Identification of the Atmospheric Vortex Shedding Behind the Jeju Island Based on MODIS and ERA5 Data

A fascinating phenomenon often occurs as pairs of vortices in two rows on the leeward of some islands. The satellite remote sensing has been a major facilitator to observe this phenomenon. In this study, we conduct a statistical analysis on the occurrence of vortex street by processing 20-year cloud images from the MODIS remote sensing data around Jeju Island in the northwest of Pacific Ocean. The criteria for forming an atmospheric vortex shedding (VS) are also discussed with ERA5 hourly data. The results indicate that the steady upstream flow reaching a certain speed under a strong temperature inversion is beneficial to the formation of VS. Based on the speed of upstream flow and the intensity of temperature inversion which are selected by machine learning methods, fitting lines are constructed to predict the occurrence of vortices, with an accuracy of more than 98%. This fitting relation can probably be used to parameterize the small-scale VS behind mountains in the meso-scale atmosphere simulation system to improve its accuracy.