Keywords
WQ simulation; artificial intelligence; remote sensing; total suspended solids; Lake Victoria
Start Date
6-7-2022 2:20 PM
End Date
6-7-2022 2:40 PM
Abstract
Freshwater lakes are a major resource for human populations. To support water quality (WQ) management for lakes, both WQ monitoring and WQ modeling are essential. WQ variables are traditionally determined by in-situ measurements. Although this method offers high accuracy, it is a costly and time consuming process. Moreover, the sampling method is not easily able to identify the special and temporal WQ variations in lakes. To overcome this limitations, the use of remote sensing (RS), as a promising tool for large-scale inland WQ monitoring, can be adopted. To simulate WQ variables in lakes, several types of models have been developed, including data-driven models and process-based models. As the processbased models require a significant amount of data (meteorological, topographical, hydrological, and WQ data), their applications have been limited. On the other hand, with the development of artificial intelligence (AI) techniques, AI models have been widely applied to model WQ variables in lakes. In this study, we investigate the application of several AI techniques coupled with RS water quality (turbidity) data for prediction of total suspended solids (TSS) in large lake basins. Specifically, the study aims to develop a robust AI model for simulating TSS in Lake Victoria basin, using the basin precipitation data and the TSS from inflow channels into the lake. To develop the AI model, the freely available turbidity data for the lake is used as a reference data. The results indicate that AI-based models are potential tools that can be adopted for TSS predictions in large lake basins. Additionally, this study illustrates the potential of the use of remote sensing data to support model development, as an alternative to in-situ measurements, especially in data-scared regions.
Water quality simulation of lakes: coupling artificial intelligence techniques and remote sensing data
Freshwater lakes are a major resource for human populations. To support water quality (WQ) management for lakes, both WQ monitoring and WQ modeling are essential. WQ variables are traditionally determined by in-situ measurements. Although this method offers high accuracy, it is a costly and time consuming process. Moreover, the sampling method is not easily able to identify the special and temporal WQ variations in lakes. To overcome this limitations, the use of remote sensing (RS), as a promising tool for large-scale inland WQ monitoring, can be adopted. To simulate WQ variables in lakes, several types of models have been developed, including data-driven models and process-based models. As the processbased models require a significant amount of data (meteorological, topographical, hydrological, and WQ data), their applications have been limited. On the other hand, with the development of artificial intelligence (AI) techniques, AI models have been widely applied to model WQ variables in lakes. In this study, we investigate the application of several AI techniques coupled with RS water quality (turbidity) data for prediction of total suspended solids (TSS) in large lake basins. Specifically, the study aims to develop a robust AI model for simulating TSS in Lake Victoria basin, using the basin precipitation data and the TSS from inflow channels into the lake. To develop the AI model, the freely available turbidity data for the lake is used as a reference data. The results indicate that AI-based models are potential tools that can be adopted for TSS predictions in large lake basins. Additionally, this study illustrates the potential of the use of remote sensing data to support model development, as an alternative to in-situ measurements, especially in data-scared regions.
Stream and Session
false