A Study on Improved Langley Plot Calibration Methods Using Noise Reduction for Field Solar Spectral Irradiance (SSI) Observation Instruments

Accurate spectral and radiometric calibration is critical for precise Solar Spectral Irradiance (SSI) and Aerosol Optical Depth (AOD) retrievals in ground-based observations. This study introduces a pixel-based real-time noise deduction method and evaluates its performance using laser sources, Fraunhofer dark lines, and an improved Langley plot calibration. The proposed approach addresses challenges in long-term field SSI monitoring, including spectral noise variation and frequent calibration requirements for wavelength and responsivity corrections. The pixel-based noise deduction method effectively suppresses spectral dark noise to 0 ± 0.890, outperforming temperature-based corrections by 0.6%. Wavelength accuracy tests with laser sources and Fraunhofer dark lines demonstrate high consistency, with δλ < 0.3 nm, while spectral calibration uncertainty is assessed at 0.195 nm to 0.299 nm. The improved Langley plot achieves spectral responsivity differing by only 0.80% from the standard Langley plot and enhances AOD correlation with CE318 by 0.9–2.7% (RMSE: 0.002–0.003), significantly improving AOD observation accuracy. This work advances the development of field SSI hyperspectral observation and calibration, improving the accuracy of SSI and AOD measurements and contributing to the study of environmental changes and climate dynamics.