| Due to the dual effects of climate change and overgrazing,the eco-hydrological degradation of the grassland inland river basin in China is becoming more and more serious.It is particularly urgent to accurately describe the mechanism and process of coupling,mutual feeding and cooperative evolution of watershed eco-hydrological processes.In this paper,the Xilin River Basin(XRB)of semi-arid steppe inland river basin was selected as the research area,and the research methods including field observation experiment,eco-hydrological mechanism research,distributed model development and construction,historical process simulation analysis and future model scenario prediction are adopted.Multi-factor,multi-scale,and long series comprehensive experimental observation data of meteorological,vegetation,soil and runoff were obtained,and a variety of new vegetation indices and soil infiltration scale transformation models were proposed to reduce the measurement error of vegetation-soil assimilation data.A distributed eco-hydrological model of dynamic process was constructed,which integrated multi-source and multi-information,to simulate the eco-hydrological process at 3-hour scale in the region with incomplete historical measurements,and to distinguish the reciprocal feed-back response process of ecological and hydrological elements.The ecological and hydrological trends of the basin under five common future scenarios in CMIP6 were predicted,and suggestions for external management and planning were put forward.The main achievements of this paper are as follows:(1)Combined with remote sensing data of optical infrared satellite Landsat-8 band operation and synthetic aperture radar Radarsat-2 polarization decomposition,the study compares 5 traditional vegetation indices with 11 new mixed vegetation indices and a deep learning neural network optimized by particle swarm optimization.Downscaling accurately inverts the vegetation coverage of the basin with a resolution of 8m.By establishing the scale transfer function of double ring infiltration,the infiltration process of different diameter of infiltrometer was deduced and the scale effect threshold caused by the change of inner diameter was calculated,which completed the removal of scale effect of steady infiltration rate of watershed soil,and achieved the purpose of meeting the measurement accuracy and saving consumables.(2)A distributed dynamic process model(DDPM)was proposed and constructed for the steppe inland river basin in arid and semi-arid regions.Eight nested modules of the model were used to simulate the water-heat-carbon cycle in XRB from 1980 to 2020.The snow cover in 3-hour days was identified by dynamic evpotranspiration,and the changes of seven evapotranspiration components at annual,seasonal,daily,and 3-hour scales were reconstructed,compared with measured data,R~2>0.85,RMSE<0.1mm.For semi-arid grasslands river winding and changeful,water steep rise steeply characteristics,considering dynamic real long river,the river curves and the convergence and overflow process,careful depict the river floodplains and flood peak due to flood the transit and arrival time(the accuracy of runoff simulation using two kinds of meteorological driving data is R~2=0.947,NSE=0.945;R~2=0.932,NSE=0.905),briefly described the basin diving underground water level and flow field distribution features.Based on the dynamic sunrise and sunset time dividing day and night,the carbon cycle of soil,vegetation,and ecosystem,as well as the variation process of soil heat flux,sensible heat flux,latent heat flux and three kinds of water use efficiency were quantitatively described at various time scales,providing data and theoretical support for exploring the interaction and mutual feed-back relationship of eco-hydrology in grassland watershed.(3)By analyzing the change process of more than ten eco-hydrological factors in the XRB every five years,the influential factors of carbon emission,the important process of grassland river runoff,the frequency of overflow and the influential factors,the study found that temperature could well explain the change of carbon emission in various time scales.The continuous dry soil moisture in the study area could not be used to explain soil respiration alone.Vegetation degradation and precipitation increase caused by overgrazing were the main reasons for the increase in the number of overflows,and there was a two-year lag between the number of overflows and global climate change.The analysis of correlation,periodicity and hysteresis among the eco-hydrological factors shows that the ecological status of vegetation in the XRB is closely related to the hydrometeorological conditions in the current year and the following years.The short-term rainfall supply easily leads to the growth of a large number of annual herbaceous plants,but this is not a sustainable recovery.Continuous wetting promotes the growth of perennials with a variety of plants,which truly contributes to the ecological restoration to the best state.(4)Based on five future scenarios of CMIP6,the trend of temperature and precipitation in the future was analyzed,and the future trend of surface soil moisture,runoff,sensible heat,latent heat flux and gross primary productivity was predicted.It was found that the medium-high concentration emission path(RCP7.0 and RCP8.5)showed more unstable factors,including irregular fluctuations and extreme weather,and thus affected the occurrence of higher frequency of drought and flood,longer duration of continuous fluctuation,stronger intensity of vegetation ecological degradation,more precipitation and higher potential evapotranspiration.The use of lower concentration of discharge and more moderate development will be directly beneficial to the eco-hydrological status and functions. |
PDF Full Text Request