Assessing data assimilation frameworks for using multi-mission satellite products in a hydrological context.

Area of Science:

• Hydrology
• Remote Sensing
• Data Assimilation

Background:

• Satellite remote sensing provides vast datasets for improving hydrological process understanding.
• Data assimilation integrates observations into numerical models using dynamic or data-driven methods.

Purpose of the Study:

• To assess Square Root Analysis (SQRA) and Kalman-Takens data assimilation frameworks.
• To integrate satellite measurements (GRACE TWS, AMSR-E, SMOS soil moisture) into a hydrological model.
• To compare the performance of different satellite data assimilation combinations.

Main Methods:

• Implemented a Square Root Analysis (SQRA) filtering scheme.
• Utilized the data-driven Kalman-Takens technique.
• Assimilated Gravity Recovery And Climate Experiment (GRACE) terrestrial water storage (TWS) and soil moisture data from AMSR-E and SMOS.

Main Results:

• Simultaneous assimilation of satellite data by SQRA or Kalman-Takens significantly improved hydrological states.
• Results showed better agreement with independent in-situ measurements.
• The Kalman-Takens approach achieved comparable results to the dynamical SQRA method with lower computational expense.

Conclusions:

• Both SQRA and Kalman-Takens are effective for hydrological data assimilation.
• The Kalman-Takens method offers a computationally efficient alternative to traditional dynamical approaches.
• Combined satellite data assimilation enhances hydrological state estimation.