IMAGe Brown Bag- Modeling Snow Water Equivalent in the Rocky Mountain Region

07/05/2016 - 12:00pm to 1:00pm
ML - Chapman Room

Colette Smirniotis
San Diego State University

Accurately modeling and predicting snow cover presents a scientific challenge, in part because it is difficult to measure this quantity directly. Thus, indirect measurements are also used to infer snow cover. Given the choice of different indirect measurements and also different algorithms to retrieve snow cover from these data, there is no single, standard gridded snow data set that can be used for climate model evaluation. The goal of this project is to use a Bayesian statistical model to create a blended product of snow water equivalent (SWE) from multiple data sets and in the process also quantify the uncertainty of this across different locations. To combine the data sets, a new version of the R package LatticeKrig is implemented that can handle the “change of support” problem common in spatial statistics and inherent in this project. LatticeKrig models the spatial process with a multiresolution basis function representation in a computationally efficient manner. Monthly snow predictions from the blended product are calculated along with the standard errors of the predictors.

Tuesday, July 5, 2016
12:00 pm - 1:00 pm
Mesa Lab, Chapman Room
(Bring your lunch)