| The Response mechanism between vegetation and drought is intricate,the impact of drought on vegetation activity will change in different conditions.Furthermore,vegetation activity will affect regional climate,which will provide feedback on drought.In this work,the Standardized Precipitation Evapotranspiration Index(SEPI)was calculated through meteorological station data on Loess Plateau over the 2000-2016.Based on the SPEI,we explored the temporal and spatial characteristics of drought in this regional.Moreover,the Rotating Empirical Orthogonal Matrix(REOF)was used to investigated the drought feature partition.In addition,we analyzed the response of vegetation activity to drought in different condition by the same period Normalized Difference Vegetation Index(NDVI)and assessed the vegetation vulnerability to drought.Results showed that:1.Loess Plateau showed a trend of humidification during the study period.The SPEI increasing area was 76.79% of the whole regional,and about 10.53% area has significant increase trend which was concentrated in Liupan and Baiyu Mountain.The spatial distribution of SPEI trend in autumn and winter were consistent with the annual.The annual trend of SPEI was 0.18/10 a over the last decades.In each season,the summer SPEI trend was 0.27/10 a,indicating the summer humidification was the most obvious.The drought events on Loess Plateau were mainly mildly and moderately arid,while the annual frequency of mildly and moderately arid were decreasing during the study period.2.The Loess Plateau can be divided into four characteristic arid areas: the southeast area;the northeast area;the northwest area and the southwest area.The southeast area showed humidification trend during spring and summer and aridification trend during autumn and winter,while the other areas showed humidification trend in each season.The SPEI of each subarea have fit consistency,specially between the southeast area and northeast area;northeast area and northwest area.The metrological elements and underlying surfaces of each subarea has significate differences,the difference in precipitation was the most and the difference in land use was the smallest.3.The correlation between NDVI and SPEI was positive in growing season.The correlation increased from April to July and then decreased,during June to September the correlation between two indices was significant.Grass and shrub seemed to be the most sensitive to drought,the significant correlation area reached 62.54%;followed by cultivated land,the significant correlation area was 52.97%;Forest was the least sensitive to drought.In different terrain,the correlation between NDVI and SPEI was higher in hilly and low mountain,while lower in low plain and high mountain.The correlation has a nonlinear relationship with the change of elevation,humidity and temperature.With the increasing of elevation,the correlation firstly increased and then decreased,the maximum correlation appeared at 1500-200 m;with the decreasing of humidity,the significant correlation area increased to 70.26% at the wetness 0.2-0.4 and then decreased to 36.13% at the wetness less than 0.2;with the increasing of annual accumulated temperature,the significant correlation area increased from 17.92% at the range of less than 2000 d·? to 60.50% at the range of 2000 d·?-3000 d·?.4.Based on the response of vegetation activity on drought,we assessed the vegetation vulnerability to drought by Geography Detector and Analytic Hierarchy Process.Results showed that: The effect of factors on the vulnerability were: humidity>terrain>land use>elevation>temperature.On Loess Plateau,the vulnerability was mainly medium vulnerability(30.13%)and high vulnerability(36.62%).The spatial distribution of vulnerability was low on southeast and high on northwest.The determinant of vegetation vulnerability to drought was mainly humidity and terrain.During 2005-2015,the vulnerability on Loess Plateau was generally decreased. |
PDF Full Text Request