ANNOUNCEMENTS
This study investigates the application of remote sensing techniques and Machine Learning approach for water quality analysis and algal bloom detection. Exploratory Data Analysis was performed for a lake in Chandpur , Faridabad using in-situ water quality data measurements. Through a correlation analysis between ground-based water quality data and Sentinel-2 satellite band values, the best-fit bands were identified based on the highest correlation values for various parameters such as pH, dissolved oxygen, conductivity, turbidity, nitrate, and phosphate. Subsequently, another objective was to analyse and compare two primary approaches for algal bloom detection: an index-based method using the Surface Algal Bloom Index (SABI) and a classification-based method utilizing the Random Forest (RF) algorithm. The SABI and RF algorithm were applied to map algal blooms during both summer (April-May) and winter (October-November) months. The results revealed consistent overall accuracy for both seasons, with the SABI achieving 93.92% accuracy and the RF algorithm achieving 98.33% accuracy for summer, and 93.92% and 99.40% accuracy, respectively, for winter. The analysis also uncovered significant seasonal variations in algal bloom intensity, with higher blooms observed during summer attributed to favorable environmental conditions such as warmer temperatures and increased nutrient availability. Conversely, lower algal bloom intensity was observed during winter months due to cooler temperatures and reduced sunlight. The study concludes that while both SABI and RF algorithm proved effective in detecting algal blooms, the RF algorithm demonstrated greater robustness and accuracy across different seasons, making it a more reliable tool for continuous monitoring and management of algal blooms in water bodies. These findings contribute to a deeper understanding of water quality dynamics and highlight the potential of remote sensing and machine learning technologies in environmental monitoring and management initiatives.
Keywords: Water quality , EDA , Machine Learning , Random Forest , Accuracy assessment.
