[Fast and accurate extraction of ring-down time in cavity ring-down spectroscopy]

Area of Science:

• Atmospheric chemistry and spectroscopy
• Development of analytical algorithms

Context:

• Cavity ring-down spectroscopy (CRDS) is a powerful technique for measuring atmospheric trace gases like the NO3 radical.
• Accurate and efficient extraction of ring-down time is crucial for high-precision and high-speed CRDS measurements.

Purpose:

• To compare the performance of five common algorithms (FFT, DFT, LRS, LM, LS) for extracting ring-down time from CRDS signals.
• To evaluate algorithms based on noise immunity, accuracy, precision, fitting speed, and optimal waveform length.

Summary:

• Simulations and experiments show that the linear regression of the sum (LRS) and Levenberg-Marquardt (LM) algorithms offer superior noise immunity and accuracy.
• The LRS algorithm demonstrates significantly faster fitting speeds compared to LM, with comparable accuracy under experimental conditions (0.2% noise level).
• An optimal fitting waveform length of five to ten times the ring-down time minimizes fitting result standard deviation for all tested algorithms.

Impact:

• The LRS algorithm is identified as a desirable and efficient method for processing CRDS data, particularly for NO3 radical measurements in atmospheric research.
• This research contributes to improving the speed and accuracy of atmospheric gas concentration monitoring.