Paper detail

Ordering dynamics of snow under isothermal conditions

We have investigated the morphological evolution of laboratory new snow under isothermal conditions at different temperatures $T=-3,-9,-19 ^{\circ}C$ by means of X-ray tomography. The collective dynamics of the bicontinuous ice-vapor system is monitored by the evolution of the two-point (density) correlation function $C(\mathbf{r},t)$ and a particular thickness distribution which is similar to a pore size distribution. We observe the absence of dynamic scaling and reveal fundamentally different classes of length scales: The first class comprises the mean ice thickness and the (inverse) specific surface area (measured per ice volume) which increase monotonically and follow a power law. The dynamic exponent is in accordance with coarsening of a locally conserved order parameter. A second class of length scales is derived from the slopes of $C(\mathbf{r},t)$ at the origin in different coordinate directions. All these scales show a slower growth with anomalous power law scaling. The existence of two different power laws is quantitatively consistent with coarsening of fractal aggregates and reveal the persistence of power law correlations in the initial condition where the snow contains dendritic structures. At low temperatures these structures persist even over an entire year. A third class of length scales can be defined by the first zero crossings of $C(\mathbf{r},t)$. The zero crossings display a non-monotonic evolution with a strong anisotropy between the direction of gravity and horizontal directions. We attribute this behavior to larger scale structural relaxations of the ice network which apparently leave the small scale interfacial relaxations unaffected. However, vice versa it remains the question how structural mobility is induced by interfacial coarsening.