Mode-concept for individualising the Z-R-relation for precipitation radar

雷达对降水的遥感技术通常允许根据需要在线确定面雨量，例如作为径流和洪水预报模拟的输入数据。传统上，根据雷达反射率计算降雨率时采用的是一个幂函数，该函数要么是根据经验得出的气候学平均值，要么是基于理想平衡情况的理论结果。将得到的雨量与测量的雨量进行比较，结果表明，当采用静态幂函数Z-R关系时，雷达像素级的误差可达5倍。以前根据降水或天气类型描述水滴大小分布的许多尝试都没有成功，主要是因为人们认为只有大尺度的气象结构对区分是重要的。在本建议中将要发展的模式概念中，Z-R关系应通过叠加几个滴模来确定，从而允许也考虑小规模气象现象。不同水滴模式的相对贡献将根据现有的水滴大小分布的长期观测和辅助气象标准来系统地确定，如降水强度、云的垂直扩展、层状或对流类型的降水、零度层的高度、垂直反射率剖面的类型等。首先应确定典型降水类型和过程的最多四个离散雨滴模式的平均相对贡献。有了在雷达图像或数据集中识别这些降水过程和类型的可能性，以后就可以从查询表中得出降雨率，从而可以得出滴状模式的最佳组合。将利用AQUARADAR合作伙伴的地面散射计和微雨雷达在特别观察期(SOP)期间测量的综合液滴光谱来调查和验证经验确定的关系的有效性。

The remote sensing technique for precipitation by means of radar generally allows for the online determination of areal precipitation, as needed for example as input data for runoff and flood forecast simulations. Traditionally the calculation of precipitation rates from radar reflectivities applies a power law conversion function which is either derived empirically as a climatological mean or theoreticall based on the ideal equilibrium case. A comparison of derived rainrates with measured ones shows errors up to a factor of 5 at radar pixel level when a static power law Z-R relation is applied. Many previous attempts to characterise drop size distributions in terms of precipitation or weather types have not been successful mainly because it was assumed that only large scale meteorological structures are of importance for discrimination. In the mode concept to be developed in this proposal the Z-R relation shall be determined by superimposing several drop modes, thus allowing to take also small scale meteorological phenomena into account. The relative contributions of the different drop modes will be determined systematically from available long-term observations of drop size distributions and ancillary meteorological criteria like precipitation intensity, vertical extension of the cloud, stratiform or convective type of precipitation, height of the zero degree layer, type of vertical reflectivity profile, and others. Primarily the average relative contributions of up to four discrete dropmodes shall be determined for typical precipitation types and processes. With the possibility to identify these precipitation processes and types in the radar image or data sets the precipitation rate can be later derived from look-up tables allowing for the derivation of optimal combinations of the drop modes. The validity of the empirically determined relations will be investigated and validated by the use of the comprehensive droplet spectra measured during the Special Observation Period (SOP) from ground based disdrometers and Micro Rain Radars of the partners witing AQUARadar.