Wind lidar signal denoising method based on singular value decomposition and variational mode decomposition.

Area of Science:

• Atmospheric Science
• Optical Remote Sensing
• Signal Processing

Background:

• Wind lidar systems are essential for atmospheric measurements but are susceptible to noise, which can degrade performance.
• Traditional denoising methods may not adequately address the complex noise characteristics in wind lidar signals.
• Improving signal-to-noise ratio (SNR) and detection range is critical for reliable wind speed estimation.

Purpose of the Study:

• To propose and evaluate a novel denoising method for wind lidar signals.
• To assess the performance of Singular Value Decomposition (SVD) and Variational Mode Decomposition (VMD) individually and in combination (VMD-SVD).
• To demonstrate the effectiveness of the VMD-SVD method in improving detection range and wind speed accuracy.

Main Methods:

• A denoising approach integrating Singular Value Decomposition (SVD) and Variational Mode Decomposition (VMD) was developed.
• Performance evaluation was conducted using a covariance matrix-based lidar signal simulation model.
• The VMD-SVD method was applied to actual wind lidar signals for validation.

Main Results:

• The VMD-SVD method demonstrated superior performance compared to VMD and SVD alone.
• An output SNR improvement of approximately 12 dB was achieved at an input SNR of -9 dB.
• Application to actual lidar data showed increased detection range and accurate wind speed estimation, even with fewer pulse accumulations.

Conclusions:

• The combined VMD-SVD denoising method significantly enhances wind lidar performance.
• The technique improves detection range and maintains wind speed accuracy, offering a substantial advantage over existing methods.
• The VMD-SVD approach allows for achieving greater detection distances with improved temporal resolution, increasing efficiency.