Quality control of marine big data—a case study of real-time observation station data in Qingdao

Abstract

Off shore waters provide resources for human beings, while on the other hand, threaten them because of marine disasters. Ocean stations are part of off shore observation networks, and the quality of their data is of great signifi cance for exploiting and protecting the ocean. We used hourly mean wave height, temperature, and pressure real-time observation data taken in the Xiaomaidao station (in Qingdao, China) from June 1, 2017, to May 31, 2018, to explore the data quality using eight quality control methods, and to discriminate the most eff ective method for Xiaomaidao station. After using the eight quality control methods, the percentages of the mean wave height, temperature, and pressure data that passed the tests were 89.6%, 88.3%, and 98.6%, respectively. With the marine disaster (wave alarm report) data, the values failed in the test mainly due to the infl uence of aging observation equipment and missing data transmissions. The mean wave height is often aff ected by dynamic marine disasters, so the continuity test method is not eff ective. The correlation test with other related parameters would be more useful for the mean wave height.

Research on quality control methods for surface temperature observations via spatial correlation analysis

“…In recent years, scholars have conducted extensive research on the quality control method of surface meteorological observations, which focused on single‐station quality control under temporal dimension and regional multistation networked quality control under spatial dimension (Tao et al ., 2018; Qian et al ., 2019; Cerlini et al ., 2020; Carminati et al ., 2021). Wade and Barnes proposed the inverse distance weighting method (IDW), which uses the Euclidean distance between the target station and the neighbouring stations to determine the weights (Feng et al ., 2004).…”

Studies on marine dynamic disasters

“…This project also carried out a series of studies around the data of marine dynamic disasters. Qian et al (2019) used hourly real-time ocean station observation data to investigate quality control techniques and methods for determining the most suitable quality control approach for the data collected at the Xiaomaidao observation station in Qingdao, China. The typhoon wind hazard risk for the city of Shenzhen was analyzed using a Monte-Carlo simulation and the YM wind fi eld model (Guo et al, 2019).…”