Iron and manganese are commonly present in groundwater supplies used by many water systems. The presence of iron and manganese in the drinking water is not harmful to human bodies. However, higher concentration causes discoloration, staining, turbidity and bad taste problems. It also form iron oxide or manganese dioxide accumulations in pipes. In the present work, low cost methods have been evolved for the removal of iron and manganese from ground water using Rice Husk based Activated Carbon (RHAC) and Sugarcane Baggase based Activated Carbon (SBAC). Exhaustive experiments have been conducted in the Laboratory on sediment columns with variable depths of 10, 20, 30, 40, and 50 cm. Four graded types of sediment collected from the banks of Rivers Brahmani, Koel, Sankha and Budhabalanga were studied for proper removal of suspended solids by maintaining sufficient infiltration rates. It has been observed that the infiltration rate is maintained for the soil column of 50 cm without having any impact of the suspended solids. The Iron and manganese concentrations are found to be removed 100% after passing through the filter material in both the cases.
Precise and reliable suspended sediment load (SSL) prediction models can help ensure the integrated water resources management in a river basin. This study considered two deep learning (DL) models i.e. gated recurrent unit (GRU) and long short term memory (LSTM) for daily SSL modeling in the Missouri River at Omaha, NE gauging station in the United States. The established models are verified with various statistical measures. The assessment of prediction accurateness of the DL models showed that the GRU was the best model in SSL prediction. The coefficient of determination was 0.871 for the GRU and 0.865 for LSTM. Besides, the GRU model has a lesser root mean square error (RMSE) and mean absolute error (MAE) compared to the LSTM. In summary, both the developed DL model can be used competently in SSL modeling. However, GRU utilizes less training constraints and thus uses a lesser amount of memory, performs quicker, and trains faster than LSTM, while LSTM needs additional data to be more accurate.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.