基于A-Train卫星编队资料的热带气旋强度估算

准确估算热带气旋(TC)强度,对于预测TC发展、减少财产损失具有重要的意义。前人将TC看作满足静力平衡和梯度风平衡的轴对称涡旋系统,基于云顶高度、云顶温度、海表面气压等物理量建立了TC强度估算模型,该模型未考虑环境垂直风切变对TC强度的影响。本文提出一种修正模型,通过统计拟合手段将垂直风切变加入原模型中。从2006—2015年的Cloud Sat资料中筛选出穿心个例共63个。针对云雷达(CPR)数据特点,提出根据反射率因子的垂直分布确定眼墙和外围边界位置的方法。分别用原模型与修正模型对这63个TC个例进行强度估算。与最佳路径数据相比,原模型结果总体偏大,尤其对风切变较大、强度较小的个例估算效果不佳。修正模型对于风切变大于5 m·s~(-1)的个例误差明显减小,平均绝对误差MAE、均方根误差RMSE和平均绝对误差百分比MAPE分别从5.8 m·s~(-1)、7.7 m·s~(-1)和19.5%变为3.5 m·s~(-1)、4.9 m·s~(-1)和11.5%。在一定强度范围内,修正模型估算效果随着强度的增强而提升。修正模型对于成熟阶段的个例效果更好,北半球的估算精度高于南半球,纬度越高,估算误差越小。试验结果表明,用该修正模型估算TC强度是可行的,可以对现有的技术进行辅助和补充。

Estimating tropical cyclone intensity based on A-Train data

The accurate estimation of tropical cyclone (TC) intensity is of great significance for predicting the development of TC and mitigating property loss. TC can be seen as an approximately axisymmetric vortex, nearly in a state of hydrostatic and gradient wind balance in previous studies and a TC intensity estimation model based on cloud top height, cloud top temperature, sea surface temperature was built without consideration of the effects of environmental vertical wind shear on TC intensity. A revised model was presented in this study, which added the vertical wind shear into the original model by means of statistical fitting. 63 eye overpass cases were selected from the CloudSat data product of 2006-2015. According to the characteristics of CPR data, a new method was put forward to determine the locations of the eye wall and the outer region based on the vertical distribution of the radar reflectivity factors. The intensities of the 63 TC cases were estimated by using the two models. The results of the original model are generally larger than the best track data. The errors of the revised model are significantly reduced, especially for the cases whose vertical shears are larger than 5 m^.s^-1. The Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percent Error (MAPE) decrease from 5.8 m·s^-1, 7.7 m·s^-1, 19.5% to 3.5 m·^-1, 4.9 m·^-1, 11.5%, respectively. Within a certain intensity range, the performance of the revised model enhances as the intensity increasing. The revised model is more suitable for the cases at the mature stage. The estimation accuracy of the northern hemisphere is better than that of the southern hemisphere. The estimation error is smaller at higher latitudes. The results show that it is feasible to use this revised model to estimate the TC intensity, which can be used to supplement existing satellite-based technologies.