Multiscale Characterization and Correction of Latitude‐Dependent Biases in CSES Magnetic Vector Measurements

Latitude‐dependent systematic biases in satellite magnetic vector measurements can significantly contaminate the estimation of external magnetic fields and electromagnetic induction signals. Unlike previous geomagnetic satellite missions, magnetic vector data from the China Seismo‐Electromagnetic Satellite (CSES) exhibit pronounced latitude‐dependent biases (latitude effect) primarily caused by boom deformation and sensor misalignment, posing challenges for the reliable use of these data in applications. In this study, we present a multiscale and directional framework to systematically characterize and correct the latitude‐dependent biases in CSES magnetic vector measurements. By removing the modeled internal geomagnetic fields and analyzing the residuals in the curvelet domain, we identify the dominant spatial scales and directional features associated with the latitude effect. A targeted multiscale correction strategy is then implemented at characteristic scales and directions. The effectiveness of the proposed correction approach is evaluated using statistical analyses, cumulative probability distributions, spherical harmonic coefficients, and magnitude‐squared coherence (MSC). The corrected data show substantially improved temporal coherence, reduced systematic amplitude distortions, and more stable inducing and induced field coefficients, indicating enhanced physical consistency of the external magnetic field estimates. These results demonstrate that, although the latitude effect is specific to the CSES mission, the proposed multiscale bias analysis method is essential for the reliable use of satellite magnetic vector data involving external field estimation and electromagnetic induction, which shows good potential for space and Earth scientific research.