CORRELATION STUDY OF MICRO-UAV THERMAL IMAGERY ON SEA SURFACE TEMPERATURE (SST) DISTRIBUTION PATTERNS FOR SCHOOLING FISH DENSITY DETECTION

The success of precision fishing methods using Unmanned Aerial Vehicles (UAVs) heavily relies on the sensor's accuracy in detecting thermal anomalies on the water surface. The dynamic characteristics of coastal airspace present mechanical challenges in the form of turbulent flow, which affects vehicle stability and the reading accuracy of the drone's thermal radiation sensors. This study analyses the correlation between the thermal radiation intensity captured by a micro-UAV camera (DJI Mavic 4 Pro Thermal) and actual Sea Surface Temperature (SST) patterns, as well as evaluates the influence of coastal aerodynamics. Field experiments were conducted in the coastal waters of Sumur Beach, Pandeglang, Banten, totaling 50 sorties from February 1, 2026, to May 1, 2026. Through thermodynamics and fluid mechanics approaches, this research models atmospheric transmission corrections and analyzes image distortion caused by variations in the Reynolds Number ($Re$). The results indicate that coastal wind speeds exceeding 10 m/s increase Re into the turbulent regime, which degrades the spatial quality of thermal imagery due to mechanical vibrations. With the application of radiation corrections, the accuracy of schooling fish detection using UAVs reached 94%, far outperforming conventional satellite mapping resolutions that are often biased by low-resolution cloud cover. This research concludes by formulating the optimal operational envelope for micro-UAVs in supporting precision fishing.