Algorithm Development for Quality Control of RangeFinder Wave Time Series Data at Ocean Research Station

Abstract

Ocean Research Station is a comprehensive ocean research infrastructure built to collect and provide real-time data with state-of-the-art observation systems. This station operates various research facilities, and the data is actively used for research in each field. In particular, since wave data is produced through the remote-sensing observation method, they are valuable as observation data. Unlike the direct observation method, they can stably collect data and store long-term observation data on severe weather. However, there is a limitation that outlier may occur due to the process of measuring and collecting data or the effects of temporary environmental changes. This study improves the remote wave data quality of MIROS RangeFinder installed at the Socheongcho Ocean Research Station using an algorithm and confirms the reliability by comparing it with the wave buoy observation data operated by the Korea Meteorological Administration. The algorithm used is a Spike removal median filter, which is applied after adjusting the conditions and variables according to the characteristics of the Socheongcho wave data. Also, we developed a new filter using a variable threshold to increase the outlier detection rate. As a result, the final algorithm shows a high outlier detection rate, but some outliers are still not removed, so improvements are required to enhance the algorithm’s performance. In addition, continuous research on the remote-sensing wave observation method is required to compensate for the limitation that the accuracy is slightly lower than that of the direct observation method.