Abstract: Atmospheric duct is an important environmental factor affecting the performance of radio systems such as radar detection. In order to quantitatively evaluate the oversea propagation effect caused by atmospheric duct, a statistical analysis method of oversea events based on measured AIS propagation data is proposed. This paper concentrates on the observation test of AIS conducted in 2019 along the Yellow Sea coast. The line-of-sight and maximum distance of AIS signal propagation are simulated using the effective earth radius and the parabolic equation method.Based on the annual measured data of AIS maximum distance and the set distance threshold, atmospheric duct trans-horizon events were identified by eliminating those with very short durations and merging those with small interval times in the trans-horizon data. Subsequently, the trans-horizon events were statistically analyzed. The measured data indicates that there were 205 instances of atmospheric duct trans-horizon events in the observed sea area throughout the year, with the longest event lasting 27.7 hours and an average duration of 3.2 hours. Atmospheric duct trans-horizon events are rare during winter but occur more frequently from March to September, with the highest frequency in June, reaching 42 occurrences. This study can provide important data support and a scientific basis for evaluating the potential of duct trans-horizon propagation, optimizing the use of wireless systems, and avoiding abnormal propagation interference.