Estimating maritime snow density from seasonal climate variables.

Snow density is a complex parameter that influences thermal, optical and mechanical snow properties and processes. Depth-integrated properties of snowpacks, including snow density, remain very difficult to obtain remotely. Observations of snow density are therefore limited to in-situ point locations. In maritime snowfields such as those in Australia and in parts of the western US, snow densification rates are enhanced and inter-annual variability is high compared to continental snow regions. In-situ snow observation networks in maritime climates often cannot characterise the variability in snowpack properties at spatial and temporal resolutions required for many modelling and observations-based applications. Regionalised density-time curves are commonly used to approximate snow densities over broad areas. However, these relationships have limited spatial applicability and do not allow for interannual variability in densification rates, which are important in maritime environments. Physically-based density models are relatively complex and rely on empirical algorithms derived from limited observations, which may not represent the variability observed in maritime snow. In this study, seasonal climate factors were used to estimate late season snow densities using multiple linear regressions. Daily snow density estimates were then obtained by projecting linearly to fresh snow densities at the start of the season. When applied spatially, the daily snow density fields compare well to in-situ observations across multiple sites in Australia, and provide a new method for extrapolating existing snow density datasets in maritime snow environments. While the relatively simple algorithm for estimating snow densities has been used in this study to constrain snowmelt rates in a temperature-index model, the estimates may be also be used to incorporate variability in snow depth to snow water equivalent conversion.