Remote Sensing Monitoring On Drought In Mount Tai Based On Modified TVDI

Drought is one of the most common and complex natural disasters in the world.It not only causes huge losses to human production and life and social economy,but also pose s a major threat to the ecological environment.Compared with the traditional point sampling method,remote sensing monitoring has obvious advantages in time and space scale,and can make up for the shortage of remote mountainous areas which is not conducive to the setting of monitoring points.Drought remote sensing monitoring research has become a hot spot at present.Among them,Temperature Vegetation Drought Index(TVDI),which was extracted from the feature space constructed by land surface temperature(LST)and vegetation index(VI),has been widely used in drought monitoring However,LST needs high-precision thermal infrared remote sensing data.Many images have no thermal infrared band and need to be superimposed with other images,and the spatial resolution of thermal infrared band is usually low.Mount Tai is a world natural and cultural heritage and a national geopark,and a typical forest distribution area in warm temperate monsoon climate zone.It is of great practical and research value to study its drought.To solve the problem that LST needs thermal infrared band,on the basis of previous work experience and referencing the construction method of TVDI feature space,T-indexes were constructed with Landsat OLI bands,and the optimal T index was selected to replace LST forming feature space with FVC to obtain MTVDI.The purpose was to provide a convenient reference method for drought monitoring in Mount Tai,and also could provide a new idea for the application of feature space method in drought remote sensing monitoring.The main research contents and results were as follows:(1)The T-indexes significantly correlated with LST was selected.By using multi temporal Landsat OLI bands,69 T-indexes were constructed through band combination,and four T-indexes with the highest correlation with LST were screened out:(B4-B1)/(B1+B4),(B7-B1)/(B1+B7),(B4-B2)/(B2+B4)and(B7-B2)/(B2+B7),and the correlation coefficients were between 0.61 and 0.76.(2)The MTVDI with significant correlation with TVDI was determined.Referencing the construction principle of TVDI,the feature space was constructed with four selected optimal T-indexes and FVC respectively,and the dry and wet edges were fitted to obtain MTVDI;the optimal MTVDI determined by the correlation analysis with TVDI was from the feature space composed of(B7-B1)/(B1+B7)and FVC,and the correlation coefficients of 11 phase were between 0.63 and 0.75.(3)The accuracy of the optimal MTVDI was verified.The correlation coefficient between TVDI and MTVDI was 0.72,and the correlation coefficient between the field measured soil moisture data of Mount Tai and MTVDI was-0.76,which indicated that MTVDI could retrieve drought at a certain extent.(4)The spatial and temporal distribution characteristics of drought in Mount Tai were clarified.Based on the 11 phase MTVDI,the spatial distribution of drought in Mount Tai showed that: the temporal variation characteristics were that the drought gradually decreased from January to July,and the drought continued to increase from July to December;the spatial distribution characteristics were that the central and western regions were humid,and the surrounding low-altitude regions were relatively dry,which were basically consistent with the TVDI retrieval results,and consistent with the actual spatial distribution of soil moisture,and MTVDI showed more precise drought distribution.In a word,the MTVDI constructed in this study has been proved to be reasonable and feasible which was an effective exploration.It could provide a theoretical reference for drought monitoring.