Automatic Identification and tracking of mesoscale eddies in Persian Gulf by pattern mining approach

Mesoscale Eddies with a diameter of less than 150 km and a time-scale of about one month are in fact rotations (distortions) which are observed in the continuous Spatio-Temporal fluid field as an Evolving pattern over space and time. In this study, using pattern mining approach, the eddy identification automatic algorithm in Persian Gulf called PGES (Persian Gulf Eddy Scan) was developed and implemented in MATLAB software. In order to prepare the input data of mentioned algorithm, at first, the ROMS numerical model with climate forcings, implemented for the Persian Gulf. The algorithm input data includes averaged daily free surface water height, sea surface temperature and salinity with a grid resolution of 5km. Finally, using this algorithm, 48 mesoscale eddies with lifespan of at least 3 weeks were identified and tracked automatically in the Persian Gulf for one year.