jams/ch/reader/view_abstract.aspx?file_no=20040571&flag=1 [19] 廖玉芳, 俞小鼎, 郭庆, 等.一次强对流系列风暴个例的多普勒天气雷达资料分析.应用气象学报, 2003, 14(6):656-662. http://qikan.camscma.cn/jams/ch/reader/view_abstract.aspx?file_no=20030683&flag=1 [20] Marshall J S, Palmer W M. The distribution of raindrops with size. J Meteor, 1948, 5: 165-166. doi: 10.1175/1520-0469(1948)005<0165:TDORWS>2.0.CO;2 [21] Changnon S A. Hailstreaks. J Atmos Sci, 1970, 27: 109-125. doi: 10.1175/1520-0469(1970)027<0109:H>2.0.CO;2 [22] Kessinger C J, Brandes E A, Smith J W. A Comparison of the NEXRAD and NSSL Hail Detection Algorithms//Preprints, 27th Conf on Radar Meteorology. Amer Meteor Soc, 1995: 603-605.