Predicting the Total Discharge from Snowpack Data between the Sierra Nevada Mountains and California’s Central Valley
MIDS Capstone Project Fall 2021

Predicting Total Discharge into California’s Central Valley from the Sierra Nevada Mountains

We illustrate and predict the groundwater discharge, as shown through the Water Table Depth ("catdef") from the Sierra Nevada Mountain Snowpack Reanalysis and other GRACE-DA variables. Using a ConvolutionalLSTM model we provide an 87% accuracy for a geospatial coordinate grid of 33 x 37, an approximate resolution of 36 Km2.  We investigated the variations of snowpack to assess the possible impact on Water Table Depth by varying snowpack input. Finally, we evaluated the benefit of offsetting the Snowpack input on improving our overall model.

Predictors (Snowpack & RZMC), Actual Target, & Predicted Target (CATDEF)
Predictors (Snowpack & RZMC), Actual Target, & Predicted Target (CATDEF)

Last updated:

December 6, 2021