Study On Forest Biomass And Carbon Storage Estimation Of Cunninghamia Lanceolata In Minjiang Watershed Based On RS And GIS

Forest is the main part in terrestrial ecosystems, and forest biomass estimate is the foundation of terrestrial ecosystem carbon cycle and carbon dynamic analysis, which plays a very important role in the global carbon cycle research. In order to assess the function of forest in the global carbon balance properly, based on geomatics, the compatible biomass estimating models of singe tree and stand forest of Cunninghamia lanceolata and the non-linear remote sensing model of Cunninghamia lanceolata in Minjiang Watershed are built, using 2003 TM remote sensing image, field investigation data, previous research data collected and the 2003 forest survey datum of Fujian Province. The biomass and carbon storage of Cunninghamia lanceolata in Minjiang Watershed are estimated and analysed spatially by overlay. It comprehensively reflects the conditions and spatial distribution of biomass and carbon storage of Cunninghamia lanceolata in Minjiang Watershed. The main conclusions are as follows:(1) Based on the concept of compatible forest biomass model, the compatible biomass estimating models of singe tree and stand forest of Cunninghamia lanceolata in Minjiang Watershed are built by ambipolar non-linear simultaneous equations, using the 2007 field investigation data, previous research data of Cunninghamia lanceolata and the 2003 forest survey datum of Fujian Province. It has basically solved the problem of compatible forest biomass models. And the precision of the models are higher than 97%, which can be applied to practical production.(2) With forest type's thematic classification system, based on RS platform, the various forest types in Minjiang Watershed are classified by Expert Classification Repository, Supervised Classification and Delamination Classification. And the distributing map of woodland in Minjiang Minjiang is made. The precision of forest type's classification is 86.9%, the precision of Cunninghamia lanceolata's classification is 86.5% (3) Combining with GIS technique, the non-linear remote sensing modeling system of Cunninghamia lanceolata in Minjiang Watershed is established based on Matlab platform by B-P Nerve Network. The study adopts the quantified and qualified factors such as Digital Number of bands, Vegetation Index etc provided by RS imagery, Altitude, Slope and Aspect as independent variable, and taking biomass of sample plot as caused variable. The standard B-P Nerve Network is enhanced by the measures such as disposal unitary input data, enhancement net training learning arithmetic and so on. The results show that: the Nerve Network with quantized Conjugated Grads Arithmetic (Trainscg) has faster learning training speed, but the Nerve Networks with Bayesian Regularization Arithmetic (Trainbr) is superior in total relative error, average relative error and pre-estimating precision and has more powerful generalization. Therefore, the study adopts enhancement B-P Nerve Networks improving the ability of extending networks as the final remote sensing biomass model of Cunninghamia lanceolata forest in Minjiang Watershed, and estimates the biomass and carbon storage of Cunninghamia lanceolata in Minjiang Watershed. And the distributing map of biomass and carbon storage of Cunninghamia lanceolata in Minjiang Watershed is made.(4)Based on GIS platform, taking the distributing map of biomass and carbon storage of Cunninghamia lanceolata forest in Minjiang Watershed, digital elevation map, slope map and aspect map as basic maps, the distributing law of biomass and carbon storage of Cunninghamia lanceolata in Minjiang Watershed is studied, with spatial analyzing. The results show that: the total area of Cunninghamia lanceolata forest in Minjiang Watershed is about 747729.4hm², accounting for 0.524% of those in China; the total biomass is 42291574.86t, the carbon storage is 19874647.45t C, accounting for 0.54% of those in China. The average carbon storage of Cunninghamia lanceolata is 26.58 t·hm^-2 , higher than those of China, which is 25.77 t·hm^-2. The order of biomass and carbon storage distribution of Cunninghamia lanceolata forest from large to small following altitude classes change is: lower(400m-800m) >low(<400m) >middle(800m-1200m) >higher (1200m-1600m) >highest(> 1600m). In the lower altitude area, the biomass and carbon storage of Cunninghamia lanceolata forest is the largest accounting for 54.96% of the total biomass and carbon storage of Cunninghamia lanceolata forest; in high altitude area, the biomass and carbon storage of Cunninghamia lanceolata forest is smallest, only accounting for 0.06% of total: In the flat, the biomass and carbon storage of Cunninghamia lanceolata forest account for 25.99% of total; in the hilly area, the biomass and carbon storage of Cunninghamia lanceolata forest account for 74.01% of total. The order of biomass and carbon storage distribution of Cunninghamia lanceolata forest from large to small following gradient classes change is: slope>steep slope>slow slope>urgent slope>dangerous slope.