Related Experiment Video
Updated: May 8, 2025
08:47
Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
Published on: February 9, 2024
1.2K
Prediction of surface water pollution using wavelet transform and 1D-CNN
Gaofeng Wang1, Hao Zhang1, Man Gao1
1College of Electrical and Information Engineering, Beihua University, Jilin 132021, China.
Summary
This study uses wavelet transform and 1D-CNN with UV-vis spectroscopy to accurately predict surface water pollution indicators like permanganate index (CODMn) and nitrogen levels, achieving R values over 0.98.
Area of Science:
- Environmental Science
- Analytical Chemistry
- Spectroscopy
Background:
- Surface water quality is crucial and assessed using indicators like permanganate index (CODMn), total nitrogen, and ammonia nitrogen.
- Accurate and robust methods are needed for monitoring these pollution levels.
Purpose of the Study:
- To develop and validate a spectroscopic method for predicting surface water pollution indicators.
- To evaluate the performance of a 1D-CNN model combined with wavelet preprocessing.
Main Methods:
- Collected 708 surface water samples with varying concentrations.
- Employed ultraviolet-visible (UV-vis) spectroscopy for spectral analysis.
- Utilized wavelet transform for spectral preprocessing and a 1D-CNN for prediction modeling.
Main Results:
- Wavelet transform with a fixed threshold (sqtwolog) optimized spectral data.
- The 1D-CNN model achieved coefficient of determination (R) values exceeding 0.98 on the test dataset.
- 1D-CNN demonstrated superior prediction accuracy and robustness compared to backpropagation and extreme learning machine models.
Conclusions:
- The combination of wavelet transform and 1D-CNN offers a highly accurate and robust approach for predicting surface water pollution indicators.
- This method shows significant practical value for real-time water quality monitoring.
