[S-GD01] Geodesy and Global Geodetic Observing System

[SGD01-02] Accuracy evaluation of airborne gravity data by the three-dimensional least-squares collocation method

*Koji Matsuo¹ (1.Geospatial Information Authority of Japan)

Keywords:Geodesy, Gravity, Geoid

The Geospatial Information Authority of Japan (GSI) has been conducting a nationwide airborne gravity survey since 2019, with the aim of developing a highly accurate gravimetric geoid model for Japan. The target accuracy of the geoid model is about 3 cm. To achieve this, it is necessary to obtain airborne gravity data with an accuracy of 1-2 mGal. Commonly, the accuracy evaluation of the airborne gravity data is done by “cross-over validation” and “comparison with Global Geopotential field Models (GGMs)”. The cross-over validation is a method to evaluate the quality of airborne gravity data by comparing the gravity values at the intersection of multiple survey lines with similar flight altitudes. This method allows us to evaluate the internal consistency (repeatability) of airborne gravity data, it cannot evaluate the absolute certainty of the observed gravity values. On the other hand, the comparison with GGMs can evaluate the absolute certainty of airborne gravity data in principle, but it may be insufficient to perform a precise evaluation of airborne gravity data because the spatial resolution and accuracy of GGMs are not high enough for this use. In this study, we evaluate the absolute accuracy of airborne gravity data by the three-dimensional least-squares collocation (3D-LSC) method. Here, we use both surface gravity data and a GGM (EGM2008). The evaluation areas are the Kanto and Chubu regions. The airborne gravity surveys there were conducted using a TAGS-7 gravimeter, in which the spacings of each main line and validation line are set to 10 km and 50 km, respectively. The flight altitudes were about 5000 m in the Kanto region and about 3000 m in the Chubu region, respectively. Since the quality of surface gravity data is low in coastal ocean zones, we perform the evaluation only over the land area. Consequently, while the RMS of the difference between the airborne gravity data and the EGM2008-derived gravity values are 2.47 mGal in the Kanto region and 3.16 mGal in the Chubu region, that between the airborne gravity data and the 3D-LSC-estimated gravity values are 1.49 mGal in the Kanto region and 1.35 mGal in the Chubu region. Considering that the accuracy of the surface gravity data used is approximately 1.1 mGal (Matsuo and Kuroishi, EPS 2020), it can be said that the accuracy of the airborne gravity data is approximately 1 mGal according to the principle of error propagation. In conclusion, it is confirmed that our airborne gravity data has an accuracy of 1 mGal level and will be useful for developing a 3 cm-accurate geoid model for Japan.