Geometrical Characterization of Precipitation Patterns
Geometrical Characterization of Precipitation Patterns
复制标题
DOI：
10.1175/2010jhm1298.1
复制
发表时间：
2011-04-01
影响因子：
3.8
通讯作者：
Sorooshian, Soroosh
中科院分区：
文献类型：
--
作者：
AghaKouchak, Amir;Nasrollahi, Nasrin;Sorooshian, Soroosh
卫星估计和天气预报模型使观测和预测大空间尺度的降水成为可能。尽管在观测降水模式方面取得了重大进展，但空间模式的特征描述仍然是一项挑战。空间格局的定量评估方法是未来发展预测降水的空间范围和格局的关键。在这项研究中，降水模式的特点是使用三个几何指标：（i）连接性指数，（ii）形状指数，和（iii）分散性指数。使用多个例子，所提出的指数的应用进行了探索，在卫星降水图像的模式分析和验证数值大气模式的几何特性。结果表明，所提出的指数可以合理地用于不同模式的相对比较（例如，多个字段与空间观察）的连接性、组织和形状。
Satellite estimates and weather forecast models have made it possible to observe and predict precipitation over large spatial scales. Despite substantial progress in observing patterns of precipitation, characterization of spatial patterns is still a challenge. Quantitative assessment methods for spatial patterns are essential for future developments in prediction of the spatial extent and patterns of precipitation. In this study, precipitation patterns are characterized using three geometrical indices: (i) a connectivity index, (ii) a shape index, and (iii) a dispersiveness index. Using multiple examples, the application of the proposed indices is explored in pattern analysis of satellite precipitation images and validation of numerical atmospheric models with respect to geometrical properties. The results indicate that the presented indices can be reasonably employed for a relative comparison of different patterns (e.g., multiple fields against spatial observations) with respect to their connectivity, organization, and shape.