Knowledge of pore-water pressure(PWP)variation is fundamental for slope stability.A precise prediction of PWP is difficult due to complex physical mechanisms and in situ natural variability.To explore the applicabilit...Knowledge of pore-water pressure(PWP)variation is fundamental for slope stability.A precise prediction of PWP is difficult due to complex physical mechanisms and in situ natural variability.To explore the applicability and advantages of recurrent neural networks(RNNs)on PWP prediction,three variants of RNNs,i.e.,standard RNN,long short-term memory(LSTM)and gated recurrent unit(GRU)are adopted and compared with a traditional static artificial neural network(ANN),i.e.,multi-layer perceptron(MLP).Measurements of rainfall and PWP of representative piezometers from a fully instrumented natural slope in Hong Kong are used to establish the prediction models.The coefficient of determination(R^2)and root mean square error(RMSE)are used for model evaluations.The influence of input time series length on the model performance is investigated.The results reveal that MLP can provide acceptable performance but is not robust.The uncertainty bounds of RMSE of the MLP model range from 0.24 kPa to 1.12 k Pa for the selected two piezometers.The standard RNN can perform better but the robustness is slightly affected when there are significant time lags between PWP changes and rainfall.The GRU and LSTM models can provide more precise and robust predictions than the standard RNN.The effects of the hidden layer structure and the dropout technique are investigated.The single-layer GRU is accurate enough for PWP prediction,whereas a double-layer GRU brings extra time cost with little accuracy improvement.The dropout technique is essential to overfitting prevention and improvement of accuracy.展开更多
The groundwater level has been continuously decreasing due to climate change and long-time overexploitation in the Xiong’an New Area,North China,which caused the enhanced mixing of groundwater in different aquifers a...The groundwater level has been continuously decreasing due to climate change and long-time overexploitation in the Xiong’an New Area,North China,which caused the enhanced mixing of groundwater in different aquifers and significant changes in regional groundwater chemistry characteristics.In this study,groundwater and sediment pore-water in drilling cores obtained from a 600 m borehole were investigated to evaluate hydrogeochemical processes in shallow and deep aquifers and paleo-environmental evolution in the past ca.3.10 Ma.Results showed that there was no obvious change overall in chemical composition along the direction of groundwater runoff,but different hydrochemical processes occurred in shallow and deep groundwater in the vertical direction.Shallow groundwater(<150 m)in the Xiong’an New Area was characterized by high salinity(TDS>1000 mg/L)and high concentrations of Mn and Fe,while deep groundwater had better water quality with lower salinity.The high TDS values mostly occurred in aquifers with depth<70 m and>500 m below land surface.Water isotopes showed that aquifer pore-water mostly originated from meteoric water under the influence of evaporation,and aquitard pore-water belonged to Paleo meteoric water.In addition,the evolution of the paleoclimate since 3.10 Ma BP was reconstructed,and four climate periods were determined by theδ18O profiles of pore-water and sporopollen records from sediments at different depths.It can be inferred that the Quaternary Pleistocene(0.78‒2.58 Ma BP)was dominated by the cold and dry climate of the glacial period,with three interglacial intervals of warm and humid climate.What’s more,this study demonstrates the possibilities of the applications of pore-water on the hydrogeochemical study and further supports the finding that pore-water could retain the feature of paleo-sedimentary water.展开更多
The distribution of nitrate, nitrite, ammonia, phosphate and silicate in pore-water and their exchange between sediments and overlying waters （benthic flux） were determined at nine locations on the shelve of Huangha...The distribution of nitrate, nitrite, ammonia, phosphate and silicate in pore-water and their exchange between sediments and overlying waters （benthic flux） were determined at nine locations on the shelve of Huanghai and East China Seas. On the basis of the redox potential of sediments and nutrients distributions in the pore-waters, it is found that the benthic sediments are being in a suboxic to anoxic environment in the Huanghai and East China Seas. The nutrients distribution in the pore-waters is mainly controlled by the sediment environment, and ammonia is the major inorganic nitrogen in the pore-waters. On the basis of benthic fluxes of nutrients calculated us- ing Fick＇s first law, there is remarkable efflux of ammonia, dissolved inorganic nitrogen（DIN）, phosphate and silicate from the sediments to the overlying waters in the study area, and their benthic fluxes are 299.3-2 214.8, 404.4-2 159.5 , 5.5-18.8 and 541.3-1 781.6 μmol/（m^2·d） respectively, and perhaps they are the major source of dissolved inorganic nitrogen, phosphate and silicate for the overlying water. At most stations, the nitrate flux was from the overlying waters to the sediments, which suggests that suboxic organic matter decomposition via denitrification is dominated in the most area of Huanghai and East China Seas. High benthic fluxes appearing in the coastal area and relatively low benthic fluxes occurring in the shelf area are found and are consistent with pri- mary productivity zoning in the study area. On the other hand, the ammonia flux displays an exponential decrease with water depth increase and an exponential increase with the bottom water temperature. However, others do not display this trend.展开更多
基金supported by the Natural Science Foundation of China(Grant Nos.51979158,51639008,51679135,and 51422905)the Program of Shanghai Academic Research Leader by Science and Technology Commission of Shanghai Municipality(Project No.19XD1421900)。
文摘Knowledge of pore-water pressure(PWP)variation is fundamental for slope stability.A precise prediction of PWP is difficult due to complex physical mechanisms and in situ natural variability.To explore the applicability and advantages of recurrent neural networks(RNNs)on PWP prediction,three variants of RNNs,i.e.,standard RNN,long short-term memory(LSTM)and gated recurrent unit(GRU)are adopted and compared with a traditional static artificial neural network(ANN),i.e.,multi-layer perceptron(MLP).Measurements of rainfall and PWP of representative piezometers from a fully instrumented natural slope in Hong Kong are used to establish the prediction models.The coefficient of determination(R^2)and root mean square error(RMSE)are used for model evaluations.The influence of input time series length on the model performance is investigated.The results reveal that MLP can provide acceptable performance but is not robust.The uncertainty bounds of RMSE of the MLP model range from 0.24 kPa to 1.12 k Pa for the selected two piezometers.The standard RNN can perform better but the robustness is slightly affected when there are significant time lags between PWP changes and rainfall.The GRU and LSTM models can provide more precise and robust predictions than the standard RNN.The effects of the hidden layer structure and the dropout technique are investigated.The single-layer GRU is accurate enough for PWP prediction,whereas a double-layer GRU brings extra time cost with little accuracy improvement.The dropout technique is essential to overfitting prevention and improvement of accuracy.
基金The study was financially supported by the National Natural Science Foundation of China(41807220)the Open Fund Project of Hebei Key Laboratory of Geological Resources and Environment Monitoring and Protection(JCYKT201903)the projects of the China Geological Survey(DD20160239 and DD20189142).
文摘The groundwater level has been continuously decreasing due to climate change and long-time overexploitation in the Xiong’an New Area,North China,which caused the enhanced mixing of groundwater in different aquifers and significant changes in regional groundwater chemistry characteristics.In this study,groundwater and sediment pore-water in drilling cores obtained from a 600 m borehole were investigated to evaluate hydrogeochemical processes in shallow and deep aquifers and paleo-environmental evolution in the past ca.3.10 Ma.Results showed that there was no obvious change overall in chemical composition along the direction of groundwater runoff,but different hydrochemical processes occurred in shallow and deep groundwater in the vertical direction.Shallow groundwater(<150 m)in the Xiong’an New Area was characterized by high salinity(TDS>1000 mg/L)and high concentrations of Mn and Fe,while deep groundwater had better water quality with lower salinity.The high TDS values mostly occurred in aquifers with depth<70 m and>500 m below land surface.Water isotopes showed that aquifer pore-water mostly originated from meteoric water under the influence of evaporation,and aquitard pore-water belonged to Paleo meteoric water.In addition,the evolution of the paleoclimate since 3.10 Ma BP was reconstructed,and four climate periods were determined by theδ18O profiles of pore-water and sporopollen records from sediments at different depths.It can be inferred that the Quaternary Pleistocene(0.78‒2.58 Ma BP)was dominated by the cold and dry climate of the glacial period,with three interglacial intervals of warm and humid climate.What’s more,this study demonstrates the possibilities of the applications of pore-water on the hydrogeochemical study and further supports the finding that pore-water could retain the feature of paleo-sedimentary water.
基金supported by the Major State Basic Research Program of China under contract No.G1999043704the National Natural Science Foundation of China under contract No.40206012.
文摘The distribution of nitrate, nitrite, ammonia, phosphate and silicate in pore-water and their exchange between sediments and overlying waters （benthic flux） were determined at nine locations on the shelve of Huanghai and East China Seas. On the basis of the redox potential of sediments and nutrients distributions in the pore-waters, it is found that the benthic sediments are being in a suboxic to anoxic environment in the Huanghai and East China Seas. The nutrients distribution in the pore-waters is mainly controlled by the sediment environment, and ammonia is the major inorganic nitrogen in the pore-waters. On the basis of benthic fluxes of nutrients calculated us- ing Fick＇s first law, there is remarkable efflux of ammonia, dissolved inorganic nitrogen（DIN）, phosphate and silicate from the sediments to the overlying waters in the study area, and their benthic fluxes are 299.3-2 214.8, 404.4-2 159.5 , 5.5-18.8 and 541.3-1 781.6 μmol/（m^2·d） respectively, and perhaps they are the major source of dissolved inorganic nitrogen, phosphate and silicate for the overlying water. At most stations, the nitrate flux was from the overlying waters to the sediments, which suggests that suboxic organic matter decomposition via denitrification is dominated in the most area of Huanghai and East China Seas. High benthic fluxes appearing in the coastal area and relatively low benthic fluxes occurring in the shelf area are found and are consistent with pri- mary productivity zoning in the study area. On the other hand, the ammonia flux displays an exponential decrease with water depth increase and an exponential increase with the bottom water temperature. However, others do not display this trend.
