Modeling Hydrological Responses To Ecological Restoration In The Upper Reaches Of Beiluo River Basin,Loess Plateau,China

Loess Plateau belongs to a semi-arid climate region with the characteristics of unevenly distributed water resources and transitivity between the soil,climate and vegetation,which also lead to the results of destroyed vegetation,broken landform and loess soil in the Loess Plateau.Hence,the area gradually became one of the most serious soil erosion regions in the world.In order to prevent the soil erosion,a series of soil and water conservation measures were implemented,which lead to that the land use/ cover obviously changed and impacted the ecosystem,the climate change,hydrology and water resources in the Loess Plateau.These changes attracted many scholars attention in various research field.Therefore,the upper reaches of Beiluo River Basin was selected as research area in this study.The methods including sliding average,Sen Slope and Mann-Kendall test were selected to analyze the changing trends of meteorological and hydrological data.The Landsat TM imagine data were employed and the vegetation coverage were derived to analyze how the vegetation restored and the trends of vegetation distribution on the landform.Based on land use maps in the year of 1995,2000 and 2010 combined with RS and GIS,digital watershed model was established and quantitatively analyzing the contribution from land use/cover change and climate change to the variation of runoff.And the changed trend of runoff,evapotranspiration and soil water content were analyzed.The temporal and spatial variation of soil erosion were also analyzed based on that database.In the end,according to the different land use and climate scenarios,the hydrological responses were analyzed.The main conclusions were as follows:(1)Precipitation of the upper reaches of Beiluo River Basin from 1963-2012 nonsignificantly increased and the annual increment was 0.45 mm.The annual average precipitation was 439.33 mm.The changed period of precipitation were roughly divided into three period during the 50 years.From 1963 to 1978,the year was non-significant decreasing,and the linear trend was-2.3 mm·a-1.The average annual precipitation was 416.9mm.From 1979 to 2001,the year was non-significant increasing period,and the linear trend was 0.2 mm·a-1.From 2002 to 2012,the year was decreasing period,and the linear trend was-0.4 mm·a-1.(2)The Mann-Kendall method was selected to analyze the change trend of runoff and sediment.The results showed that the runoff and sediment obviously decreased(P<0.05),and the average annual variability were-0.32mm·a-1 and-170.18 t·km-2·a-1,respectively for runoff and sediment.The change trend of annual runoff from 1963 to 2012 can be divided into three stages.From 1963 to 1978,the annual runoff slowly decreased,and the linear trend was-0.9mm·a-1.From 1979 to 2001,the annual runoff gradually increased,and the linear trend was 0.3 mm·a-1.From 2002 to 2012,the annual runoff obviously increased,and the linear trend was-0.7 mm·a-1 at a significant levele of P<0.05.The change trend of annual sediment load from 1963 to 2012 also was divided into three stages.From 1963 to 1978,the annual sediment load slowly decreased,and the linear trend was 10375.1 t·km-2·a-1.From 1979 to 2001,the annual sediment load gradually increased,and the linear trend was 10375.1 t·km-2·a-1.From 2002 to 2012,the annual sediment load obviously increased,and the linear trend was 148.9 t·km-2·a-1 at a significant levele of P<0.05.(3)Since 1999,the implementation of the Grain for Green project was successful in improving the vegetation coverage in the study area.The changes in land use types at the three years are calculated.The results showed that the dominant land use types were farmland and grassland,which accounted for 46.79% and 49.94% of study area in 1995,respectively.By 2000,farmland and grassland always dominated the region,but farmland gradually decreased by 4.1%,and grassland and forest increased by 4.26% and 0.14%,respectively.Until 2010,grassland and forest dominated the region.Farmland decreased by 28.1%,whereas the grassland and forest rapidly increased by 9.23% and 18.01%,respectively,compared with 2000.Generally,from 1990 to 2000,the farmland slowly decreased,whereas the grassland and forest slowly increased.From 2000 to 2010,farmland rapidly decreased,whereas the grassland and forest rapidly increased.This change in trend was consistent with vegetation coverage under the Grain for Green project.The results also showed that the average vegetation coverage in the three years in the Upper Reaches of Beiluo River Basin gradually increased from 20.21% to 51.22% during the periods of 1995 to 2014.The trends of vegetation distribution on landform were analyzed in the upper reaches of the basin.It showed that the extent or the area of low vegetation coverage transforming to high vegetation coverage tends to be greater with the elevation lower,slope gradient larger,and slope aspect inclining to be shady.The difference of vegetation coverage inclined to enlarge from 1995 to 2007 and 2014.(4)The verified SWAT model was used to evaluate the hydrological response to climate variation and land use/cover change.From 1986 to 2012,the periods were divided into three stages.The year of 1986 to 1995 was baseline period.The year of 1996 to 2005 and 2006 to 2012 were the two changing periods.The results showed that the runoff depth gradually decreased in the two changing periods.The contribution of human activity to runoff changes were respectively 1.3 times and 3.5 times of climate change,respectively for the second changing period and the third changing period.The results also showed that with the increased vegetation coverage from 1995 to 2010,the annual surface runoff during these three years and the changing rates during these three periods presented a decreasing gradient from northwestern to the southeast and from north to south.Evapotranspiration gradually increased during these years and presented a decreasing trend from south,southwest to north,northeast part.And the soil water content was similarly with runoff and gradually decreasing,and presenting a decreasing trend from north,northeast to south,southwest part.(5)Combined with RUSLE and GIS,the temporal and spatial variation of soil erosion in 1990,2000 and 2010 were analyzed.The results showed that the study area suffered from primary land use changes,with increasing grassland and forest and decreasing farmland from 1990 to 2010.Based on that,the average soil erosion modulus decreased from 18189.72 t/(km2·a)in 1990 to 7408.93 t/(km2·a)in 2000 and 2857.76 t/(km2·a)in 2010.Compared with 1990,the average soil erosion modulus decreased by 59.0% and 84.3% for 2000 and 2010,respectively.Benefiting from the increased vegetation coverage and improved ecological environment,the soil erosion in this study area clearly declined.This research also found that the distribution of the three years of soil erosion was similarly based on topographic factors.The average soil erosion modulus gradually increased with the increase of the slope gradient,and 76.08% of the total soil erosion was concentrated in the region with a gradient more than 15 degrees.The soil erosion modulus also varied with slope aspects in the order of sunny slope>half-sunny slope>half-shady slope>shady slope.(6)After analyzing the extreme land use simulation results,we designed spatial optimized land use scenario based on the characteristics of terrain and land use in this study area.Compared with land use in 2010,farmland area decreased from 14.59% to 9.99%,grassland area decreased from 63.43% to 49.79% and forest area increased from 20.95% to 39.2% in optimized scenario.Accordingly compared with 2010,the runoff depth decreased by 16.98% and average soil erosion modulus decreased by 18.65% in optimized scenario.The hydrological response to the future climate scenarios were also analyzed,and the results showed that in yearly scale the simulated runoff depth gradually increased both in the scenarios of RCP4.5 and RCP8.5 from 2021 to 2050.In monthly scale,the runoff depth gradually decreased from January to March and from October to November both in the scenarios of RCP4.5 and RCP8.5.In the scenario of RCP4.5,the runoff depth probably increased from June to August.In the scenario of RCP8.5,the runoff depth probably decreased from June to August.