Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Etats Unis → 2018 → Understanding groundwater variability : Modeling groundwater storage change in southern California

San Diego State University

Understanding groundwater variability : Modeling groundwater storage change in southern California

Rother, David E

Titre : Understanding groundwater variability : Modeling groundwater storage change in southern California

Auteur : Rother, David E.

Université de soutenance : San Diego State University

Grade : Master of Science (MS) in Geography 2018

Résumé
This thesis details the development, calibration, and application of an atmosphere-land-groundwater model that is used to simulate groundwater storage change in southern California between November 2005 and December 2016. The model consists of three separate components : the Weather Research and Forecasting (WRF) atmospheric model, the Simplified Simple Biosphere (SSiB) model, and a modified version of the Simple Groundwater Model (SIMGM). A full coupling of the WRF/SSiB/SIMGM allows for a complete simulation of the physical and biophysical processes that influence groundwater storage, as well as their interactions with regional climate. The primary dataset used to validate the results of the model is of terrestrial water storage provided by NASA’s Gravity Recovery and Climate Experiment (GRACE) satellites. The analysis of in-situ USGS groundwater well water table depth measurements indicate that GRACE adequately represents groundwater storage fluctuations in the region. A series of physics scheme combinations within the WRF model were tested to determine which most accurately simulated patterns of precipitation and temperature during November 2005 - February 2006. Further sensitivity tests were performed on the SSiB/SIMGM model, specifically on the equation representing groundwater discharge. Time series analyses comparing the atmosphere-land-groundwater model and GRACE observational data suggest that the model tends to simulate terrestrial water storage more accurately during wet periods than in dry, however, it underestimated the intensity of the negative groundwater anomaly compared to the offline model. The statistical tests used to evaluate the model’s performance, including mean bias and correlation, indicate that the WRF/SSiB/SIMGM coupled model captured precipitation and temperature with good accuracy, however, the model underestimated the drought signal displayed in the GRACE terrestrial water storage dataset. Overestimation of precipitation over Arizona and along California’s south-eastern border was the primary cause of excess soil moisture and water storage in these areas.

Présentation et version intégrale

Page publiée le 9 janvier 2022