Spatiotemporal Features Of Precipitation And Simulation Of Surface Runoff In Xiaohe River Basin

Xiao River is the second largest tributary of the Fen River,locates in the northeast of the middle section in Fen River Basin.Spatiotemporal features of precipitation and surface runoff in the area have important effects on social and economic development,ecological environment improvement as well as wellbeing.It has great significance to study the precipitation and surface runoff in the area for scientific management of water resources.Based on the datasets both monthly total precipitation of 24 observation stations distributed in the Xiaohe river basin from 1971?2011 and daily runoff from Lujiazhuang hydrological observation station from 1995?2010,We used the Mann-Kendall mutation test,moving t-test method,Morlet wavelet analysis and GS+V9 to analyze the spatiotemporal distribution features of the precipitation,furthermore the ArcSWAT2009 model was used to explore the applicability in Xiaohe River Basin(XHRB).The results were as follows:(1)The results from trend analysis of Mann-Kendall mutation test showed that the annual average precipitation of Xiaohe river basin from 1971 to 2011 exhibited a decreasing trend(Q<0 or Z<0)in annual precipitation,and in summer,flooding and winter seasons,respectively,in XHRB.The corresponding decrease rate was-5.39 mm/10a,-7.28 mm/10a,-5.35 mm/10a and-0.24 mm/10a,respectively;but the precipitation showed an increase trend in spring and autumn(Q>0 or Z>0)with an increasing of 4.44 mm/10a and 3.45 mm/10a.The statistics Z value showed that the change trend of precipitation in the same time series was not obvious.(2)The analysis from four methods of cumulative anomaly method,the MK mutation test,yamamoto method and moving t-test method all showed that mutation in annual precipitation occurred in 1979 and 1997,and that mutation in precipitation in both spring and winter occurred in 1982 and 2000.No mutation in precipitation occurred summer and autumn seasons.Few overlap occurred for four test results.(3)The cycling analysis results of annual average precipitation from Morlet wavelet method showed that main precipitation cycle duration in the period of 41 years was 6a,2a,4a,32a and 16a;32a,22a,4a,7a and 13a for the rainfall variation in spring;6a,2a and 18 a for the rainfall variation in summer;6a,2a,33a,and 16 a for the rainfall variation in flood season;9a,2a,32a and 21a for the rainfall variation in autumn;13a,3a,6a and 32a for the precipitation variation in winter.Until 2011 the precipitation at all temporal scales showed a unclosed abundant precipitation period,indicating that there is a probability of increasing trend in precipitation in all temporal scale in the future of a few years or a decade years in the basin.(4)The analysis results using Geostatistical Analyst Module of Arcgis9.3 and GS+V9 showed that the co-kriging interpolation has the highest precision in XHRB,and this method is applicable to all series;the precipitation distribution in space in XHRB is generally characterized by relatively abundant on both ends of the north and south,and less on both sides of east and west.The more precipitation was in the central strip of the basin,showing a northwest-southeast direction;Precipitation variation rate(Q)in the basin generally showed a trend that the variation was more in the east part than in the west part,and more in north than in south in the basin.(5)The analysis result of modeling and simulation the surface runoff using the SWAT model for XHRB showed that the SWAT model could be well used to simulate changing process of surface runoff of Xiao River basin(Calibration period r=0.88,Ens=0.71;Validation period r=0.78,Ens=0.73),the trends of observed and predicted were similar.Obviously,the SWAT model can meet the requirements of simulation precision,and it is suitable for the Xiao river basin runoff simulation.