基于机载探空和雷达观测资料的冰相粒子谱特征分析  

cn Investigating ice particle size distribution characteristics using collocated aircraft in-situ measurements and radar observations

作　　者：李安文 吕迎辉 赵坤[1,2] 黄浩 胡向峰 段英 LI Anwen;L Yinghui;ZHAO Kun;HUANG Hao;HU Xiangfeng;DUAN Ying(Key Laboratory of Mesoscale Severe Weather/MOE/School of Atmospheric Sciences,Nanjing University,Nanjing 210023,China;State Key Laboratory of Severe Weather and Key Laboratory of Radar Meteorology,CMA,Beijing 100081,China;Hebei Provincial Weather Modification Center,Shijiazhuang 051432,China)

机构地区：[1]南京大学大气科学学院/中尺度灾害性天气教育部重点实验室,南京210023 [2]中国气象局雷达气象重点开放实验室/灾害天气国家重点实验室,北京100081 [3]河北省人工影响天气中心,石家庄051432

出　　处：《气象科学》2024年第5期936-948,共13页Journal of the Meteorological Sciences

基　　金：国家自然科学基金资助项目(42025501,42175159);中央高校基本科研业务费专项资金资助项目(020714380177);江苏省社会发展重大科技示范项目(BE2023766);国家灾害天气国家重点实验室开放项目(2022LASW-A01,2023LASW-A01)。

摘　　要：为获取不同种类冰相粒子拟合谱参数的统计关系,本文利用2017年5月22日的飞机原位探测数据,对中国北方地区的冰相粒子谱进行研究。针对机载高容量降水光谱仪(High Volume Precipitation Spectrometer,HVPS)观测数据,提出了一种基于决策树方法的图像分类算法,可将水凝物粒子分为小冰晶、柱状冰晶、聚合状雪花和雨滴,并提取出水凝物粒径、轴比和数量等特征,结合采样体积计算沿飞行轨迹的各类水凝物的粒子谱。利用观测粒子谱的各阶矩将其拟合为Gamma谱分布,并分析谱分布参数间的统计关系,该关系可作为雷达反演模型的约束条件。为了验证拟合粒子谱参数的准确性,利用T矩阵方法计算沿飞行轨迹的各类水凝物散射特性,并使用雷达模拟器结合拟合粒子谱模拟雷达变量。结果表明,模拟雷达反射率与附近S波段雷达观测结果吻合良好,在整个飞行过程中平均绝对误差为19.55%,这可能是由飞行观测和雷达观测的采样体积不同以及T矩阵方法中参数假设所造成的。基于观测构建中国北部背景下粒子谱和雷达变量间的物理关系,可以改进冰相粒子谱的雷达反演算法,为进一步分析冰相微物理过程,提高数值预报的准确性提供基础。In order to obtain the statistical relationship of fitted particle distribution parameters of various kinds of ice phase particles,the PSDs of ice hydrometeors in North China were studied using aircraft in-situ observations on 22 May 2017.A new particle habit classification algorithm based on decision tree method for High-volume Precipitation Spectrometer(HVPS)images was introduced,which can automatically classify hydrometeor into small ice crystal,column,aggregate snow,and rain and can detect hydrometeor characteristics such as particle size,aspect ratio,and quantity.Based on the hydrometeor characteristics detected by this classification algorithm and instrument sampling volume,the PSDs along the flight track were calculated and fitted to Gamma distributions using multiple moments of the PSDs.The statistical relations between different parameters of the Gamma distributions were then examined,which can be used as constraints in building radar retrieval algorithm.To verify the observed PSDs,radar variables were simulated by radar simulator that uses the gamma fitted PSDs along the flight track and single scattering properties of hydrometeors calculated using the T-matrix method.The simulated radar reflectivity shows good agreement with a nearby operational S-band radar observations,with a mean absolute error of 19.55%over the entire flight.This difference may be caused by different sample volume of the airborne observations and radar observations,and uncertainties in parameters used in T-matrix calculation.The collocated observed PSD characteristics in North China can help to improve the retrieval algorithms of ice PSDs with radars,which can improve the analysis of the ice phase microphysical processes and would help improve the microphysical parameterization of numerical modeling in the future.

关 键 词：冰晶分类 微物理特征 粒子谱 雷达模拟器 

分 类 号：P407.7[天文地球—大气科学及气象学]