Assessment of Gravity Filtration Techniques for Shipborne Marine Gravity Observations Over the Mediterranean Sea

Marine gravity provides vital geodetic and geophysical information for several applications, such as regional geology, tectonic structures, and oceanographic processes. Filtering these data represents a crucial pre-processing step to eliminate any erroneous signals and improve the data quality for better analysis and interpretation. This study investigates five gravity filtration techniques to improve the quality of shipborne marine gravity data collected over the Mediterranean Sea. The assessed techniques include Least Squares Collocation (LSC), Interquartile Range (IQR), three-sigma limit, Inverse Distance Weighted (IDW) interpolation, and refinement using Geopotential Earth Models (GEMs). The study employs 66229 shipborne gravity anomaly data points acquired from the International Gravimetric Bureau (BGI). The results indicate that the IDW interpolation and LSC techniques are the most effective. IDW interpolation can provide a balance between error detection and computational time, while LSC requires a significantly longer processing time. It is also shown that using any recent GEM yields almost the same performance. Moreover, a cross-check between the assessed techniques shows that they all accepted about 67% of the data as not erroneous. These findings can be exploited to significantly improve the accuracy of marine gravity datasets for different geodetic and geophysical studies.