Correlation-based LiDAR wind sensing using SiPM: methodology and indoor validation

Cross-correlation wind LiDAR estimates wind speed by tracking advected aerosol-backscatter texture, enabling compact, cost-effective systems using intensity returns only. We evaluate silicon photomultiplier (SiPM) processing for this task using a dual-beam, single-receiver architecture by comparing two in-gate intensity estimators: peak amplitude and pulse area. A Monte Carlo SiPM response model is used to compare the two estimators at the device-readout level in the linear-counting regime, and indoor fog-tunnel experiments are used separately to evaluate wind-retrieval performance at the system level. Over the tested (U_ref,N) plane, it achieves a lower mean relative retrieval error (8.75% versus 9.81%) and a larger fraction of operating points below 10% relative error (74.18% versus 62.64%). Under the tested near-linear operating conditions, these results provide quantitative guidance for observable selection in SiPM-based correlation wind retrieval and support the feasibility of compact, fixed-range implementations.