[A-CC26] Glaciology

[ACC26-09] Accuracy, limitation and improvement of snow grain size measured using the optical instruments

*Teruo Aoki¹, Akihiro Hachikubo², Motoshi Nishimura¹, Tomonori Tanikawa³ (1.National Institute of Polar Research, 2.Kitami Institute of Technology, 3.Meteorological Research Institute, Japan Meteorological Agency)

Keywords:snow grain size, specific surface area, HISSGraS, IceCube, BET method

Snow grain size and the specific surface area (SSA, the ratio of surface area to mass or volume) of snow grains are important microphysical parameters for governing the surface albedo and for characterizing the evolution of snow metamorphism. Field measurements of these parameters are often performed using optical instruments such as the well-established instrument IceCube (A2 photonics, France) (Gallet et al., 2009). However, conducting multipoint observations, such as satellite validation, with instruments like IceCube is too heavy and time-consuming. Recently, we developed the Handheld Integrating Sphere Snow Grain Sizer (HISSGraS), which can be applied directly to the snow surface, snow pit face, and snow samples collected in a sampling case (Aoki et al., 2023). To evaluate the accuracy of HISSGraS and IceCube, we measured the wide range of SSAs for different types of natural snow and artificial ice grains using these instruments under temperature-controlled conditions. We compared the results with those obtained using the gaseous adsorption method (Brunauer–Emmett–Teller: BET) as a reference method. The measurements were conducted in a cold laboratory at the cold room temperature (T_a) of -20 °C at Kitami Institute of Technology. The measured snow shapes include precipitation particles (PP) reserved at T_a = -50 °C for a few days to one month after the snowfall, the same PP reserved at T_a = -30 °C for one month, and artificial spherical ice particles with different sizes. A total of 35 snow samples were measured. The range of SSA for all snow samples measured with the BET method was from 4.5 to 82.8 m² kg⁻¹ (mean value 51.5 m² kg⁻¹). The SSAs measured with the HISSGraS agreed well with those measured with IceCube (determination coefficient R² = 0.98). In contrast, the SSA measured with the HISSGraS (IceCube) correlated statistically with the BET-derived SSA (R² = 0.80 (0.85)) but were remarkably underestimated for SSA > 60 m² kg⁻¹. This underestimation is likely attributed to the loss of photons emitted from a light source of the optical instruments in the target PP under measurement, where snow density is low and light absorption is small due to small grain size. The correlation in SSA between the HISSGraS (IceCube) and BET methods improved to R² = 0.86 (0.87) when the values for such low-density snow samples were excluded. To improve the accuracy of SSA with optical instruments for the PP samples with snow density lower than 160 kg m^-3, we artificially compressed five PP samples to 2.7 times higher snow density on average. We measured the SSAs with HISSGraS and IceCube and compared them with the BET-derived SSA. The results showed that the SSA measured with HISSGraS and IceCube, which were significantly underestimated for low-density PP, increased to the value 10% lower than the BET-derived SSA of the uncompacted PP samples, whereas the SSA measured with the BET method remained almost the same as before compression. We conclude from these results that the accuracy of SSA derived with the optical instruments can be improved, resulting in an underestimation of 10%, by compressing the snow samples in the case of low-density PP.