Merging datasets from NASA SnowEx to better understand how snow falls and moves.

A modelers ability to estimated snow depth and snow water equivalent (SWE) for mountain snow depends strongly on their understanding of snowpack spatial distribution and the soundness of the models initial conditions. This work focuses on the application of data science and efficient algorithms to find optimal locations in mountainous terrain to act as initital conditions for machine driven interpolation methods. By using graphs, collections of pairwise relationships between objects, to describe our data we were able to efficiently search, partition, and understand snowpack from the persepetive of this data structure.