A parameter identifiability and estimation study in Yesilirmak River

Area of Science:

• Environmental Science
• Water Resource Management
• Environmental Modeling

Background:

• River water quality models require accurate parameter estimation, a complex process with significant challenges.
• Existing models often face difficulties in parameter calibration due to large numbers of variables and data limitations.

Purpose of the Study:

• To systematically calibrate and validate a river water quality model using dynamic field data.
• To identify and estimate key model parameters for improved simulation accuracy.
• To assess the effectiveness of an integrated optimization algorithm for parameter estimation.

Main Methods:

• A QUAL2E-derived model using continuous stirred-tank reactors (CSTRs) for eleven pollution constituents was employed.
• Parameter identifiability was assessed using sensitivity analysis and collinearity index, identifying 8 key parameters.
• An integration-based optimization algorithm coupled with sequential quadratic programming was used for parameter estimation.
• Dynamic field data from the Yesilirmak River in Turkey were collected and utilized for calibration and validation.

Main Results:

• The analysis identified 8 parameters within the identifiable range for the water quality model.
• The integrated optimization algorithm successfully estimated the model parameters.
• Calibrated model predictions showed good agreement with observed river water quality data from the Yesilirmak River.

Conclusions:

• The proposed systematic calibration and validation procedure effectively estimates model parameters for river water quality simulation.
• The method provides a reliable approach for dynamic simulation of river water quality, crucial for effective water resource management.
• Accurate parameter estimation enhances the predictive capability of water quality models, aiding in environmental protection efforts.