Predicting the Ecological Quality of Rivers: A Machine Learning Approach and a What-if Scenarios Tool.
Saved in:
| Title: | Predicting the Ecological Quality of Rivers: A Machine Learning Approach and a What-if Scenarios Tool. |
|---|---|
| Authors: | Politikos, Dimitris1 (AUTHOR) dimpolit@hcmr.gr, Stefanidis, Kostas1 (AUTHOR), Varlas, George1 (AUTHOR), Papadopoulos, Anastasios1 (AUTHOR), Dimitriou, Elias1 (AUTHOR) |
| Source: | Environmental Modeling & Assessment. Dec2024, Vol. 29 Issue 6, p1059-1077. 19p. |
| Subject Terms: | *Ecosystem health, *Water quality, *Bodies of water, *Environmental monitoring, Machine learning |
| Geographic Terms: | Greece |
| Abstract: | Monitoring the ecological status of rivers is essential for protecting freshwater biodiversity and ecosystem health. The main objective of this work was to predict the ecological quality of Greek rivers using a machine learning approach based on the Extreme Gradient Boosting (XGBoost) classifier. We used a dataset that comprises ecological, physicochemical, geomorphological, and sample-related parameters collected from the national monitoring network of Greek rivers as well as climate parameters from the ERA5-Land dataset. More specifically, we developed multiple models that predicted the ecological quality class derived by four quality elements (QEs) that are benthic macroinvertebrates, benthic diatoms, fish, and physicochemical quality. The Shapley Additive exPlanations (SHAP) approach was implemented for quantifying the contributions of the predictors on the quality class. We finally developed a web interface tool that can simulate what-if scenarios to predict the quality class under altered environmental conditions. Our findings showed that total phosphorus, nitrate, and ammonium were important predictors for benthic macroinvertebrates, benthic diatoms, and physicochemical quality, whereas for fish, predictors related with the geomorphology (e.g., altitude and slope) had a higher influence. The SHAP plots revealed the synergistic effect of predictors on the quality classes, highlighting a negative effect of increased nutrients on achieving the good quality class based on macroinvertebrates and diatoms and a positive relationship between altitude and slope with the good ecological quality class based on fish. Furthermore, the web interface could provide a useful tool for water managers to predict quality classes for water bodies under what-if scenarios. [ABSTRACT FROM AUTHOR] |
| Copyright of Environmental Modeling & Assessment is the property of Springer Nature and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | GreenFILE |
|
Full text is not displayed to guests.
Login for full access.
|
|
Be the first to leave a comment!
