Comparison And Application Of Spatiotemporal Sediment Delivery Ratio Algorithm In Yanhe Watershed

Investigating the evolution of the erosion and sediment yield environment in a typical loess watershed under ecological restoration conditions and analysing the relevant patterns of spatio-temporal sediment delivery ratio in the watershed could provide a scientific basis for promoting scientific management and further high-quality development of the watershed.However,few studies have assessed the applicability of different approaches to modelling and evaluating the spatio-temporal sediment delivery ratio in watersheds.In this study,three methods including the SWAT,RUSLE models coupled with the sediment connectivity index IC and RUSLE models coupled with the SEDD models,were used to calculate the spatio-temporal sediment delivery ratio in the Yanhe Watershed,and deeply analyzing the relevant laws of the watershed erosion and sediment yield process.Finally,the optimal model algorithm for the sediment delivery ratio in the Loess Hill watersheds is derived through comparative analysis and small watershed validation.The research shows that as follows:（1）The SWAT model and SDR_SWAT of the Yanhe Watershed were constructed to assess the spatial and temporal variation of runoff erosion power and sediment delivery ratio in the watershed using nested basin theory over different periods.The spatial and temporal variability of runoff erosion power is obvious,with the maximum value decreasing from 13.28×10^-4 m⁴·s^-1·km^-2 in 1985 to 5.02×10^-4 m⁴·s^-1·km^-2 in 2015,the runoff erosion power has obvious spatial scale effect in space,showing the characteristics of large in tributaries and small in main streams in space.When the nested basin area is less than 1000 km²,the runoff erosion power increased in a power function trend as the area decreases.Therefore,it is necessary to focus on small-scale runoff erosion events.The SDR_SWAT calculated based on the SWAT model is highly variable at the annual scales,maintaining high values in the 1980s and 1990s,especially in the 1990s,the SDR_SWATexceeded or was over 0.7 in several years,while after 2000,the SDR_SWAT remained at 0.2and 0.6.On the monthly time scale,the sediment delivery ratio was closely related to rainfall and runoff,showing large values in the larger flood season from July to September,indicating that large rainfall and runoff events within the year could often transport more sediment.（2）RUSLE results show that the average erosion of the basin in the 1980s and 1990s exceeded 80 t·hm^-2·a^-1 and even 100 t·hm^-2·a^-1 in some years,while it dropped significantly to about 20 t·hm^-2·a^-1 after 2000,especially in around 2015.This is mainly due to the substantial erosion attenuation caused by the continued revegetation of the watershed,and the interannual variation between the two is negatively correlated significantly,but the upstream areas,including subbasins 1,2,3,4,6 and 7,still exhibit high erosion intensities that require focused attention.（3）The sediment connectivity index IC and SDR_IC were constructed to develop an integrated modelling approach for identifying erosion and sediment yield in the watershed,in order to visually assess the spatial and temporal characteristics of erosion and sediment yield processes of the Yanhe Watershed from 1985 to 2020,and to clarify the relationships between IC and subbasin erosion.The results show that the spatio-temporal patterns of vegetation cover had a strong influence on the distribution of IC,and the local minima in IC（less than-13）,which occurred in 1995,1998,and 2010,were caused by the increase in vegetation in some local areas.The proposed IC-Erosion diagram achieved visual expression of watershed erosion and sediment connectivity,show that from 1985 to 1999,areas with high connectivity but low erosion accounted for more than 60%of the watershed area,indicating a dynamic mismatch between sediment connectivity and erosion.From2000 to 2004,over 40%of erosion occurred in a relatively small number of areas with high connectivity and high erosion match areas（about 20%）,which can be defined as critical areas for erosion,which will help in developing land use planning and identifying critical areas for soil and water conservation.The SDR_IC calculated based on the IC model showed the spatio-temporal variation of sediment delivery ratio in the study area over the years,with a more significant downward trend in SDR_IC on the time scale,especially after 2000,when the sediment delivery ratio remained around 0.5 and remained on a downward trend since 2015,with the SDR_IC value converging to 0.4;while the spatial distribution characteristics of SDR_IC were consistent with IC,showing the characteristics of"large in gullies and valleys,small in slopes".（4）The SEDD model and SDR_SEDD was constructed to reflect the spatial sediment delivery ratio characteristics of the watershed as a whole.The upstream area has a strong sediment delivery capacity with the SDR_SEDD exceeding 0.6,and less than 0.5 in the middle reaches,where sediment delivery was relatively small.However,the temporal characteristics of SDR_SEDD’s expression are not obvious.The main temporal variation in SDR_SEDD was found in some subbasins with larger sediment delivery ratio in the upper reaches,where SDR_SEDD decreased to some extent,but the overall sediment delivery ratio in the watershed does not change significantly over a long series of time scales,with the annual mean SDR_SEDD concentrated in the range 0.54-0.58 and less variable,and the sediment delivery ratio in the subbasins mostly concentrated in the range 0.35-0.70.（5）Through model comparison and analysis,the effective IC-based sediment delivery ratio algorithm(SDR_IC)with good applicability in the Loess Hill watersheds was selected for promotion and application.The algorithm was well applied to the erosion and sediment yield simulation of the Qingjian River Watershed,the validation basin,and the results showed the spatio-temporal patterns of erosion and sediment yield in the watershed from 2006 to 2012.The erosion in the watershed has been controlled to some extent,but soil and water conservation work still needs to be strengthened in the midstream area;the sediment delivery ratio in some of the downstream areas was above 0.7,and these areas could be identified as the key area for sediment management in the watershed.The model algorithm was also validated with the measured sediment transport data,and the results showed that the algorithm has a good simulation effect.The model approach is therefore feasible,applicable and generalisable,and provides a comprehensive way of identifying the erosion and sediment yield processes and spatio-temporal sediment delivery ratio in the watershed,which can provide key information for sediment management in the watershed and provide a scientific basis for the government to carry out targeted soil and water conservation measures and ecological management work in different areas of the watershed.