An Algorithm Research Of Gust Front Identification For CINRAD

Gust front is a microscale weather phenomenon that often causes serious groundwind and strong wind shear, therefore shorttterm forecasting, nowcasting and the civilaviation systems had paid high attention to the studies of gust fronts. In this paper,synoptic characteristics of gust fronts and the influence to meteorological elements wereanalyzed. As weather radar products’ color code would affect the effect of radar imagesand recognition of gust front echo, configuration rules of color code were discussed, andan improved color code was presented for radar basic products. Before radar dataaccessed the finally recognition algorithm, some methods about quality control shouldbe done, containing noise filtering, fill the lacked points, data smoothing and velocityambiguity correcting. Based on the works above, two recognition subalgorithms fornarrow-band echo and convergence line were designed. Then the front line was locatedby the algorithms’ results and some quantitative physical quantities which couldrepresent the intensity of gust front were presented. Finally, actual cases were used toanalyze and inspect the identification effects. Some conclusions could be presentedfrom this article:(1)Gust front have feedback effects to the maternal storm, when the storm with agust front moved slowly, warm moist air could be pushed away from maternal storm bythe cold outflow, the maternal storm would disappear rapidly, while, when the storm andthe gust front moved fast, warm moist air might be lifted into the maternal stormcontinuously by the divergent cold outflow, the maternal storm got supplementaryenergy, so it would keep developing strongly. When a gust front across, it might causesome weather phenomenon just like: temperature descending, air pressure rising andwind speed increasing.(2) The improved color code for reflectivity could make the transition fromperiphery to center of the stratiform cloud echo more apparent, could make the colorresolution of the convective cloud more clear, could make the strong center moreoutstanding, could effectively highlight the gust front’s linear characteristics ofnarrow-band echo. The range of the improved color code of radial velocity dynamically changed with PRF could help to show the large speed area, air jet, wind shear and speedfuzzy under different PRF effectively. The improved spectral width color code can validshow turbulence intensity and the linear edge of a gust front, furthermore, there is astrong visibility in the area with high spectral width and a apparent color transition inthe region with low spectral width.(3) Gust front on reflectivity images show a narrow-band echo, the intensity is5-30dBz, the width is2-10km, the length is50-200km. There is a certain distancebetween gust front and maternal storm, this distance continuously increase from theprimary stage to the decline stage of gust front. The reflectivity in narrow-band wasalmost consistently, but higher than the surrounding echo at least5dBz. Gust front onthe radial velocity image show a wind speed convergence or a wind shear. The radialvelocity value decreased continuously as the distance extended when the radial acrossthe front line. Spectral width field of gust front characterized by high spectral widthvalue is4-12m/s. Before forming a gust front, the strong center of the maternal stormwould decline rapidly, this point can deem to be a important parameter on gust frontwarning.(4) K-neighborhood frequency method could filter out noise echoes and repair thelack points well, median filtering could effectively restrain the pulsation of the radialvelocity data and eliminate the residual noise, the optimal one-dimensional radial andone-dimensional tangential velocity ambiguity correcting technique could solve theproblem of speed fuzzy effectively.(5) Based on the features of gust front on the reflectivity and radial velocity image,a algorithm called “bilateral grads” and the convergence line identification algorithmwere designed.These two algorithms were used to recognize the narrow-band echo onthe reflectivity image and wind shear or wind speed convergence on the radial velocityimage, then the front line would be located by the results from them.(6) Some actual cases were used to examine the algorithm, the results showed that:When the characteristics of narrow-band or convergence line were obvious, thealgorithms could recognize the gust front successfully and perfectly, while, when thenarrow-band was fractured, mixed with ground objects or the radial velocity data wasincomplete because of the distance folded, the algorithms would be limited. As for the98radar volume scan data, the algorithms could identified out70.