Simultaneously identifying micron-scale aerosols, cloud droplets, and ice crystals within mixed-phase clouds is essential for advancing cloud microphysical research, yet it presents substantial technical challenges. This study proposes a novel phase identification method based on the Ground-based Fog and Aerosol Spectrometer (GFAS; DMT Inc.), which is rigorously validated through cloud chamber experiments and in-situ observations. Cloud chamber experiments conducted at −14 °C, −10 °C, and −6 °C yielded distributions of the polarization ratio (<em>δ</em> = (<em>S</em>−<em>P</em>)/(<em>S</em>+<em>P</em>)) and total scattering counts (TotalCounts = <em>S</em>+<em>P</em>) for highly non-spherical ice crystals. Based on these measurements, quantitative thresholds were established to classify three distinct particle regimes: ice crystals (<em>δ</em> > −0.33, TotalCounts < 1.68 × 10²), water droplets (<em>δ</em> < −0.82, high TotalCounts), and a transition zone (−0.82 ≤ <em>δ</em> ≤ −0.33) corresponding to weakly non-spherical ice or aerosols. Applied to ground-based observations at Mt. Liupan in Northwest China from January to April 2024, the threshold method was applied to a complete in-cloud glaciation cycle. The GFAS effectively distinguishes highly non-spherical ice crystals from pure water droplets, while the TotalCounts threshold achieves a separation efficiency exceeding 98 % in distinguishing pure ice crystals from aerosols. Although aerosol <em>δ</em> values primarily fall within the transition zone (−0.82 and −0.33) and overlap with weakly non-spherical ice crystals, preliminary discrimination remains achievable using a joint constraint of <em>δ</em> and TotalCounts. This method provides a temperature-independent diagnostic tool for quantifying the glaciation degree in mixed-phase clouds, offering critical observational support for understanding cloud microphysical processes.