Study On Water Cycle Evolution And Nitrogen/Phosphorus Loss In Reservoir Watershed

The water cycle evolution in large-scale reservoir basins and the driving mechanism of the associated process are the theoretical basis for water safety guarantee in the reservoir area.The Three Gorges Reservoir area(TGRA)is located in the hinterland of southwestern China and in the middle and lower reaches of the Yangtze River,which has played a huge role in flood control,power generation,shipping,water supply and other benefits.During its 16 years operation,the mechanism of the water cycle,climate effects and water-soil-eco-environment systems have changed much more complexly,which is of great significance for Green and healthy development.Based on the analysis of water cycle and rainstorm changes in TGRA before and after the impoundment and combining with field observation and artificial rainfall experiments under rainstorm conditions,we revealed the relationship between different runoff components and the loss of nitrogen and phosphorus nutrients.The results can explain how the water cycle drives the loss of nitrogen and phosphorus nutrients.At the same time,a distributed water cycle model for small watersheds in the reservoir area is established that builds a set of parameterized schemes.Finally,the climate model data acts as input is used to identify the future water cycle and water environment evolution trends in small watersheds.The main conclusions are as follows:(1)Before and after impoundment in TGRA,there were no significant changes in annual precipitation.During the storage period(September-October),it increased significantly after impoundment than before.The annual average temperature in drainage period ascended much more obviously,however,it decreased during the impounding period.The annual evaporation represented the same trend with the annual average temperature.The annual surface runoff showed decreasing trends,and it was consistent with annual precipitation during the impoundment.The annual surface soil moisture(SSM)in TGRA showed decreasing trends.But the SSM from January to February and August to December increased.There were no obvious variation characteristics between precipitation and runoff in Wanxian station.The point of variation between evaporation and flow appeared in 1982,and the precipitation-evaporation one appeared in the 1990-1992 interval.In Wuxi station,there was no obvious variation of these relations.(2)The hourly precipitation(P)in TGRA from 2008 to 2018 showed that the frequency of the hourly P that exceed 20 mm appeared increasing trends.After 2015,the scope that occurred rainstorm(P>20mm/h)grew up.The highest hourly P was 76.83mm,which happened near the Dianjiang County.The results of water vapor and flux showed that the water vapor mainly focused on 850hPa height and came from southwestern directions.The TGRA was divergent area during these periods.(3)Field observations showed that soil discharge(Qs)contributed the most to the total discharge.The three different slope fields produced different Qs that the 10° plot was the most and the 15° one was smallest.The surface discharge(Qr)in these three plots represented significant correlations with the concentration of PO4+-P.When the slope is 10°,there was an obviously positive correlation between Qr and TP,TN.When it is 5°,the Qr was significant correlated with the concentration of NH4+-N,NO2--N,and TP.The correlations between Qs and these indices were not so significant.On the other hand,the small watershed observation indicated that the formula of the precipitation and runoff was y=53.847x+93.344.The correlations of outlet runoff and nutrient concentration were not obvious.(4)Artificial rainfall experiments under rainstorm conditions showed that the rainfall intensity(RI),slope,discharge and sediment were key factors which impact on nutrient loss.When RI exceeded 80mm/h,the average Qr increased with slope.When RI<40 and slope<5°,there was no Qr,Qs ascended with RI.There was a quadratic term relationship between Qr and TN concentration(P<0.05).The relationship between Qr and TP concentration was linear,but not significant.In 20° plots,Qs and TN concentration are quadratic terms,but others were linear correlations.Other nutrient indices fluctuated with Qr and Qs.(5)Based on SWAT hydrological model,we simulated the runoff and non-point pollution load in different precipitation frequencies.In 2014(wet year),the TN and TP load were 4570.32t and 814.44t,respectively.In 2010(normal year),the TN load was 2329.83t and the TP load was 442.21t,In 2012(dry year),the TN and TP load were 690.11t and 50.30,respectively.At the end,we used the climate model data in different scenarios to drive the SWAT model that we built.The predicted runoff and non-point load can help the local government to make decisions.