PREDICTION OF TEMPORAL DISSOLVED OXYGEN CONCENTRATIONS IN A LAKE USING REMOTE SENSING AND MACHINE LEARNING

Date

2024-8-14

Author

Ünalan, Utku Berkalp

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

122
views

232
downloads

Cite This

Dissolved Oxygen (DO) levels are vital for aquatic life, especially under the stress of climate change, making their monitoring essential for effective lake management. However, local measurements are often costly and time-consuming, whether they are with field campaigns or gauges. This study develops an approach and evaluates the feasibility of using remote sensing and machine learning techniques for temporal monitoring and estimation of DO concentrations in a shallow eutrophic lake. As DO cannot be directly measured with optical sensors, the study first identifies optically sensitive parameters that correlate with ground DO measurements. Here, chlorophyll-a (Chl-a), temperature, and depth are found to be statistically significant in predicting DO. Then, the explanatory parameters are predicted using remotely sensed data. Finally, the developed approach achieved an R² of 0.89 and 0.64 and a mean absolute error of 0.81 mg/L and 1.29 mg/L, for locally measured and predicted test data sets. Results indicated a high potential for estimating the non-optical DO parameter with selected optical parameters. The developed approach presents an alternative to model continuous temporal variations in DO, which may make decision-making on lake management more sustainable.

Subject Keywords

Remote Sensing, Machine Learning, Dissolved Oxygen, Remote Sensing of Water Quality

URI

https://hdl.handle.net/11511/111004

Collections

Graduate School of Natural and Applied Sciences, Thesis