Land Hydrology Studies in North America Using GRACE and Hydrology Models
The need for a reliable land hydrology model that can monitor the amount of water stored on and beneath the Earth’s surface on a regional and global scale has become very important, especially in overpopulated areas or regions that already suffer from shortage of freshwater. The main objective of th...
| Main Author: | |
|---|---|
| Other Authors: | , , , , |
| Format: | Doctoral or Postdoctoral Thesis |
| Language: | English |
| Published: |
Schulich School of Engineering
2020
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/1880/111902 https://doi.org/10.11575/PRISM/37727 |
| Summary: | The need for a reliable land hydrology model that can monitor the amount of water stored on and beneath the Earth’s surface on a regional and global scale has become very important, especially in overpopulated areas or regions that already suffer from shortage of freshwater. The main objective of this thesis is to examine the hydrology signal in North America using a combination of land hydrology models and satellite gravimetry products coming from the GRACE satellite mission. Our analysis emphasizes on the post-processing of GRACE data. More specifically, we define a detailed framework for the extraction of hydrological signals from GRACE data by removing the contribution of non-hydrologic geophysical components and using advanced processing techniques. In order to carry out this objective, we improve the most frequently-used filtering methods for the suppression of correlated errors from GRACE data, and develop more refined algorithms for their implementation. We formulate a selective decorrelation of GRACE data using machine learning and show that our new approach mitigates the over-filtering effects of the conventional decorrelation. We also solve the instability and inaccuracy problems related to the calculation of isotropic Gaussian filter coefficients and develop new expressions that simplify their evaluation. We assess the GRACE data and the hydrology models, and find a satisfactory level of agreement between them, with an averaged RMS difference of 3.9 cm in terms of equivalent water height. We then combine these independent datasets and develop two combined hydrology models for the monitoring of monthly terrestrial water storage and groundwater storage variations. We examine their seasonal and long-term variations and provide useful insights for the spatiotemporal evolution of water masses in North America from 2003 to 2014. For the most part, North America is affected by negative long-term trends of terrestrial and ground water changes that are more evident in Hudson Bay and southern North America, ... |
|---|