| Sparse mixed forest with trees,shrubs,and green herbaceous vegetation is a typical landscape in the afforestation areas of northwestern China.The vegetation is very sparse in this area,and the aboveground biomass of sparse woody vegetation is the dominant carbon sink in this area.It is a great challenge to estimate the woody aboveground biomass of a sparse mixed forest with heterogeneous woody vegetation types and background types by using remote sensing data.Therefore,this paper explores the estimation methods and applications of sparse woody aboveground biomass using optical remote sensing data at different scales.The main work of this paper is as follows:(1)The object-oriented machine learning algorithm was used to obtain the coverage information of trees,shrubs and herbaceous based on high-resolution UAV images,and three stratification schemes(non-stratification,stratification based on two vegetation types of tree and shrub,stratification based on five tree species)were used to construct the canopy coverage-woody aboveground biomass models.And estimation the sparse woody aboveground biomass using the UAV-corrected GF-2.(2)Six vegetation indices(NDVI,RVI,MSAVI,TCG,NDMI,NIRv)of Landsat-8OLI were used to construct different vegetation index-biomass models,and the optimal vegetation index was selected for estimating sparse woody aboveground biomass.(3)The vegetation index-woody aboveground models were constructed based on six vegetation indices and three herbaceous coverage levels of plots of UAV images interpretation,and the optimal vegetation index was selected for estimating sparse woody aboveground biomass.Interpretation of optical remote sensing images at different scales and the aboveground biomass estimation method of sparse woody,the research results of this paper are as follows:(1)High-resolution UAV remote sensing can be used to obtain priori knowledge of woody vegetation types and canopy coverage of trees,shrubs and herbaceous in sparse mixed tree-shrub forest.It can be used to improve the accuracy of vegetation canopy coverage which extracted from GF-2 remote sensing images.(2)The accuracy of coverage-woody aboveground biomass models is strongly influenced by vegetation type.The finer the classification of the woody vegetation is,the more accurate the coverage-woody aboveground biomass model is.(3)When woody vegetation type was not clear,the vegetation index-woody aboveground biomass model performed better in estimating sparse woody aboveground biomass.The accuracy of the vegetation index-woody aboveground biomass models was easily affected by green herbaceous vegetation coverage background,and considering the green herbaceous vegetation coverage background can significantly improve the accuracy of each vegetation index-woody aboveground biomass model for estimating woody aboveground biomass. |
