Remote Sensing Estimation Of Vegetation Biomass And Allometry Characteristics Of Different Plants In Minqin Desert Area

Plant biomass is a reflection of ecosystem productivity,and the accumulation of biomass is one of the most important functions of organisms.The inversion of above-ground biomass in desert areas by remote sensing technology can reflect the productivity level of the study area,which is important for the study of regional carbon cycle and energy flow.Unlike forest ecosystems,desert areas are extremely water-scarce and the dominant species are mostly drought-tolerant shrub species,and the biomass allocation mechanism has its uniqueness.Remote sensing technology can be used to invert the growth of above-ground vegetation by optical principle and combine with plant anisotropic growth model to study the biomass allocation in the study area,which is conducive to an in-depth study of the interspecific biomass allocation mechanism of plants in the protected area.In this paper,Minqin Liangucheng Nature Reserve in Gansu was selected as the study area,and individual biomass modeling was first conducted for the dominant shrubs in the study area.Then,using the multispectral remote sensing data Landsat8-OLI,the vegetation indices with large correlation with biomass were selected to invert the above-ground biomass of the study area and select the best prediction model.And the plant distribution was analyzed according to the study area and spatial autocorrelation.Finally,the desert vegetation growth mechanism was studied based on the anisotropic growth of typical plants by applying the R language package(SMATR),and the influence of meteorological factors(temperature,rainfall)on the biomass allocation strategy was investigated,and the main findings were as follows:1.Biomass estimation models were developed for eleven typical dominant plants in the reserve,and the goodness of fit of individual biomass estimation models for dominant plant communities in the study area was greater than 0.7.The best-fit variables were canopy area S(C1×C2)for membranous jatropha.Basal diameter D for Artemisia annua.Plant volume V(SH)for camelina,date palm,lemon,and bitter bean.The best fitting variable for red sand,serrated white thorn,and small fruit white thorn is D~2H.The best fitting variable for Haloxylon ammodendron and Pearl Salsola is DH.2.In the univariate regression model fitting of vegetation index and aboveground biomass in the study area,the NDVI index has the best fitting effect with aboveground biomass.The optimal regression fitting model is the NDVI quadratic regression model(y=932.865x2009.516x~2-37.790,R~2=0.824,MAE=6.205665,RMSE=7.987759).In the multiple linear regression equation,the fitting equation R~2 with NDVI,EVI,and GRVI as independent variables is the largest(Y=471.384 NDVI-2.925 EVI-7.074 GRVI+4.335,R~2=0.872,SEE=4.689860,MAE=3.391084,RMSE=4.494362).Perform accuracy tests on different types of fitting models,and the multiple linear regression model has the highest accuracy,the smallest error,and the most stable.Therefore,it is determined that the optimal fitting model is the multiple linear regression model.3.Using the optimal aboveground biomass regression model to estimate the AGB of Liangucheng Nature Reserve,a distribution map of aboveground biomass was obtained.The aboveground biomass of the reserve was approximately 1.59*10~5 tons,with significant regional differences.The biomass range of different zones is between 0-37g/m~2 to 304-495g/m~2.The distribution of aboveground biomass has a significant spatial correlation,Moran’s I=0.926,showing a significant aggregation state.4.The four typical dominant plants selected in the reserve had different patterns of biomass distribution among different organs and above and below ground,and there were large interspecific differences.The proportion of leaf biomass:Nitraria tangutorum Bobrov>Reamuria soongorica>Calligonum mongolicum>Sophora alopecuroides.The proportion of stem biomass:Nitraria tangutorum Bobrov>Calligonum mongolicum>Reamuria soongorica>Sophora alopecuroides,and the proportion of root biomass:Sophora alopecuroides>Calligonum mongolicum>Nitraria tangutorum Bobrov>Reamuria soongorica.The above-ground and below-ground biomass of bitter bean and white spur showed a similar growth relationship,and the above-ground and below-ground biomass of red sand and sand date showed a significant allometric growth relationship.5.Mean annual temperature and mean annual rainfall had important effects on biomass allocation.Aboveground and belowground biomass increased significantly with increasing temperature;aboveground biomass increased significantly with increasing rainfall,while belowground biomass and total biomass tended to decrease.Plant root-shoot ratio allocation was not affected over a smaller range of temperature and rainfall variation.