Singular Spectrum Analysis to Correct the Seasonal Variation in the FCDR of FY-3A/B/C VIRR Reflective Solar Bands

The Visible and Infrared Radiometer (VIRR) onboard China’s FY-3A/B/C satellites has delivered essential global earth observations for over 15 years, enabling critical applications in cloud dynamics, vegetation assessment, and environmental monitoring. However, VIRR lacks onboard calibration systems for its visible/near-infrared channels, results in progressive radiometric degradation due to cumulative space radiation and detector aging, challenging the generation of stable long-term climate datasets e.g., the Fengyun satellite climate data records (FCDR). By integrating Simultaneous Nadir Overpass (SNO) cross-calibration with Aqua/MODIS references, we identified pronounced seasonal fluctuations in long-term recalibration coefficients, particularly for the 0.86 m. To isolate these effects, Singular Spectrum Analysis (SSA) decomposed the coefficient series into three components: trend, seasonal fluctuations, and residuals. A hybrid calibration model was then formulated by integrating the isolated trend and seasonal features. Validation across globally distributed pseudo-invariant sites confirmed enhanced radiometric stability. The work highlights the necessity of accounting for seasonally modulated calibration artifacts, which were previously unaddressed in operational protocols, to ensure the stability and accuracy of climate data records (CDRs).