Marine waters assessment using improved water quality model incorporating machine learning approaches

Area of Science:

• Marine Biology
• Environmental Science
• Water Quality Assessment

Background:

• Marine ecosystems rely on good water quality, influenced by various factors.
• Existing Water Quality Index (WQI) models face uncertainty issues in assessments.
• Accurate WQI is crucial for the health of marine organisms and ecosystems.

Purpose of the Study:

• To introduce and evaluate two novel WQI models: weighted quadratic mean (WQM) and root mean squared (RMS).
• To assess marine water quality in the Bay of Bengal using the proposed WQI models.
• To address and reduce uncertainty in WQI calculations for improved accuracy.

Main Methods:

• Developed and applied two new WQI models: WQM and RMS.
• Utilized seven key water quality indicators: salinity, temperature, pH, transparency, dissolved oxygen, total oxidized nitrogen, and molybdate reactive phosphorus.
• Assessed spatio-temporal water quality in the Bay of Bengal.

Main Results:

• Both WQM and RMS models classified water quality as 'good' to 'fair'.
• No significant difference was observed between the weighted (WQM) and unweighted (RMS) model results.
• The models demonstrated high sensitivity (R² = 1) and effectively reduced WQI uncertainty.

Conclusions:

• The developed WQM and RMS WQI models accurately assess marine water quality.
• These novel approaches offer a more reliable method for WQI calculation, reducing existing uncertainties.
• The study confirms the effectiveness of these WQI models in evaluating the health of marine environments.