Research On Estimation Of Aboveground Biomass And Aboveground Carbon Storage For Tamarix In Changyi National Marine Ecological Special Protection Reserves Based On Remote Sensing

Biomass and carbon storage of vegetation estimation is one of the hot topics of many scholars study.It is an important indicator for Biomass to measure the production and stability of ecosystem and carbon storage is an important indicator of the health of the biosphere.Wetland is one of the largest carbon pool in the world,the area accounted for land area of 4%-6%,but carbon reserves account for 35% of terrestrial carbon storage,it has a strong carbon sequestration capacity and a huge potential for carbon sequestration.Wetland vegetation is an important component of coastal wetlands and play a important role of energy fixed and organic material producers in the ecosystem.Spartina alterniflora,tidal beach Suaeda,Tamarix,reed vegetation often grow on coastal wetlands.As a typical specie in coastal wetland,Tamarix grows in intertidal and supratidal zone and is very important for protection of the coast and the improvement of the shoal.The "NAN HONG BEI LIU" wetland restoration project have been put forward clearly under the "13th Five-Year" plan of national economic and social development,which show that Tamarix plays a key role in coast protection.The method of estimating above ground biomass(AGB)and aboveground carbon storage can be divided into three categories,which are the method of ground measurement,integrated model method and remote sensing(RS)monitoring method respectively.Compared with the traditional ground measurement method and integrated model method,the way of vegetation biomass of wetland is estimated by remote sensing,which can combine the field characteristics with the remote sensing feature and solve some problem,for an example,wetland environment is complex and difficult to enter.In addition,remote sensing monitoring method can save manpower and material resources,and can carry on large-scale spatial inversion of biomass and carbon storage,and analyzes the spatial pattern characteristics,However,the research of estimating Tamarix biomass and carbon storage based on domestic GaoFen("GF-1")image is not a lot,information about it is less.Under the support of “Coastal zone dynamic monitoring and demonstration application of National High Resolution Project” and “Marine public welfare project of Development of Carbon Sequestration Information Collection,Supplementary Investigation and Field Data Acquisition System for Typical Coastal Wetlands”,the paper take major areas of National Marine Ecological Special Protection Reserves as the study area,Pleiades high resolution image is used to complete extraction of Tamarix distribution area in the study area,GF-1 data which is developed by ourselves and in-stiu measured data is used to establish Tamarix AGB and aboveground carbon storage estimation model and the model is applied in Study area.At the same time,the result estimated by converting biomass to carbon storage is compared with by remote sensing directly.The main results are as follows:1)Fusion data(spatial resolution is 0.5m)of Pleiades Multispectral and Panchromatic is used to extract distribution range of Tamarix and NDVI is established with object-oriented classification method based on sample.Fused image and NDVI gray data are segmented or merged.For the purpose of improving the classification accuracy of Tamarix,it is found that the optimal segmentation scale is 30 and the merging scale is 80 and surface features in the study area are divided into Tamarix,salsa,miscellaneous grass,roads,water and shoals,bare land,the overall classification accuracy is 88.6% and the Kappa coefficient is 0.81,classification results is ideal,besides,production accuracy and user accuracy of Tamarix are 87.6% and 95.71%,respectively.In order to better extraction accuracy of Tamarix,Manual visual interpretation method is chosen to classification details by contrasting with original classification image.The result show that distribution area of Tamarix is 1235.1 hectares,accounting for 64.5% of the total area of the study area.2)Twelve remote sensing information variables were studied by using GF-1 PMS 2 data.The results of correlation analysis showed RS information factors such as the image band 3(B3),RDVI,RVI,DVI,MSAVI,NDVI,SAVI,which is correlated respectively with Tamarix aboveground biomass;By biomass fitting regression analysis,MRE,RMSE error and measured value-estimated scatter plot comprehensive evaluation,The Research revealed determination coefficient of BP neural network and RVI cubic polynomial model are relatively high,which is 0.78 and 0.77,MRE of two estimation models is 14.44% and 26.79%,their RMSE is 0.09kg/m2 and 0.16kg/m2,respectively,so the models estimation accuracy are higher and are used to estimate AGB of Tamarix individually,however the application result of BP neural network model is not ideal,it have no practical significance.Conversely,application effect of cubic polynomial model with RVI is better,the average AGB of Tamarisk is 0.38kg/m2 and the total biomass of 19336 t in the study area.3)Aboveground carbon storage of study area is estimated based on AGB RS estimation theory.The study of correlation analysis reveal that RS information factors like the image band4(B4),MSAVI,GDVI,NDVI,SAVI,DVI,RVI,EVI,which is correlated respectively with Tamarix aboveground carbon density,and the estimated results of DVI S function model is better than others.The determination coefficient of S function model with DVI is 0.69,MRE and RMSE of model is 25.36%,0.04kg/m2 separately.In addition,in-stiu measured Tamarix AGB and carbon density data is applied to calculate average carbon content rate of Tamarix,average carbon content rate is 0.33.The carbon density of the study area is calculated based on RS estimation results of AGB.By comparing the calculated results both of them,study show that the method based on RS for direct inversion of carbon density is better than that aboveground carbon density is estimated based on RS estimation results of AGB.Estimated results of carbon density in the study area based on RS reveal that the Tamarix woodland average carbon density is 0.38kg/m2 and the total aboveground carbon storage is 8057.78 t.