| Severe convective weather is characterized by small spatial scale,strong suddenness and great destructive power,which sometimes brings great threat to people’s life and property safety,so it is of great scientific and practical importance to make fast and accurate forecasts and warnings.In this paper,we discuss severe convective weather in two parts: potential prediction and proximity warning.For the prediction of potential,the ERA5 reanalysis data,Doppler weather radar data and severe convective data are mainly used to study the three types of severe convection weather in Guangdong Province from 2010 to 2019,including hail,short-time heavy precipitation and thunderstorm gales.Potential prediction models were established for 3hours,6 hours and 12 hours,respectively.The latter two types of weather were further classified according to the large-scale circulation situation.Then NCEP GFS(Global Forecast System)forecast data were used to test the weather processes in 2020 and 2022 in Guangdong Province by Bayesian method.For the proximity warning,this paper summarizes the radar proximity warning indicators applicable to different types of strong convective weather based on the product characteristics of Doppler weather radar.The main conclusions are as follows:(1)The characteristics of environmental parameters of the three types of severe convective weather indicated that the water vapor is the most abundant when short-time heavy rainfall occurs,and the temperature difference between the middle and low layers is the least.The thunderstorm gales had large unstable energy and the highest convective effective potential energy.The static instability of hail is the largest,and the vertical wind shear from the lower to the upper layers shows a gradual increase.Compared with the previous two kinds of weather,the main difference lies in the lower 0 ℃ layer height and the obviously dry middle troposphere.(2)According to the 500 hPa circulation pattern,the short-term heavy rainfall can be divided into two troughs and one ridge type,two ridges and one trough type,the Bay of Bengal pre-trough type,marginal subtropical high type,tropical storm type and zonal type.The marginal subtropical high type has the largest instability energy and the smallest vertical wind shear,while the tropical storm type has the highest water vapor content,the smallest temperature difference between the middle and low levels,and the vertical wind shear is stronger.Thunderstorm gales can be divided into three types,among which the pre-trough type is dominated by dynamic factors and supplemented by thermal factors.The strong subsidence of cold air is the most significant characteristic of the post-trough pattern.The thermal factor plays the main role in the marginal type of subtropical high.(3)The average hit rate of all three types of severe convective weather potential forecast by Bayesian algorithm exceeds 80%,and the average hit rate after short-time heavy precipitation and thunderstorm gales typing improves by about 0.1 compared with that before typing.Furthermore,by improving the Bayesian method based on the resampling of environmental parameters and application of multi-layer conditional probabilities,the problem of unbalanced sample size of thunderstorm gale is solved.On the basis of Bayesian model excluding basic probability and combining with large-scale environmental field,the potential prediction model can be optimized,the null prediction rate can be reduced,and the prediction ability can be improved.(4)The success probability of warning for the three types of severe convective weather processes in 2020 can reach more than 90% by using the summarized threshold of severe convective weather approaching warning indicators,which can give a good early warning indication,indicating that the composite reflectivity,echo top and vertical integrated liquid water content density performs well as statistical indicators for early warning,which can provide reference for future meteorological early warning. |
PDF Full Text Request