Abstract:Low level jet is important for predicting severe convective weather. At present, the identification of low level jet is conducted mainly by handwork, which brings the problem of low efficiency, easily influenced by subjective factors. So, based on the wind field data of the sounding stations in MICAPS, we propose an automatic low level jet identification and drawing algorithm in this paper. The algorithm is based on the definition of the low level jet axis, and detects the low level jet axis from several aspects in terms of wind speed, wind direction, sounding station distribution, and the central axis position. Then after the steps of transitive closure clustering, key points extracting, low level jet axis merging and the axis smooth ing, the automatic identification and drawing of low level jet are achieved. The test result shows that the jet axis, which is automatically drawn, has the characteristics of accurate position and natural shape. Besides, it could reflect the transport path of water vapor in jet, and adapt to the complex environment of low level jet. In the 291 test data, the identification rate reaches 94.96% and false alarm is not found.