GC-IOK: An Iterative Optimization Kriging with Covariance and Continuity Constraints of Gravity Characteristics

A high-fidelity gravity background field is essential for gravity-aided navigation. Existing gravity models, which predominantly rely on satellite gravimetry, are often insufficient in terms of both accuracy and spatial resolution for practical navigation applications. It is therefore crucial to enhance these models with near-surface gravimetric measurements. While spatial interpolation is commonly used to grid such observations, current approaches suffer from significant shortcomings: function-fitting methods prioritize mathematical optimization over the physical structure of the gravity field, while conventional Kriging techniques do not adequately incorporate spatial continuity between adjacent grid points. To overcome these limitations, this paper proposes a novel Gravity‑Characteristics Iterative Optimization Kriging (GC‑IOK) method, which explicitly integrates the spatial covariance and continuity properties of the gravity field. The approach employs a local gravity anomaly covariance function to quantify stochastic uncertainty in Kriging interpolation and further utilizes the continuous distribution characteristics of the field to iteratively refine local gridding results, thereby improving overall model accuracy. Validation was conducted using EIGEN‑6C4 model data across diverse geomorphological regions in China—including deserts, plateaus, karst mountains, oceans, and plains. Results show that in gravity backgrounds rich in local extrema and dominated by high‑frequency signals, the proposed method more effectively captures local continuity and reduces the average RMSE by 0.3–0.7 mGal compared to Ordinary Kriging. This study thus provides a transferable framework for high‑fidelity grid‑based gravity background modeling.