Research On Response Simulation Of Environmental Hydrological Process And Risk Assessment Of Water Pollution To Changing Environment In Huai River Basin

Water cycle as a link among the earth system "Geosphere-Biosphere-Atmosphere",is the core issue of global change research.Affected by the global warming,global hydrological cycle exacerbated and extreme hydrological events e.g.floods,droughts,etc.occurred frequently.On the other hand,intensive human activities exacerbated the environment changes e.g.land use changes and large-scale water conservancy constructed in the rivers.With the trend of global warming,measured runoff and water resources in northern rivers of China decrease obviously,and the frequency and severity of extreme drought and flood are increasing.In addition,large-scale water conservancy constructed alters the runoff process and exacerbates water pollution in the rivers.The inherent nonlinearity and time-variation in hydrological processes become increasingly significant with intensive anthropogenic activities and climate change forcing.Therefore,the hydrology statistical models,hydrological models,hydrodynamic models,water quality models and non-point source models,etc.should be made full use of,so as to accurately describe the variations in hydrological factors and water quality elements at basin scale,and identify and quantify the flow regime changes and underlying water quality problems due to anthropogenic activities and climate change,and further assess the combined risk of nonstationary water quality.Supported by the National Grand Science and Technology Special Project of Water Pollution Control and Improvement(2014ZX07204-006)and the National Basic Research Program of China(2010CB428406),the Huai River Basin(HRB)which polluted seriously and regulated highly by human activities,was selected for the case study.Based on basin hydrology theory,this study was conducted by using the knowledge of hydrological statistics,systematic hydrology,water pollution migration and transformation mechanism and hydrodynamics,etc.Thus,the spatial and temporal variations of hydrological and water quality elements were detected,and the anthropogenic activities and topographic factors influencing water quality variations were identified;The influence mechanisms of land use on water quality variations,along with scale-process interactions on the variations at multiple scales was revealed.Mathematical models describing hydrological and water quality processes were established to quantitatively evaluate the impacts of climate change and human activities on the hydrological and water environment process,and to explore the mechanism of future climate change impacts on annual runoff.Finally,time-varying marginal distribution and copula models were conducted to quantitatively evaluate water quality combined risk.The main work can be summarized as follows:(1)The significances of the research on the numerical simulation of watershed hydrological and water environment processes under a changing environment,and the assessment of water environment combined risk,were described.The research progress of climate change impacts on water cycle,the coupling simulation of watershed water quality and quality,hydrological response to the changing environment and combined risk of water environment were reviewed in this field.(2)A variety of diagnostic techniques for trends and step changes were used to detect the spatio-temporal variations in the hydrological and meteorological series.The results demonstrated that there were no significant descent trends in precipitation and runoff and significant step changes in runoff occurred in 1992.Moreover,the impacts of climate change and human activities on runoff were quantitatively evaluated combined with the hydrological models DTVGM,SWAT and HBV and elasticity coefficients method.A total of 36 models including 18 runoff models and 18 runoff coefficient models are examined to select the better models matching the observed runoff for the application of future runoff prediction.Schreiber-Holland model presents a better argument for the measured runoff.Under future climate scenarios,the temperature and precipitation increase but runoff decrease compared the mean annual runoff during the period 1960-2010.Runoff changes exhibit a positive correlation with precipitation but a negative correlation with temperature,and is more sensitive to changes in precipitation than temperature.Precipitation increase required for runoff maintenance at various increases in temperature is calculated and break even range is given in contour plot.Climate elasticity has a high variation magnitude from the min to max along the directed line near the central symmetry point,where the numerical instability usually occurs.The increase in temperature of three future RCPs illustrates that the precipitation increase fails to offset the impacts of temperature increase and results in runoff decrease.(3)In this study we conducted the spatio-temporal statistical detection of water quality variation using multi-statistical analyses in the HRB,which is the sixth largest river in China and have been a serious polluted problem in aquatic environment.Future,anthropogenic actives and natural factors influencing water quality variations were identified,and the influence mechanisms of land use on water quality variations,along with scale-process interactions on the variations at multiple scales were revealed.The analysis demonstrated significant improvement of water quality during 2005-2014 in the HRB.There was a weak and moderate positive spatial autocorrelation for water quality parameters NH3-N and CODmn while DO exhibited a weak positive spatial autocorrelation for the whole period and in 2010s.Two cluster centers of significant high concentrations were detected for DO and TP at Mengcheng and Huaidian respectively,while four cluster centers of significant low concentrations for DO at Wangjiaba and Huaidian in the 2010s.Water temperature,regulated flow and load of water quality variables exhibited a significant correlation to water quality variation.Water quality variations could be determined differently from each station for each water quality parameters.(4)In the upper and middle regions of HRB,an integrated numerical model was established to describe the hydrological and water quality processes.The hourly flow and water quality data from the period 2007 and 2006 and 2008 are for calibrating and validating the model respectively.The evaluation indicators demonstrated good agreement between the observed and simulated hydrographs for the calibration and validation periods,and could well reflect the water quality variations in regulated river basins.In addition,Huaidian and Fuyang dams could promote water quality improvment.On the contrary,Bengbu dam exacerbated the water quality deterioration in the downstream.(5)SWAT model were calibrated and validated for modelling future streamflow and nutrient parameters(nitrogen and TP)load with the historical measured data over the period 2000-2010.The daily flow and water quality data from the period 2000-2006 and 2007-2010 are for calibrating and validating the established model respectively.The performance of non-point source pollution model was evaluated using the determination coefficient(R2),Nash-Sutcliffe efficiency(NSE)and percent bias(PBIAS).The results indicated that R2 and NSE values of the calibration period for streamflow in three watersheds range from 0.75 to 0.92 when the range in validation period is 0.76 to 0.88.In both calibration and validation period,the PBIAS values fall within the range between-10%and 13%.As for N and TP load in both calibration and validation period,R2 and NSE values range from 0.50 to 0.75 while the PBIAS values have a higher change range between-42.8%to 24.2%.Streamflow was more sensitive to climate change than land cover change,and the effect would be enhanced when climate and land cover changes occur concurrently.Monthly streamflow was significantly influenced by climate change that decreased from August to November in three study watersheds(Bengbu,Lutaizi and Wangjiaba)while increased from December to February and April to June.Nitrogen and TP load from August to November of three watersheds almost exhibit a significant decrease trend except for TP load in Wangjiaba while increase from April to June.At the watershed scale,nitrogen load during the 2010s is higher in Bengbu under RCP4.5.The changes in TP load form November to March under RCP 8.5 are more significant than RCP2.6 and RCP4.5.(6)Making fully use of the long time series of monitoring data,Markov method was applied to reveal the joint transition laws of water quality combined risk.The first-order Markov joint transition probabilities of water quality state combing with NH3-N and CODMn were different depend on initial water quality state.The larger joint transition probability had the trend toward state Class Vw,Class IV and Class ?,and then tended to Class II,Class V and Class I.The nonstationary had been revealed both in marginal models and copula models between the NH3-N and CODMn series.Time and%land cover of farmland and urban blocks were employed as the explanatory variable for estimating the nonstationary distribution parameters,and showed that the explanatory variable time had a better performance for explaining the changes of distribution parameters.Frank copula had a better performance for describing the dependence structure between NH3-N and CODMn in this study.In modelling the dependence structure changes,time-varying copula had an obvious advantage compared with constant parameter copula.Besides,the downward trend presented in dependence parameter illustrated that the dependence between NH3-N and CODMn will be weaken as the time varies.The largest synchronous encounter risk probability of NH3-N and CODMn occured while they simultaneously reached Class V.The asynchronous encounter risk probability was largest when NH3-N and CODMn was inferior to class V and class IV water quality standard respectively.