Research On Remote Sensing Model Of Arborous Layer Forest Carbon Storage

As the world’s most important forest ecosystem, forest takes important position in the global carbon cycle, so the research on the forest carbon reserves is of great significance under the background of many problems caused in the current global climate change. Traditional methods of estimating forest carbon stocks need to be observed in the wild in carbon storage, which consume large amounts of human and financial resources, will destroy the forest. Even now the most commonly used remote sensing estimation techniques; you have to collect some sample datum. To address these issues, this research launched a new approach:all major forest types in Fujian province as the research object, with the average temperature and precipitation grids and soil quality factors to rapidly build a database of productivity model, calculated forest productivity. Based on the survey plot data and the second such inventory data of Fujian province in 2003 year, combined with forest productivity fitting curve of forest age in small classes, calculation of tree ages forest productivity. Calculation of forest biomass utilization of forest productivity will be cumulative productivity each year, cumulative count using a forest growth curve correction, as the plot of forest biomass. After that use the remote sensing image, build the forest biomass and carbon storage of remote sensing model. Main conclusions of the study are as follows:(1) Based ArcGIS and ERDAS, using international common of USLE method and reference both at home and abroad calculation soil erosion method, using ERDAS software to get vegetation management c factor Grill grid data, and terrain LS factor Grill grid data, and precipitation erosion force r factor Grill grid data, and soil k factor Grill grid data, then through the function overlay of ERDAS make all factor figure superimposed even take, obtained Fujian province of soil erosion model, generated Fujian province soil erosion module figure.(2) Make full use of ERDAS9.2,ArcGIS10 software, average temperature, average rainfall in Fujian province based on data and main factors of soil quality data, successful establishment of climatic productivity of forests, forest productivity raster databases, modeling features to build a forest in Fujian province using ERDAS grid database of climatic productivity of forests, forest productivity, calculated from these three types of forests of Cunninghamia lanceolata and Pinus massoniana and broad-leaved forest in forest productivity.(3) Use the most recent a survey plot data in Fujian province and the last two types of inventory data, raster data combined with forest productivity has been fitting curve of forest age-productivity of different tree species, tree ages calculated forest productivity. Will be cumulative productivity each year, the cumulative value fixed as derived from forest biomass for the corresponding year, plot is used for remote sensing modeling of biomass.(4) select the band gray value related of bio-volume and the vegetation index, extracted the band of gray value, enhanced type vegetation index (EVI), and ratio vegetation index (RVI), and return a of vegetation index (NDVI), and poor value vegetation index (DVI), through DEM get elevation, and slope and slope to figure, amounted to 16 a variable for established bio volume model of since variable, using gradually return analysis law established Fujian province forest bio volume of remote sensing estimates model. Tests concluded that respective model coefficient of determination are 0.674,0.770,0.943.(5) Based on spectral information and investigation of remote sensing image data, use unsupervised classification and expert knowledge to extract land use types in Fujian province, as well as the Chinese FIR, pine and hardwood forests and other forest types feature information, and set up a corresponding grid database.(6) Based on ERDAS software, utilize average carbon content rate coefficient of forest tree species commonly used (0.5) and forest biomass estimation by remote sensing remote sensing estimation model for modeling forest carbon stock in Fujian province, and modular arithmetic are raster data images. Three forest of Cunninghamia lanceolata and Pinus massoniana and broad-leaved forest carbon stock raster data type for the corresponding mask to get the 2003 Chinese FIR and Masson pine and hardwood forest carbon storage of raster data in Fujian province, statistical results are as follows:4 966 300 5.96 t of fir forest carbon reserves, pine forest carbon reserves 462 113 98.9 t, hardwood forest carbon reserves 9 008 405 7.06t.2003 year fujian forest carbon density, cunninghamia lanceolata forest for 3 4.97 483 Mg/hm2, horsetail pine to 2,7.06 393 Mg/hm2, broad-leaved forest is 5 1.34 536 Mg/hm2, a total of 41.6 Mg/hm2. Based on Pseudo invariant features principle of relative radiometric calibration method for carbon storage model calibrated to establish 2012 carbon model of Cunninghamia lanceolata and Pinus massoniana and broad-leaved forest in Fujian province, carbon stock results for statistics:5 287 620 3.45 t of fir forest carbon reserves, pine forest carbon reserves 511 560 18.8 t, hardwood forest carbon reserves 9 936 469 2.36 t.