Estimation Of Soil Carbon And Nitrogen Storage In A Typical Wetland In The Lesser Khingan Mountains

Although wetland area accounts for only 6 %-8 % of the global land area,wetland carbon storage accounts for 1 / 3 of the global soil organic carbon pool,which is one of the important carbon pools in terrestrial ecosystems.The swamp wetland is the main wetland type.Because of its long-term over-humid growth environment,the organic matter can not be completely decomposed,so the organic matter can be accumulated in a large amount,which plays an important role in the carbon and nitrogen cycle of the ecosystem.Under the background of climate warming and permafrost degradation,the carbon and nitrogen accumulation and carbon and nitrogen cycle of marsh wetlands have become a topic of widespread concern in academia.Xiaoxing ’an Mountains is located in the middle and high latitudes of China.It is the southern boundary of permafrost degradation moving northward.Swamp wetlands are widely distributed.Accurate estimation of soil carbon and nitrogen storage in Xiaoxing ’an Mountains wetland has important influence and significance for understanding regional carbon and nitrogen cycle and climate change.In this paper,the soil samples of the surface layer(0-10cm),middle layer(10-20cm)and bottom layer(20-30cm)were obtained by field sampling of Wuyiling swamp wetland and Dazhanhe swamp wetland in the typical permafrost area of Xiaoxing ’an Mountains,and the soil bulk density and water content were calculated.Explore the correlation between environmental factors and soil organic carbon and total nitrogen at different depths,including meteorological factors(temperature,precipitation),vegetation factors(NDVI,EVI,infrared band,near-infrared band,short-wave infrared band)and topographic factors(DEM,aspect,slope),and the correlation between carbon and nitrogen.The random forest model was used to predict soil organic carbon and total nitrogen,and the reserves of soil organic carbon and total nitrogen were estimated.The conclusions are as follows :1、The soil organic carbon content decreased with the increase of soil depth,and the soil total nitrogen in the middle layer was greater than that in the surface layer and the bottom layer.The content of soil organic carbon and total nitrogen decreased with the increase of soil depth.The carbon and nitrogen content of different wetland types is generally higher than that of forest and shrub.2、Among the environmental factors,NDVI,B5 and water content were significantly positively correlated with soil carbon and nitrogen,while temperature and bulk density were significantly negatively correlated with soil carbon and nitrogen,and other environmental factors were weakly correlated or irrelevant.3、In the three machine learning model methods of random forest,support vector machine and decision tree,the random forest model has the highest training accuracy in each soil depth.4、The random forest model predicts that the estimated soil organic carbon storage in the Wuyiling swamp wetland is 517594 t,the estimated soil total nitrogen storage is21629.5 t,and the estimated soil organic carbon and total nitrogen storage in the Dazhanhe local swamp wetland are 151136 t and 11137.4 t,respectively.