Study On Convective Incipient Based On Himawari-8meteorological Satellite Monitoring

Convective activities are increasingly affecting people's lives,often accompanied by bad weather,which seriously interferes with normal aviation activities and causes economic losses,and even casualties.At present,the research on convection mainly focuses on the development and mature stage of convection,while the research on the initial stage of convection still has some limitations.If we can accurately and effectively predict the initial stage of convection in advance,it will play an inestimable role in the aviation industry,meteorological industry and ecological environment,and we can also make targeted disaster prevention and reduction before the occurrence of strong convection.Therefore,this thesie proposes a set of high-resolution algorithms for the monitoring of multi-channel convection events in summer and Himawari-8,which is suitable for the specific application of high-resolution AHI in Shanghai.The algorithm is based on the improvement of mb06 algorithm.The core of the improvement is to establish a new field of interest and threshold,count the CI events in Shanghai from May to July,quantify and visualize the atmospheric state or cloud top characteristics represented by 8 places of interest through these events,and obtain the scatter diagram.After screening and analyzing the threshold,7 fields of interest are finally retained.The overall algorithm is divided into image fusion,cumulus mask,target cloud tracking and multi-channel time series.Firstly,the image of visible light channel and infrared channel is fused to improve the image resolution,then the data is pre screened through cumulus mask,and then the new cumulus object tracking method is used to track the target cloud.Finally,combined with multi-channel time series technology,the corresponding thresholds of seven fields of interest are established as the recognition conditions for predicting CI events,As long as a region is greater than or equal to 6 identification conditions,it is considered that the region will have convective initiation after 0-1h.In the subsequent algorithm verification,taking the area larger than 35 d Bz on the radar map as a reference for the first time,it is found that the algorithm in this thesis has achieved good recognition effect,with an overall hit rate of 95.18% and a false alarm rate of 43.04%;It not only improves the hit rate and reduces the false alarm rate,but also advances the prediction time.Continuing to analyze three typical cases,it is found that the algorithm in this thesis has a good identification effect for well-developed deep convection,but there are some limitations for CI events near many developed or mature convective clouds and some shallow convection.