An Incremental Spatial Sampling Layout Strategy For Regional Soil Carbon Flux Estimation

Carbon dioxide and other greenhouse gases discharged from ecosystem directly affectclimate change. At the same time, accurate measurement of regional soil carbon flux plays animportant role in studying ecosystem carbon cycle. At present, single point measurement, fluxtower, and the use of remote sensing method combined with model are major methods forcalculating soil carbon flux among which sampling methods are applied to proceed single pointmeasurement of soil carbon flux. Due to strong spatial-temporal heterogeneity,flux estimationerror of area flux total cost is bigger using sampling measurement method. Combine tower of fluxand the means of remote sensing method to calculate carbon fluxes through empirical modelwhose uncertainty is still large.Emerging distributed self-organizing data collection network, Wireless Sensor network isapplied to monitor regional soil carbon flux, achieving sustainable and synchronous monitoring inarea. The positioning of sensor is the basis of this study.Based on layout problems in regional internet soil carbon flux,this paper puts forward a kindof layout strategy with incremental monitoring points (MDI): evenly determines a number ofinitial sampling points in the study area. Choose the largest dispersion among themulti-directional interpolation discrete degree formed with sampling sites in order to determinethe appropriate space sampling point layout and to achieve accurate estimation of regional soilcarbon flux with less sampling points.The number of initial point of MDI layout strategy. The optimal number of initial point is sixor seven has a certain influence on the layout of monitoring points. The right amount of initialpoints achieves guarantee to fully cover the entire monitoring area with fewer costs.Advantages of MDI layout policy relative to other layout policies. When sampling pointsn=10,MDI layout strategy on carbon flux estimation error is a little bit lower than random layoutstrategy and higher than uniform layout policy. With the increase of measuring points,error ofthese three layout strategies are all reduced. When sampling points n=40, MDI layout policy oncarbon flux estimation error (0.028)is12.5%lower than uniform layout policy (0.32) and30.0%lower than random layout policy(0.04).Relationship between estimation error and monitoring points to regional soil carbon flux. With the increase of monitoring points, estimation error of area soil carbon flux decreases.Decrease rate of MDI layout policy is faster than the rate of uniform layout and randomdistribution, and the fixed value of the trend error is smaller.Finally conduct incremental soil carbon flux monitor，SCF-S2002，use MDI layout strategyto finish experiment on monitoring soil carbon flux in experimental area, receiving distributionresults of SCF-S2002in Internet of things and spatial distribution results of soil carbon flux. Thisexperiment illustrates MDI layout strategy can carry on incremental layout monitoring stationsand proceed layout of increasing monitoring points according to different environment.