Recently, the team of Professor Zheng Yi of the School of Environmental Science and Engineering of Southern University of Science and Technology published the latest research results in the top journal “Water Resources Research” in the field of water resources. According to reports, the research results solve the technical bottleneck encountered by the environmental Internet of Things in the application of water environment monitoring and management.
In the study, Han Feng, Research Associate Professor of the School of Environmental Science and Engineering, SUSTech, was the first author of the paper, Hu Zhaoping, a master’s student, was the second author, Zheng Yi was the corresponding author, and the co-authors included Professor Chen Nengwang of Xiamen University.
Research diagram Courtesy of SUSTech
Online sensors are the most sensitive “sensory systems” of the environmental Internet of Things, which can provide high-frequency, near-real-time data for environmental monitoring and management. In the management of river basin water environment, compared with the traditional manual sampling-laboratory analysis, online sensors have outstanding advantages in cost and timeliness. However, sensors installed in natural water bodies are susceptible to biofouling, background ions, and other factors, are difficult to maintain and calibrate, and data errors are often significant. Therefore, how to fully and effectively use online sensor data in water environment monitoring, early warning and management has become a technical bottleneck in real work.
In this regard, the research team developed a multi-source data assimilation method (BCMSO) based on strict Bayesian analysis, which uses the watershed water quality model as the platform and integrates conventional water quality monitoring data (low frequency, high cost, low error) and online sensor data (high frequency, low cost, high error) to significantly improve the ability of river water quality monitoring and forecasting.
Through rigorous mathematical derivation, this study theoretically proves the rationality of BCMSO, and reveals through numerical experiments that BCMSO can effectively eliminate the influence of sensor data error on water quality forecasting, and significantly reduce the uncertainty of water quality forecasting. In Zhangzhou City, Fujian Province, China, which is an important honeyfruit producing area, agricultural nitrogen non-point source pollution is prominent, and the BCMSO method has been further used in the Fengpuxi River Basin of the province for the management of nitrogen pollution.
The research team of Xiamen University has observed and studied the basin for many years, and installed online sensors at the exit of the basin to obtain conductivity data at high frequencies (every 20 minutes). There is a significant positive correlation between conductivity data and nitrate concentration, which can be used as an alternative observation parameter for river nitrate concentration. Based on the BCMSO method, the assimilation of conductivity data by SWAT water quality model is realized, which makes the prediction of nitrate concentration in the outlet section of the basin more accurate and significantly reduces the uncertainty range of the forecast value. This reduction in uncertainty can reduce the installation cost of nitrogen management safety margin in the basin from 106 million yuan/year to 34 million yuan/year.
According to reports, the study not only solves the important technical problems in the application of environmental Internet of Things, but also shows the great potential and broad prospects of environmental big data for environmental management, which is of great significance for the current construction of digital twin basins and smart environmental protection in China. (Source: China Science News, Diao Wenhui)
Related paper information:https://doi.org/10.1029/2022WR033673