Water And Sediment Characteristics And Their Responses To Landscape Pattern Changes In High Risk Regions Of Soil Erosion In The Huaihe River Basin

The Huaihe River Basin（HRB）is an important crop product base of China,which supported about 25%of the total marketable grain by about 12%of the total agricultural land of China,plays an essential role in guarantee the Chinese crop security.The agricultural land is one of the main sources of the soil erosion,the large amount of agricultural land within the regions lead to the serious soil erosion,further decrease the soil fertility and crop yields.Agricultural economy is one of the main economic industry within the HRB,it is not suitable to promote the effective measurement for soil and water conservation such as“Grain for Green Program”.In addition,landscape pattern is one of the characteristic of the distribution of the land use,has a significant impact on the runoff and sediment process.Understanding the effects of landscape pattern on runoff and sediment process in the watershed could bring the possible to realize the purpose of soil and water conservation by adjusting the characteristica of landscape pattern without greatly change the number of various land use type.The HRB has been chosen for the study area,the spatio-temporal variation of the rainfall erosivity within the HRB during 1960～2018 has been analyzed to confirm the high risk regions for the soil erosion,and then,the long-term regression model has been established for the rainfall erosivity within the HRB and the large scale atmospheric circulation.In addition,a distributed hydrological model was established for the high risk regions of the HRB for the soil erosion,the model parameters were calibrated in a distributed manner;the spatio-temporal variation of the runoff and sediment processes within the high risk regions for the soil erosion were analyzed in multiple scale based on the model output of the calibrated hydrological model.Finally,the variation of the historical landscape pattern within the high risk regions for the soil erosion were analyzed,and the effects of typical landscape pattern charactristics on runoff and sediment processes and the annual differences were analyzed qualitatively and quantitatively.The main conclusions were as below:（1）During 1960～2018,the annual average rainfall erosivity of the HRB,Upper HRB（UHRB）,Middle HRB,Lower HRB,and Yishusi River Basin was 4280,5060,4068,4886,and4089 MJ?mm?hm^-2?h^-1,respectively.The Upper HRB,where rainfall erosivity would be continued increased in the future,is the high risk regions of the HRB for the soil erosion.On the continuous annual and monthly scale,the Arctic Oscillation and North Pacific Index were both significantly correlated with rainfall erosivity with various lag time;the monthly Arctic Oscillation,Southern Oscillation Index,Ni?o3.4 Sea surface temperature,and North Pacific Index in different months were all exist significant correlation with rainfall erosivity.The large scale atmospheric circulation indice of each month have the potential to be the predict indicators for the rainfall erosivity within the watershed.（2）The runoff and sediment processes in the UHRB during 1962～2019 were divided into four simulation periods by the double accumulation curve test method:P1（1962～1980）,P2（1981～1991）,P3（1992～2009）,and P4（2010～2019）.The SWAT（Soil and Water Assessment Tool,SWAT）models were established and the model parameters were calibrated based on the measured runoff and sediment records in 3 hydrological stations within the UHRB.The calibrated SWAT model could simulate the runoff and sediment processes within the UHRB well.The performance of the runoff simulation of different stations in most periods reached the excellent level with NSE（Nash-Sutcliffe efficient coefficient,NSE）value higher than 0.75,the performance of the sediment yields simulation of different stations in most periods reached the acceptable levels with NSE value higher than 0.50.Among the four simulation periods,the performance of runoff and sediment yields were best during P3,with a NSE value ranges0.84～0.93,and 0.67～0.91,respectively.The SWAT model can be used to simulate the water-sediment process in the upper reaches of Huaihe River.（3）The average annual runoff（R）and the average annual RC in the UHRB showed significant spatial differences,and the relationship between runoff and sediment yields varied obviously in different decades.The sediment yields and erosion modulus were decreased significantly（P<0.05）within the UHRB,and the soil erosion was mainly concentrated among June～September.The average annual R of the UHRB has a significant spatial difference.On the whole,the average annual R decreases from the southwest to the northeast,and the erosion modulus decreases from the southwest to the northeast.In addition,we need to focus on the prevention and control of soil erosion in Xinxian and Tongbai counties in the upper reaches of the Huai River,as well as the prevention and control of river sedimentation in Guangshan and Huangchuan counties.（4）During 1980～2018,the paddy land,forest land,water area,and urban land shown an increased trend,while the dry land,brush land,and grass shown a decreased trend,and the main land use conversions were the exchanges between dry and paddy land,and large amount dry and paddy land converted into forest and urban land.The boundaries between paddy fields and other landscape types showed an increasing trend,and the spatial distribution was also more concentrated than in the past.Human activities had a significant impact on the spatial distribution of paddy fields.The boundary changes between drylands and other landscape types fluctuate greatly,with a decrease in scattered drylands and a more concentrated landscape distribution.The dominant landscape within the watershed is more dispersed than in the past,with the impact of human activities gradually increasing,the connectivity between different patches decreasing,and the fragmentation and diversity of the landscape increasing.（5）Most typical landscape pattern metics at the paddy land,dry land,and landscape level have significant correlations with annual and monthly R,SEM,and SSC（P<0.05）.When the proportion of paddy land increases or the distribution becomes more concentrated,the annual R,annual SEM,and annual SSC will decrease.When the proportion of dry land increases or the complexity of patches decreases,the annual R,annual SEM,and annual SSC of the watershed will increase.R increased with the landscape edge becomes more complex,which increases the hydraulic erosion of the watershed.The diversity of landscape within the watershed helps to reduce R and SY,thereby significantly reducing the SSC in the channel.The PLSR model established using various landscape pattern metrics as independent variables can better simulate changes in R,SEM,and SSC.Some paddy land landscape pattern characteristics（complexity of patch shape,density of distribution,and type of surrounding landscape）,dry land landscape pattern characteristics（area ratio,concentration of dryland distribution,and edge length）,and landscape level landscape pattern characteristics（overall fragmentation,single patch area size,and diversity）have significant annual differences in their impact on runoff and sediment processes within a watershed.The research results contribute to achieving sustainable land resource utilization and water resource management in the watershed.