Exploring Thermospheric Disturbance Patterns Through Space-Borne Accelerometer Measurement Errors: A Weighted Accelerometer 1B Dataset of GRACE C

Satellite measurements are essential for understanding Earth's complex system, yet they often lack a reliable a-priori covariance matrix. This study presents a new methodology to enhance the reliability of satellite measurements by deriving experimental covariance matrices from the original observation. We focus on the accelerometer measurements (1A dataset) from the GRACE (Gravity Recovery and Climate Experiment) C satellite. Using autocorrelation analysis, we create a block-diagonal covariance matrix for the 1A dataset. We then apply a low-pass Gaussian filter that integrates this covariance matrix into the least squares estimation, resulting in a refined 1B dataset that minimizes spikes and spurious accelerations while preserving measurement error. Our variance analysis uncovers disturbances linked to geomagnetic storms and the satellite's transitions through Earth's shadow and terminator, with fluctuations notably peaking during the equinoxes.