Study On Estimation Method Of Natural Grassland Biomass Based On UAV Multi-spectral Data

Natural grassland is vast in Xinjiang,with an available area of 50 million hm~2,ranking the third in the country,the northern Xinjiang region has 29 million hm~2,accounting for more than 50%of the Xinjiang,which is an important supply area for livestock products.In order to estimate aboveground biomass of natural grassland and grasp variation trend rapidly,exactly and effectively,aerospace remote sensing technology are usually used to do it,but the spatial resolution and temporal resolution of images acquired are low.Rapid development of unmanned aerial vehicle(UAV)and lightweight sensors brings out the possibility of low altitude remote sensing and high spatial resolution and spectral resolution image.At the same time,multi-spectral images tend to produce"same-spectrum foreign body","foreign body with the spectrum"and so on;which has brought difficulties to follow-up's treatment.Higher spectral resolution also increases the correlation between adjacent bands,and consequently a large amount of information redundancy,which not only brings computational complexity but also increases the time complexity.Therefore,it is a difficult problem in the application research to reduce the dimension processing of multi(high)spectral remote sensing data.The grassland in natural rangeland,northern hillside of Tianshan Mountain was selected as a typical study area,the remote sensing image acquisition platform of UAV was built by multi-rotor UAV,which equipped with light multi-spectral camera,and then based on spectral characteristics of vegetation,the best band index(OIF)and maximum correlation coefficient(MCC)to choose the optimal band combination.Based on the correlation between aboveground biomass collected and Vegetation Indexes(VIs),we established some relational models between aboveground biomass and VIs with analysis of regression,then we determined the optimum one with comparing their precision.Finally,we inverted the aboveground biomass using satellite image.The main work and related research results are as follows:(1)Aiming at the characteristics of more bands,larger amount of data and similarity in each other,we selected typical region in Ziniquan rangeland as research area.Multi-rotor UAV equipped with Micro MCA12 Snap was used to obtain multispectral images.We divided them into three groups after analyzing band spectral feature and peculiarity of vegetation,and then we used OIF and MCC to select the optimum band preliminarily.Finally,we determined the optimal band combination after taking the functions between biomass and VIs into account.The result showed there was no strict corresponding relation between OIF and accuracy of estimate model,using MCC to obtain the optimal band combination is not very ideal,may be the way was vulnerable to the influence of various factors,the method for extracting the only single band can't reflect the characteristics of the vegetation from various angles,etc.However,it can improve the accuracy of estimation model to some extent by selecting the optimal band combination.(2)In order to estimate the biomass of natural grassland and master its variation rule accurately and effectively,we considered the terrain factors on the impact of the spatial distribution of vegetation aboveground biomass fully.Then we analyzed the correlation of biomass aboveground and VIs using multispectral images(including near infrared band)and the measured data on the ground.Finally,we used regression analysis method and established the estimation model.The results proved that different types and the characteristics of vegetation brought about the same vegetation index correspond to different biomass,which was the root cause of the low relevance between total biomass and VIs.It can improve the relevance between them after considering the terrain obviously,and the model was accurate and corresponded to reality.(3)Combined with UAV characteristics such as low cost,high aging,high resolution and satellite remote sensing advantages of large scale monitoring,we applied previous research results and combined Landsat8satellite image,collaborative estimating models were established,the estimate models accuracy were above74%,The purpose of collaborative estimation was realized.