Automatic Recognition Of Low-level Jet Based On Wind Field Information Of MICAPS

In recent years,disastrous weather occur frequently in our country.It causes great losses to the safety of life and property.The loss could be reduced if the disastrous weather being accurately forecasted and timely prevented.Jet is closely related to disastrous weather,such as rainstorm and blizzard.Therefore,it is meaningful to realize automatic identification of low-level jet.A method of automatic identification of low-level jet was proposed in this paper.This method was based on wind field information of MICAPS and features of low-level jet.The main contents of this paper are summarized as follows:1.Learning the characteristics and movement of jet,to have a more complete understanding of jet combined with MICAPS.Then the effective area of jet and the principles of data extraction were proposed.2.The principle of pretreatment was proposed based on coordinates,wind speed and direction.Then the data was pretreated.3.The cluster algorithm was proposed combined with characteristic of wind.Data was clustered based on wind speed and direction.4.The clustering results were handled based on technical specification(Revised Draft)of mesoscale weather chart.Classes which Satisfy conditions were fited,smoothed with B spline curve.Finally,automatic identification of jet stream was realized.