Estimation Of The Remote Sensing On Soil Carbon Storage Of The Ecotone Of Forest And Grassland In Hulunbuir Steppe

Forest grass ecotone is a transitional zone between grassland and forest.In view of the lack of research on soil carbon storage and the uncertainty of soil carbon pool,it is of great significance to monitor and estimate the carbon storage in the ecotone by combining remote sensing with ground sampling.In this experiment,the soil samples in the forest grassland ecotone between Hailar District and Wuerqihan Forestry Bureau of Hulunbuir city were measured for organic carbon,and the indoor reflectance data of the surface soil?0-10cm?was obtained by using the ground object spectrometer FS3 of ASD company.The narrow band data measured by FS3 instrument was fitted to the wide band data of Landsat 8 image on average Six vegetation indexes?NDVI,DVI,RVI,osavi,rdvi,EVI?,four salinity indexes?SI1,SI2,Si3,NDSI?and one brightness index?BI?were calculated according to the formula.The correlation between the calculated indexes and the actual soil organic carbon?SOC?content was analyzed,and the indexes with significant correlation were selected for model fitting,and Landsat 8 image was used The inversion is carried out.The results are as follows:?1?According to the investigation of the line plants in the forest grass ecotone,the forest vegetation is mainly composed of Betula platyphylla,Pinus sylvestnis,Populus davidiana and sparse forest,and the grassland vegetation is mainly composed of Gramineae,Rosaceae and Compositae.?2?In the soil of 0-30cm depth,the physical and chemical properties of soil decrease with the increase of soil depth,and the soil nutrients are mainly concentrated in the surface soil layer of 0-10cm.?3?There is a negative correlation between soil spectral reflectance and soil organic carbon content,which is significant in visible and near infrared band.?4?The vegetation normalization index?NDVI?and the soil salinity index three?SI3?have a significant correlation with soil organic carbon.Among them,the vegetation normalization index NDVI has a positive correlation with soil organic carbon,and the soil salinity index Three SI3 had a negative correlation with soil organic carbon,and the correlation coefficients r reached 0.619 and-0.636,respectively.?5?Linear regression analysis shows that the index estimation model based on indoor ASD data fitting is ideally y=0.396x-0.055?x=NDVI,R²=0.384,RPD=1.486?and y=-53.644x+0.065?x=SI₃,R²=0.404,RPD=1.582?.?6?Compared with a single index model,the multi-index combination model can significantly improve the correlation coefficient R2 of the model,but it cannot improve the prediction accuracy.?7?The correlation analysis results of the NDVI and SI3 indexes calculated from ASD data and the NDVI and SI3 indexes calculated from the image band show that the soil salinity index three?SI3?is the best image inversion index,and the established image inversion model It is:y=-19.51x+0.06?x=SI₃?,the inversion results are tested,and the verification result is the standard prediction error SEP=0.018,the ratio of standard deviation and prediction error RPD=1.27<1.4 failed to pass the RPD standard test It shows that the model based on indoor light source is relatively low in the accuracy of image inversion.