Adaptive shipborne gravimetry method holds 0.771 mGal accuracy through GNSS outages

Researchers from the Naval University of Engineering in Wuhan and Yuncheng University, China, have published an open-access method for shipborne strapdown gravimetry that copes with GNSS signal problems. The paper appeared in Communications Engineering on 5 September 2026.

The method switches between two modes. During GNSS outages it integrates the strapdown inertial navigation system with a pseudo log (SINS/PL), using a covariance-matching adaptive Kalman filter. During interference, or when signals are normal, it uses SINS/GNSS integration with a linear Huber-M robust Kalman filter.

Marine experiments covering both outage and interference scenarios gave an overall accuracy of 0.771 mGal. The authors say the approach outputs gravity data in real time without added equipment or operating cost, and is more computationally efficient and robust than existing methods.

Why it matters: Marine gravity surveys rely on satellite positioning that can be jammed or lost. A method that keeps sub-mGal accuracy through such events without extra hardware could make offshore gravity and geoid-related work more dependable for survey operators.