Temperature and grain size dependence of near-IR spectral signature of crystalline water ice: From lab experiments to Enceladus' south pole

An experimental facility has been set up to acquire infrared spectra of pure water ices with controlled grain sizes ranging from 80 to 700 mu m and temperatures ranging from 80 to 140 K at 10(-2) mbar. Forty-seven near IR spectra (1.0-5.0 mu m), selected among the total acquired spectra, permit us to determine how spectral characteristics between 1.0 and 1.8 mu m depend on both grain size and temperature. It will be shown that the conventional spectral characteristics derived from Gaussian fitting cannot be used as good temperature or grains size indicators. On the contrary, spectral band areas allow for a proper determination. In this work, we confirm the strong dependence of the 1.05 mu m, 1.30 mu m and 1.50 mu m band areas on grain size. A new linear scaling law, established from our experimental data, is proposed for grain sizes' estimations. It is also demonstrated that the temperature effect on band areas is much weaker than the grain size effect, because temperature has opposite effects on both sides of two inflection points. However, the "reduced" area, computed between the two inflection points, turns out to be a very good indicator of the ice temperature, especially for the 1.30 pm and 1.50 pm bands. A second linear scaling law, established from the experimental data, is then computed to describe the temperature dependence of these reduced areas. These near-IR spectral properties of crystalline water ice have been established from laboratory experiments. It will be shown that the scaling laws proposed in this work are indeed suitable for retrieving temperature and grain size of laboratory icy samples. But they may also be a good tool for characterizing the icy surfaces of the outer solar system. As an example, the VIMS spectral images of Enceladus' south pole have been used to retrieve both the grain size distribution around the tiger stripes area and the distribution of temperature anomalies along the ridges. The results are in very good agreement with previous studies. (C) 2011 Elsevier Ltd. All rights