Search

Quantifying snow grain size is crucial to analyze radiative transfer and mechanical interactions in the snow cover. We present a nondestructive method for fast measurements of snow optically equivalent diameter (OED). The method consists of diffuse near-infrared reflectance measurements by a compact integrating sphere setup to derive OED. This principle is realized in the handheld InfraSnow instrument. The correlation between snow OED and reflectance is calculated by applying Monte Carlo ray tracing to a 3-D implementation of the measurement geometry. Including the geometrical boundary conditions is essential to obtain a good agreement between modeled and measured InfraSnow reflectance values. In addition to InfraSnow reflectance, snow density is required as second input parameter to the OED analysis. Our InfraSnow OED measurements agree with reference OED measurements by micro computed tomography (micro-CT) within 25% for seven of the ten tested snow blocks. Furthermore, the relative differences between both measurement methods are close to the estimated uncertainties of the InfraSnow methodology. If density is measured by micro-CT and then used as InfraSnow model input to derive OED, an average agreement with the reference micro-CT OED values within 13% is found. If density is measured by a permittivity sensor, the average agreement is within 20%. [ABSTRACT FROM PUBLISHER]

Abstract: Snow density is one of the key properties to characterize a snow cover. We present diffuse near-infrared transmittance measurements with an integrating sphere setup in the laboratory. We analyze 8 snow samples taken from melt forms, decomposed, rounded, faceted and machine made snow. Reference measurements of specific surface area (optically equivalent grain size) and density are done by micro-computed tomography and used as input for transmittance calculations. A diffuse flux extinction model cannot be applied to simulate transmittance as our setup cannot be approximated by an infinite snow block thickness. Calculations with a more intricate radiative transfer model (DISORT) agree with our measurements within the estimated grain size and density variability for all probed natural snow types. Only our machine made snow shows a morphology which cannot be modeled by DISORT. Thus, our results show for the first time a direct experimental correlation between transmittance and snow specific surface area and density without the need for an empirical fitting parameter. We feel this to be an important step towards a possible high-resolution, quantitative optical measurement method to determine snow density in combination with an independent specific surface area measurement. [ABSTRACT FROM AUTHOR]