Hydrological Response to Climate and Land Use Changes in the Dry–Warm Valley of the Upper Yangtze River

The hydrological process in the dry–warm valley of the mountainous area of southwest China has unique characteristics and has attracted scientific attention worldwide.Given that this is an area with fragile ecosystems and intensive water resource conflicts in the upper reaches of the Yangtze River,a systematic identification of its hydrological responses to climate and land use variations needs to be performed.In this study,MIKE SHE was employed and calibrated for the Anning River Basin in the dry–warm valley.Subsequently,a deep learning neural network model of the long short-term memory(LSTM)and a traditional multi-model ensemble mean(MMEM)method were used for an ensemble of 31 global climate models(GCMs)for climate projection.The cellular automata–Markov model was implemented to project the spatial pattern of land use considering climatic,social,and economic conditions.Four sets of climate projections and three sets of land use projections were generated and fed into the MIKE SHE to project hydrologic responses from 2021 to 2050.For the calibration and first validation periods of the daily simulation,the coefficients of determination(R)were 0.85 and 0.87 and the Nash–Sutcliffe efficiency values were 0.72 and 0.73,respectively.The advanced LSTM performed better than the traditional MMEM method for daily temperature and monthly precipitation.The average monthly temperature projection under representative concentration pathway 8.5(RCP8.5)was expected to be slightly higher than that under RCP4.5;this is contrary to the average monthly precipitation from June to October.The variations in streamflow and actual evapotranspiration(ET)were both more sensitive to climate change than to land use change.There was no significant relationship between the variations in streamflow and the ET in the study area.This work could provide general variation conditions and a range of hydrologic responses to complex and changing environments,thereby assisting with stochastic uncertainty and optimizing water resource management in critical regions.

Online machine learning for stream wastewater influent flow rate prediction under unprecedented emergencies

Accurate influent flow rate prediction is important for operators and managers at wastewater treatment plants(WWTPs),as it is closely related to wastewater characteristics such as biochemical oxygen demand(BOD),total suspend solids(TSS),and pH.Previous studies have been conducted to predict influent flow rate,and it was proved that data-driven models are effective tools.However,most of these studies have focused on batch learning,which is inadequate for wastewater prediction in the era of COVID-19 as the influent pattern changed significantly.Online learning,which has distinct advantages of dealing with stream data,large data set,and changing data pattern,has a potential to address this issue.In this study,the performance of conventional batch learning models Random Forest(RF),K-Nearest Neighbors(KNN),and Multi-Layer Perceptron(MLP),and their respective online learning models Adaptive Random Forest(aRF),Adaptive K-Nearest Neighbors(aKNN),and Adaptive Multi-Layer Perceptron(aMLP),were compared for predicting influent flow rate at two Canadian WWTPs.Online learning models achieved the highest R2,the lowest MAPE,and the lowest RMSE compared to conventional batch learning models in all scenarios.The R2 values on testing data set for 24-h ahead prediction of the aRF,aKNN,and aMLP at Plant A were 0.90,0.73,and 0.87,respectively;these values at Plant B were 0.75,0.78,and 0.56,respectively.The proposed online learning models are effective in making reliable predictions under changing data patterns,and they are efficient in dealing with continuous and large influent data streams.They can be used to provide robust decision support for wastewater treatment and management in the changing era of COVID-19 and also under other unprecedented emergencies that could change influent patterns.