SIParCS 2015- Colette Smirniotis

Colette Smirniotis, San Diego State University

Modeling Snow Water Equivalent in the Rocky Mountain Region

(Slides) (Recorded Talk)

As global temperatures increase, there is growing demand to predict future snow more accurately. However, modeling snow is a notoriously challenging process. There is no standard dataset used to evaluate snow model predictions, although there are several observational datasets and data products commonly used for this purpose. Four of these datasets are combined into a blended product using fixed basis functions and a Gaussian Markov random field. Inference and uncertainty quantification of SWE are achieved through conditional simulations of this blended product.

Mentors: Rachel McCrary and Doug Nychka