Infrared Spectroscopy with Variable Decomposition Level Dual-Tree Complex Wavelet Transform for Quantification of Air...

Area of Science:

• Environmental Science
• Atmospheric Chemistry
• Spectroscopy

Background:

• Open-Path Fourier Transform Infrared Spectroscopy (OP-FTIR) is crucial for atmospheric pollutant detection.
• Accurate real-time background spectrum estimation is a major challenge in OP-FTIR.
• Variable environmental conditions complicate gas analysis.

Purpose of the Study:

• To develop an adaptive background spectrum estimation method for OP-FTIR.
• To enhance the accuracy and robustness of atmospheric gas detection.
• To provide a generalizable solution for complex open-path environments.

Main Methods:

• Proposed a variable decomposition level dual-tree complex wavelet transform algorithm.
• Implemented dynamic adjustment of decomposition levels based on spectral bands.
• Incorporated nonlinear compression with L1 regularization for stability.

Main Results:

• Achieved lower root-mean-square errors for C2H4 (7.04%) and C3H6 (19.37%) compared to Legendre polynomial fitting.
• Reduced relative uncertainties for CO2 and CO background gases across three field sites.
• Demonstrated improved retrieval accuracy and enhanced robustness in variable conditions.

Conclusions:

• The proposed method offers a generalizable strategy for background estimation in OP-FTIR.
• This approach supports more accurate and scalable atmospheric gas monitoring.
• The technique enhances the reliability of OP-FTIR for environmental sensing.