Distribution Of Water Temperature In The Main Stream Of Xiangxi River In The Three Gorges Reservoir And Its Effect On The Succession Of The Circulation Pattern In The Reservoir

This paper analyzes the time-varying characteristics of the main stream of the Xiangxi River estuary before and after the impoundment of the Three Gorges Reservoir,and then uses the α method of the number of reservoir water replacement times to determine the water temperature structure of the Xiangxi River estuary,according to the temperature gradient,the length of the stratification time,etc.The indicator analyzes the vertical distribution characteristics of water temperature.Combined with the air temperature,wind speed,relative humidity,rainfall,inflow water temperature,inflow rate,and outflow rate of the main stream of the Xiangxi River in the Three Gorges Reservoir from 2003 to 2013,the main influencing factors were found through cluster analysis and grey correlation analysis.In the study of the influence of the main influencing factors and the influence of the water temperature distribution on the main stream of the Xiangxi River.The LSTM time series model was established to simulate the surface water temperature of the main stream of the Xiangxi River estuary.Firstly,the main influencing factors and the measured surface water temperature are selected as the input variables of the model,the LSTM model type is selected,the parameters are set,and then the data from 2003 to 2008 are selected for model rate determination.The data from 2009 to 2013 are validated by the model and verified.Obtain the model and parameter size when the error is the smallest,so as to predict the daily surface water temperature from 2014 to 2016,and then use the Logistic curve prediction model based on the surface water temperature to determine the parameters k,b,a,and simulate the sag of the Xiangxi River estuary.To the water temperature,the vertical distribution law of the water temperature of the main stream of the Xiangxi River in the Three Gorges Reservoir is obtained.Through the encrypted water temperature of the Xiangxi River estuary,as the lower boundary condition of the Xiangxi River hydrodynamic model,the relevant data of the Xiangxi River Valley model were collected to establish the Xiangxihe Reservoir Bay based on CE-QUAL-W2.Hydrodynamic model.Sensitivity analysis and calibration of the relevant parameters of the model were carried out.Finally,the model was verified based on the measured data,and the influence of the variation of the water temperature in the Xiangxi River estuary on the succession of the Xiangxi River to the reservoir pattern was analyzed.The main results are as follows:(1)the change of water temperature in miaohe river after impoundment is obviously lagging behind that in cuntan(natural river channel).The temperature difference between cuntan and miaohe is affected by the flow size.As the flow increases,the temperature difference appears positive and negative alternations without obvious rules,namely the phenomenon of variable temperature period.The larger the flow,the longer and more obvious the duration of variable temperature period.The three gorges reservoir water temperature distribution types should be unstable stratification,emerged from the layered from 2006 to 2016,xiangxi brook river trunk stream of amplitude transformer is greater than the underlying surface temperature difference temperature difference,xiangxi brook river trunk stream of layered mainly appear in 4 ～ 5 month,part of the year 3,6 month also appear statified,emerged from layering just appeared,to the maximum,the depth of the maximum temperature gradient on the increase.The time from the temperature difference to the maximum is longer than the time from the maximum to the disappearance of the temperature difference,and the time from the stratification to the maximum stratification tends to decrease.When the flow reaches around 15000m3/s in each year,the stratification is disappearing.(2)flow,relative humidity,wind speed and temperature are the main controlling factors.The slope of temperature in the heating period is greater than that in the cooling period,so the heating rate in the heating period is greater than that in the cooling period.In thermostatic period(1 ～ 2)at the end of the end of,as temperatures rise,the water temperature in 9.00 ℃ to 13.00 ℃,the water temperature is basically the same.In the flow in the process of increasing,the water temperature changes from the beginning is not obvious(between 0.50 ℃ ～ 0.50 ℃)increases gradually,when the traffic reaches 30000 m after a/s,the water temperature changes between 0.50 ℃ and 0.50 ℃ in volatility changes.The change of average wind speed from 2003 to 2013 was lower from 2003 to 2008,with a range of 0.622m/s.From 2008 to 2013,the change range was 1.066m/s,and the wind speed was negatively correlated with the water temperature.The change of daily relative humidity from 2003 to 2013 was relatively low from January to may,and the change of relative humidity and surface water temperature was positively correlated in other periods.(3)single-layer or double-layer model is better than multi-layer model.LSTM time series model can well simulate forecasts of the three gorges reservoir surface water temperature,water temperature by the prediction results of 2014 surface temperature change range of 11.80 ～ 26.00 ℃,the lowest temperature is appear on March 10,2014,the highest temperature is appeared in the August 7,2014.2015 surface temperature change range of 11.98 ～ 27.10 ℃,the lowest temperature is appear on March 3,2015,the highest temperature is in August 18,2015.Logistic curve can well reflect the three gorges reservoir water temperature prediction formula of vertical water temperature distribution of the fitting accuracy is higher,the water temperature of maximum absolute error range of 0 ～ 0.79 ℃.Predict by determining k,a,b,the predicted results,the water temperature maximum absolute error range of 0.45 ～ 1.63 ℃.(4)is obtained by model output on June 29,2014,the water temperature distribution in the cross section of the xiang river estuary(XX01)compared with the measured values,the absolute error range in 0.02 ～ 0.95 ℃,the average of the absolute error is 0.55 ℃.On June 29,2014,the absolute error range in 0.00 ～ 1.13 ℃,the average of the absolute error is 0.41 ℃.During the fluctuation period,there were 12 inflection points of water temperature fluctuation,and the occurrence of each inflection point of water temperature was the trend change of circulation pattern.It can be seen that the fluctuation of water temperature data was larger than that of the previous three times a month.It can be seen from the dynamic succession process of circulation modes in different stages that the depth of the mixed layer at the position of XX01 has a significant amplitude around the inflection point of water temperature,which more accurately reflects the succession process of circulation modes and has a more severe impact on xiangxi river reservoir bay than the previous research results.