首页> 外文会议>International conference on image processing, computer vision, pattern recognition >Forecasting of severe Thunderstorms using K- nearest neighbor technique
【24h】

Forecasting of severe Thunderstorms using K- nearest neighbor technique

机译:使用K近邻技术预测强雷暴

获取原文

摘要

K-nn is one of the popular techniques in the field of pattern recognition. Here it has been applied for the prediction of severe thunderstorms. Three types of weather parameters i. e., moisture difference, adiabatic lapse rate and wind shear at different geopotential heights of the upper atmosphere have been taken into account for this job. Applying K-nn methodology we get more than 98% correct prediction for 'squall days' and 90% correct prediction for 'no squall days' with 12 hours lead time. Both surface as well as upper air data which are measured by radiosonde/ rawindsonde in the early morning are used in this case.
机译:K-nn是模式识别领域中的一种流行技术。在这里,它已被用于预测强雷暴。三种天气参数i。例如,这项工作已考虑到在上层大气的不同地势高度处的水分差,绝热消失率和风切变。使用K-nn方法,我们可以在12小时的交货时间内获得超过98%的“ s日”正确预测和90%的“无日”正确预测。在这种情况下,将使用无线电探空仪/ rawindsonde在清晨测量的地面和高空数据。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号