Spectral attenuation and aerosol particle size distribution

Area of Science:

• Atmospheric Science
• Optical Physics
• Numerical Analysis

Background:

• Accurate particle size distribution (PSD) is crucial for understanding aerosol properties and their impact on climate and visibility.
• Reconstructing PSD from spectral attenuation data presents an inverse problem, often ill-posed and sensitive to noise.
• Existing methods may lack robust parameter selection or applicability to complex aerosol systems like marine aerosols.

Purpose of the Study:

• To develop and validate a numerical algorithm for reconstructing particle size distribution (PSD) from spectral attenuation data.
• To introduce a method for selecting the regularization parameter based on an objective validity estimate.
• To assess the applicability and accuracy of the proposed method for marine aerosol PSD reconstruction.

Main Methods:

• The study employs a regularization technique to solve the inverse problem of PSD reconstruction.
• A novel method for choosing the regularization parameter is proposed, based on objective validity estimation of the solution.
• Numerical experiments are conducted to test the algorithm's performance and estimate accuracy within specific radius intervals.

Main Results:

• The numerical algorithm demonstrates applicability to reconstructing the PSD of marine aerosols.
• An objective method for selecting the regularization parameter improves the reliability of the reconstructed PSD.
• The study estimates the achievable accuracy of the distribution function within a defined radius interval, considering factors like spectral range and data point distribution.

Conclusions:

• The regularization-based numerical method is effective for reconstructing particle size distributions from spectral attenuation data.
• The proposed objective validity estimation provides a reliable approach for parameter selection in inverse problems.
• The method's performance is validated for marine aerosols, with insights into extending the applicable radius interval through extrapolation.