The Variation Of Climate Water And Sediment Response To The Vegetation Restoration In Northern Shaanxi Provinence In Loess Region

With the continuous development, human began to realize the importance of ecological construction.In later of the20th century, large-scale Grain for Green Project were carried out in China. Ecological environment is very frangibility in Loess Plateau, where soil erosion is seriously, since returning farmland to forest project were implementation, vegetation restoration has achieved significant results, a significant increase in vegetation coverage. To study the effect of vegetation restoration on climate and sediment change, the result not only can afford benefit assessment of restoration in Loess Plateau Ecological a reference, can also provide a scientific basis for ecological restoration and soil erosion control.In this paper Loess were selected as the study area,with GIS and RS technology and statistical methods support, based on years of meteorological data Yulin, Dingbian, Wuqi and Yan’an and2000-2010vegetation change research, analysis of vegetation change impact on the local climate; according to woodland survey data and national water census soil erosion data, analyze the impact of forest spatial distribution and its effect on soil erosion;spatial data including land use, soil type, digital elevation models and the corresponding attribute data were collected, and with years of monitoring data Wuqi hydrological stations of the study area by simulating runoff and sediment on Loess Hilly Region to study the applicability of the distributed hydrological model SWAT; runoff and sediment were simulated under different land use through the SWAT model, to analyze the role of vegetation restoration played in water and sediment reduction. Finally, the analysis results obtained are as follows:(1) Since1957, the heating trend in the study area was obvious while both annual precipitation and relative humid showed decrease trends; annualairtemperature kept increasing significantly after the GrainforGreenProject and consistent with long trend. Daily highest temperature increase most significantly in February (increase0.76-1.90℃at four stations).Daily lowest temperature increase most significantly in September (increase0.76-1.82℃at four stations). Contrast to long term trend, daily highest temperaturedidn’t increase but decrease; Monthlyrainy days decreased in spring and summer (decrease0.89-3.71days at four stations) while monthly precipitation and daily lowest relative humid didn’t change significantly. Daily lowest relative humid in September increase significantly4.19-8.23%at four stations; with comparison with change of vegetation coverage from2000-2011, climate response to vegetation was revealed to some extent. Increaseof vegetation coverage influenced air temperature and humid. As the vegetation coverage was lower and increased more, itinfluenced more on abating the heating trend and increase humid.(2) The system cluster analysisresults showed that:39Basin can be divided into four categories, there are significant difference of features of forest space distribution among the various types of watershed;(2) the increase of forest vegetation coverage is a major inhibiting factor to soil erosion, meanwhile the proportion of forest area, forest patch density, shape index addition will work in suppression of forest soil erosion, within a certain range, the woodland location near the top of the hill will promote the soil erosion.(3) Average annual runoff simulation, Nash-Sutcliffe efficiency coefficient Ens and correlation coefficient R2of the model can reach0.8above during calibration period and relative error is also smaller only-1.40%. During verification period Nash-Sutcliffe efficiency coefficient Ens and the correlation coefficient R2is also greater than0.6, the mean relative error is only3.27%, the annual runoff simulation meet the applicable accuracy of the model. Average month runoff simulation, the results of both verification and calibration period are poor; annual sediment production simulation, Nash-Sutcliffe efficiency coefficient Ens of the model were0.74and the correlation coefficient R2of the model were0.77, the average relative error is0.28%, simulation results was satisfactory during verification, the results improved but still poor after removing rainfall anomalies.(4) trend test under the1983to1992period, the measured average annual runoff was significantly increased. The averagewater reduction effectswas10.1%; period from2002to2009, the average annual runoff measured was significantly reduced, The averagewater reduction effects was0.59%; trend test under the1983to1992period, the measured annual sediment yieldwith a significant increasing trend, with an average of55.9%sediment reduction effect; period from2002to2009, the measured annual sediment yield was significantly reduced, the average sediment reduction effect was45.8%; vegetation water reduction and sediment reduction effect at a period of time are concerned with the amount of runoff and sediment yield in that period, within a certain range the correlation were positively.