Spatial calibration and uncertainty reduction of the SWAT model using multiple remotely sensed data.

Area of Science:

• Hydrology
• Environmental Science
• Remote Sensing

Background:

• Remotely sensed products are valuable for watershed modeling, offering spatial and temporal data to improve predictions and reduce uncertainty.
• Their application and effectiveness across diverse regions require further investigation.

Purpose of the Study:

• To evaluate the impact of integrating remotely sensed evapotranspiration (RS-ET) and leaf area index (RS-LAI) on watershed model predictions.
• To reduce parameter uncertainty at the watershed scale and improve spatial predictions of ET and LAI at the sub-watershed level.

Main Methods:

• Utilized the Soil and Water Assessment Tool (SWAT) model for watershed simulations.
• Applied a systematic calibration process, initially using streamflow data, then progressively incorporating RS-ET and RS-LAI as constraints.
• Assessed model performance based on streamflow, RS-ET, and RS-LAI spatial patterns.

Main Results:

• Initial calibration with streamflow and RS-ET yielded 14 satisfactory parameter sets.
• Including RS-LAI reduced the satisfactory parameter sets to six.
• Only three of these parameter sets accurately captured the spatial distribution of ET and LAI at the sub-watershed scale.

Conclusions:

• Integrating multiple remotely sensed products significantly reduces parameter uncertainty in watershed models.
• This approach enhances the reliability of simulating intra-watershed hydrological processes.
• The study broadens the applicability and utility of remote sensing data in watershed modeling.