Contact probe spectroscopy of snowpits in Finland

Microwave remote sensing of snow relies on knowledge of snow properties such as grain size for the inversion from radiances to snow mass, and numerical snow models could provide this information. However, the relationship between snow's physical grain size and its radiative properties is not necessarily obvious due to the variety in shape, size and conglomeration of snow crystals. Using the contact probe spectroscopy method suggested by Painter et al in 2007, optical grain sizes are estimated for a series of snowpits in Sodankylä, Finland (67.4° N, 26.6° E). At each location traditional microscope measurements of grain sizes, shapes and degree of conglomeration were also taken, providing over 300 pairs of measurements for comparison. Results are consistent with a linear relationship between the two methods of grain size estimation. It is found that in the wavelength range considered (950-1080 nm) the amount of conglomeration between grains has no detectable effect on the optical grain size, suggesting that modelling this process is not necessary for remote sensing at these wavelengths. The predominance of faceted crystals and prism/cup hoar at the field site limits the opportunities for exploring the effect of different grain shapes, and evidence for a difference in optical properties between hoar and faceted crystals is weak, although there are indications that microscope measurements of the extent of smaller grains in a sample is a better predictor of optical properties for depth hoar. Errors related to the spectroscopic method and spatial variation in the snow are quantified experimentally, a relatively novel approach compared with the theoretical SNR arguments applied in earlier work by Nolin & Dozier and Painter et al. These results provide new tests for physically based snow models which will be an important step towards improving snow remote sensing products that can use these models to improve their inversion process.