Multiscale Remote Sensing Monitoring About Achnatherum Splendens Aboveground Biomass In Hulun Bision

Achnatherum Splendens is perennial dense plexus non-zonal gramineous vegetation,which is a widely distributed in Inner Mongolia plateau,Loess Plateau,eastern Qinghai-Tibet Plateau and other arid and semi-arid typical steppe area.Achnatherum Splendens is not only important feed for livestock in pastoral areas,but also serves as a natural barrier to protect the grassland ecosystem.Its changes of succession is the response of vegetation to the climate and human activities in the basin,and it is an indicator of the changes in the ecological environment of the grasslands.It plays an important role in the climate of Hulun basin in arid and semi-arid regions.Therefore,timely and accurate estimation of achnatherum splendens biomass can promote scientific management and rational utilization of grassland resources,and provide data support for relevant departments to plan and manage human activities such as grazing and ensure the sustainable development of ecological grassland in Hulun Lake basin.Achnatherum splendens,as a high-density vegetation,is more different from the typical grassland vegetation.In order to reduce the spatial heterogeneity of grassland,refine the remote sensing estimation method of aboveground biomass.This paper selects the widely distributed achnatherum community in the Hulun Lake Basin as the research objective,and used three data sources,which were composed of satellites-UAV-ground measurement data,to perform biomass longitudinal upscaling transformation analysis of achnatherum splendens.That is,the field collection data of achnatherum splendens biomass was inverted through two times of upscaling,and finally study the spatial distribution of its biomass in the Hulun Lake Basin.On the basis of this,the exploration of its soil environment was carried out,which could provide a scientific basis for the follow-up study on quantitative remote sensing of achnatherum splendens.The main content were as follow:（1）Achnatherum splendens community spatial distribution extraction:according to Landsat8 images and the sampling sites,the pure achnatherum splendens community patches were selected as the training sample set to analyze its spectral characteristics（the main analyisis spectral differences between achnatherum splendens and typical steppe grass）,shape indexes,spatial relationship and texture information.The classification images were divided into the best scale and combined with the training sample to build a rule set for image classification.The spatial distribution was extracted and the result was corrected and finally verified the classification accuracy.The result showed achnatherum splendens extraction accuracy was 88%.（2）Biomass upscaling inversion of achnatherum splendens:The measured biomass data of 1m×1m was inverted to 10m×10m biomass by the fractional vegetation cover（FVC）of achnatherum splendens was calculated by RGB images of unmanned aerial vehicles.Afterwards,the relationship between FVC of achnatherum splendens quadrats and the biomass sensitivity of 10m×10m was analyzed,and the accuracy,feasibility of the scale inversion of achnatherum splendens FVC was verified.Finally,the biomass index of achnatherum splendens was built based on the Landsat8 spectral index,and the biomass of 10m×10m was inverted to the 30m×30m biomass data to obtain the biomass spatial distribution of achnatherum splendens.（3）Construction of Pure Vegetation Index Model of biomass achnatherum splendens:The construction of the pure vegetation index model of achnatherum splendens consisted of two parts:background information pixel decomposition and pure vegetation index linear biomass inversion model construction.In this paper,the improved three-band gradient difference method is used to invert the FVC of achnatherum splendens in the drainage area(R²=0.65,RMSE=0.098（P<0.001）and accuracy of EA=78.95%,and then the background information of the achnatherum splendens mixed pixel was decomposed into pixels and the pure reflectivity of achnatherum splendens was obtained.The correlation analysis between pure reflectivity and biomass of achnatherum splendens showed that the correlation between biomass and reflectance in the near-infrared region was the highest,R² was 0.43,followed by red and short-wave infrared,R²was about 0.35.The worst correlation with reflectivity of blue and green bands,R² is only about 0.04.Finally we determined the spectral indices in the construction of indica biomass inversion model:NDVI,EVI,SAVI,RVI,SIPI,NDWI（normalized vegetation moisture index）,CI_green,and NDSI₆₄.The result showed that EVI index model had the best effect,with the R² was 0.74 and the RMSE was 98.74kg/100m²,while the model validation R² was 0.87,the RMSE was 40.17kg/100m²,and the estimation accuracy EA was 61.89%.Based on this,the biomass of achnatherum splendens in the Hulun Lake Basin was inverted and the spatial distribution of its average biomass was mapped.The results showed biomass of achnatherum splendens in the Hulun Lake basin was mostly20kg/100m²,of which the biomass of achnatherum splendens（40-80kg/100m²）was mainly concentrated in the west of Hulun Lake.The highest biomass in the surrounding area of the Hulun Lake（>80 kg/100m²）.Finally,the soil environment PH,nitrogen,phosphorus and soil organic matter of achnatherum splendens community were monitored,and the correlation analysis of achnatherum splendens biomass and soil physiochemical properties was taken,results showed there was no distinct relationship between them.