Biomass And Carbon Storagr Estimation Of Reed In Yellow River Estuary Wetland Based On High Resolution Remote Sensing

Terrestrial ecosystem carbon storage is the critical issue of global climate change research. As an important terrestrial carbon pool, wetland is one of the most productive ecosystems. Wetland plant’s carbon storage can effectively reflects the health status of ecosystem. It’s dynamic change can provide basis and decision support for protection, restoration and construction of wetland.This research utilizes measured spetrum and biomass and the corresponding period’s high reslution image of Yellow River Estuary Wetland on spring, summer and autumn as material. Basing on spetrum and image to build vegetation indexs respectively. Then use vegetation indexs as independent variables and biomass as dependent variable to process a variety of single variable regressions. Choose the highest determination coefficient regression model of each vegetation to conduct accuracy testing, the highest accuracy model is the biomass estimation model. Compare the spetrum based model and the image based model, use the image based model to complete biomass inversion. Finally, research the seasonal change of biomass and carbon srorage in study area. The main results are as follows:(1) The optimal spetrum based biomass estimation model of spring, summer and autumn are respectively cubic model of NDVI, exponential model of DVI and cubic model of NDVI, models’MRE are 20.98%,20.48% and 26.50% respectively, models’RMSE are 0.177 kg/m2,0.320 kg/m2 and 0.416 kg/m2 respectively. The optimal image based biomass estimation model of spring, summer and autumn are respectively cubic model of NDVI, cubic model of NDVI and cubic model of SAVI, models’MRE (Mean Relative Eorror) respectively are 17.86%, 18.17% and 18.85%, models’RMSE (Root Mean Squared Eorror) are 0.137 kg/m2, 0.237 kg/m2 and 0.335 kg/m2 respectively. The accuracy of spetrum based biomass estimation model and image based biomass estimation model are both declined with the change of season. The accuracy of spetrum based biomass estimation model is lower than the image based biomass estimation model in the same season. The accuracy of spetrum based model in autumn is much lower than other seasons due to the change of spetrum caused by earing.(2) The mean biomass, mean carbon storage, total biomass and total carbon storage are increase with the season change. Mean biomass of spring, summer and autumn are 0.376 kg/m2,0.802 kg/m2 and 1.004 kg/m2 respectively; Mean carbon storage of spring, summer and autumn are 0.165 kg/m2,0.353 kg/m2 and 0.442 kg/m2 respectively; Total biomass are 34036.732 t,72654.927 and 90961.900 t respectively; Total carbon storage are 14976.162 t,31968.1681 and 40023.2361 t respectively.(3) Divede study area into low carbon storage, moderate carbon storage, high carbon storage, water and mudflate, research the time and space change of every carbon storge type. Percentages of water and mudflat, high carbon storage are decreased with the change of season, water and mudflat is less distrubuted at the northeast of study area, high carbon storage are main distrubuted at the near of water and cropland; percentages of low carbon storage is increase first and after decrease, while, percentages of moderate carbon storage is decrease first and after increase, low carbon storage and moderate carbon storage are distrubuted equably at study area.