Research On Spatial Distribution Of Soil Carbon Flux Based On Bayesian-kriging

The soil carbon storage is the main part of the global total soil carbon and occupies an importantposition in the terrestrial carbon cycle.Soil carbon flux is an important factor to measure soil carbonbalance of terrestrial ecosystems. The spatial distribution of soil carbon flux can reflect the distributionof carbon storage in some degree, reduce the uncertainty of carbon balance within the region, and lay afoundation for carbon trading. Based on the improved kriging interpolation algorithm, this paper do amore in-depth study for the spatial distribution of soil carbon flux.The main research results:1,Using dynamic transform point placement methods to get the data source.In the process oflong-term monitoring regional soil carbon flux, every once in a way to adjust the measurementlocations is used to reduce soil carbon flux uncertainty. Provide more reliable data basis for estimatingregional soil carbon flux.2,The bayesian-kriging interpolation algorithm method is introduced to estimate regional soilcarbon flux. bayesian-kriging indirectly use of the variable information which is measured at theestimated point around.And it is based on the weighting coefficients of spatial location and parametersof sample points and the predicted points.Then it combined with known geometric distributionobtained prediction point interpolation.Ultimately it can reduce the error value which caused by otherinterpolation algorithm and make the estimate result more accurate.3,Using SCF-S2002and dynamic transform point placement methods on the data processing andanalysis, and verification and comparison. Then based on the bayesian-kriging interpolation algorithm,estimate regional soil carbon flux. Finally complete the display of the measured regional soil carbonflux spatial distribution. From the spatial distribution we can see: soil carbon flux is higher among thestudy area where terrain is higher and vegetation types is more complex; soil carbon flux is relativelylow in the flat place where vegetation type is relatively simple.