| Drought is a very common natural disaster,and in the last hundred years,the global climate is undergoing significant changes characterised by warming,with the occurrence of extreme drought events increasing and intensifying as global temperatures rise.Long-term persistent and frequent drought events have caused regional water shortages,reduced vegetation cover,land degradation,lowered biodiversity and reduced food production,which have brought great harm to national economies,especially agricultural production.More and more scholars have begun to focus on the spatial and temporal patterns of drought events and their drivers,of which drought indices are an important research tool for detecting large-scale,long-duration drought events.Regional environmental drought is the result of the interaction and integration of meteorological-soil-vegetation systems.However,traditional single-type drought indices focus on a certain aspect of drought events(e.g.meteorology,agriculture,etc.),and it is difficult to comprehensively describe the spatial and temporal characteristics of regional drought events and their complex features,therefore,in recent years,some scholars have started to use multiple single-type indices to construct comprehensive drought indices.Most scholars use water balance method,entropy weight method,principal component analysis method and probability statistics method to construct comprehensive drought indices,but these methods cannot reflect the non-linear relationship between variables well,while kernel entropy component analysis method carries out entropy component analysis in the feature space as a way to realize continuous transformation of data,which has good non-linear processing ability.At present,in the Loess Plateau region,the construction of a comprehensive drought index based on the meteorological-soil-vegetation system and the study of the relationship between land use,vegetation cover and the comprehensive drought index are still weak.Therefore,it is necessary to carry out research on the construction of a comprehensive drought index based on meteorological drought index,soil drought index and vegetation cover index,and to analyse the relationship between land use,vegetation cover and comprehensive drought index in the Loess Plateau region,so as to provide some theoretical reference for the comprehensive and systematic drought prevention and drought relief work in the Loess Plateau region.In this study,the standardized precipitation evapotranspiration index SPEI,the standardized soil moisture index SSI and the vegetation cover index NDVI were selected to characterize the meteorological drought,and the spatial and temporal evolution of these three indices on the Loess Plateau from 1981 to 2020.The spatial and temporal variation characteristics of the NDVI were verified and analyzed.The following main conclusions were achieved:(1)The SPEI,SSI and NDVI of the Loess Plateau at different time scales(monthly,seasonal and annual scales)from 1981 to 2020 were calculated based on the meteorological data,GLDAS soil moisture data and NDVI data of the Loess Plateau from 1981 to 2020,and the spatial and temporal evolution characteristics of these three single types of drought indices were analysed.The results showed that the SPEI showed an overall decreasing trend at monthly and annual scales,but the trend was not obvious;at seasonal scales,there was a significant decreasing trend in spring,a relatively stable trend in summer,and a significant increasing trend in autumn and winter;the SSI showed a significant increasing trend at both monthly and annual scales;at seasonal scales,there was a significant increasing trend in spring and winter,and a decreasing trend in summer and autumn,with a larger decrease in summer.NDVI shows a significant upward trend on monthly,seasonal and annual scales,with large fluctuations around 2000 due to the full-scale promotion of the return of farmland to forest and grass.The meteorological index SPEI showed an overall decreasing trend during the 40 years from 1981 to 2020;spatially,the area with an increasing and decreasing trend of SPEI accounted for 55.33% and 44.67% of the Loess Plateau area,respectively,with significant increasing areas mainly concentrated in central Shanxi,central Ningxia,most areas in Guanzhong and northern Shaanxi,Jingyuan County in Gansu and eastern Qinghai.Soil drought index SSI shows an overall non-significant upward trend,i.e.the trend of soil drought is weakening;SSI-12(interannual trend of soil drought)shows an upward trend and a downward trend,accounting for 66.01% and 33.99% of the Loess Plateau area respectively,and the downward trend areas are mainly concentrated in southeastern Shanxi,central Ningxia and some parts of southern Shaanxi.The vegetation cover index NDVI shows a significant upward trend overall,i.e.the vegetation cover is increasing;the area with upward and downward trends accounts for 82.74% and 17.26% respectively,and the downward trend areas are mainly concentrated in central Shanxi,northern Ningxia,some parts of southern Shaanxi in Shaanxi,eastern Gansu and some parts of Qinghai.(2)A comprehensive drought index was constructed based on the kernel entropy component analysis method(KECA)and three single-type indices,and the spatial and temporal variation characteristics of the comprehensive drought index were analysed on the basis of accuracy verification.Through comparison with typical historical drought events,it was found that the composite drought index was more consistent with the actual situation than the single-type drought indices.Specifically,the integrated drought index shows a nonsignificant decreasing trend on both interannual and monthly scales,i.e.it shows an intensifying drought trend,and in terms of seasonal scales,the trend is non-significantly increasing in spring and summer,showing a non-significant decreasing trend in autumn,and a significant decreasing trend in winter.the area where KECA-12(integrated drought index interannual variation)shows an increasing and decreasing trend accounts for 32.58% and67.42% of the Loess Plateau area respectively.The area of KECA-12(interannual variation of the composite drought index)with an increasing trend and a decreasing trend accounted for32.58% and 67.42% of the Loess Plateau area,respectively,and the significant increasing areas were mainly concentrated in the Datong-Shuozhou area in northern Shanxi and most of the southern Inner Mongolia.The Pearson correlation coefficients of the integrated drought index and the three single types of indices were further analysed.The results show that the Pearson correlation coefficient between the comprehensive drought index and its three singletype indexes is small.(3)The relationship between land use,vegetation cover and the comprehensive drought index of the Loess Plateau in 1990,2000,2010 and 2020 was analysed.In terms of dynamic changes,the degree of drought first decreases continuously and then increases in the 2020 s,with a more obvious increase in the degree of drought in forest and arable areas.Among all NDVI types,the proportion of drought-free medium and low NDVI is the highest,followed by low NDVI and medium NDVI,and the proportion of other types of NDVI is relatively small.The systematic analysis of the integrated drought index and land use and vegetation cover of the Loess Plateau provides a theoretical basis for ecological protection and vegetation restoration work as well as agricultural production in the Loess Plateau region.While developing agriculture,ecological construction should be emphasized to promote the restoration of the ecological environment of the Loess Plateau and to reasonably adjust the land use structure and use it as a basis for improving agricultural land in the Loess Plateau. |
PDF Full Text Request