Data-Driven Analysis of Water Quality Dynamics: A Abstract Case Study of Brisbane River, Australia

Water quality monitoring is essential for maintaining ecosystem health and ensuring sustainable water resource management. Conventional approaches based on grab sampling and laboratory analysis are often costly, time-consuming, and inadequate for capturing rapid or large-scale variations. This study presents an integrated framework that combines Internet of Things (IoT) sensors, remote sensing (RS) observations, and machine learning (ML) models for evaluating and predicting water quality in the Brisbane River. A dataset comprising key physicochemical and biological parameters- pH, turbidity, dissolved oxygen (DO), and chlorophyll-a, was analyzed to assess both seasonal dynamics and short-term fluctuations. Several ML algorithms, including Random Forest, XGBoost, and Support Vector Regression, were applied to predict DO and chlorophyll-a concentrations. The models achieved high accuracy, with XGBoost showing superior performance for temporal forecasting, while RS-derived inputs enhanced spatial generalization. The integration of IoT, RS, and ML facilitated the detection of spatiotemporal patterns and provided real-time insights, demonstrating the potential of hybrid computational approaches in supporting adaptive water management strategies. These findings contribute to the development of scalable, data-driven solutions for monitoring aquatic environments under increasing anthropogenic and climatic pressures.

کلیدواژه ها:

Water quality monitoring ، Machine learning ، Remote sensing ، Internet of Things (IoT) ، Brisbane River