A Causal Framework for Understanding Water Quality Dynamics and Ecosystem Resilience

Water quality in freshwater ecosystem is essential to preserve biodiversity to ensure ecosystem resilience. However, one of the biggest challenges for environmental monitoring and management is understanding the complex relationships between water quality parameters like temperature, pH, conductance, phosphates, alkalinity, nitrates, dissolved oxygen, etc. This study investigates the interdependencies between these factors using causal inference techniques which focuses on identifying how changes in any of the factors influence others which ultimately affects overall health of ecosystem. This research study explores a causal model using observational data of water, exposing the direct and indirect relationship between water quality indicators, particularly highlighting the factors impacting dissolved oxygen levels, which plays a crucial role in supporting aquatic life. This study shows the use of causal inference model based on SHAP and LIME which can be used for sustainable decision-making of freshwater resources.