The estimating of the spatial distribution of forest biomass in China based on remote sensing and downscaling techniques

Abstract

Terrestrial ecosystems play an important role in the global carbon cycle.The implementation of the United Nations Framework Convention on Climate Change(UNFCCC) and the Kyoto Protocol have made the study of terrestrial ecosystem carbon cycling a hot topic of scientific research globally.Forest biomass is an important carbon pool for terrestrial ecosystems,so its magnitude and spatial patterns are critical in determining the carbon-exchange potential of forests.Scholars in the field have used a variety of different methods to study many aspects of China′s forest biomass.Despite progress,a high degree of spatial biomass heterogeneity has caused the current findings to vary widely,especially when the regional-scale is used in measuring the spatial distribution pattern of biomass,leaving significant uncertainties.Three traditional methods have been used in biomass research: physical plots measured on the ground,modeling and simulation,and remote sensing.Currently,statistical downscaling is a widely used statistical method which transforms large-scale,low-resolution information into regional-scale,high-resolution information.This method has recently been used effectively and achieved good results in the field of ecosystem carbon cycling.Combining remote sensing data with ground-based observations is a key step in the quantitative research of forest biomass spatial distribution patterns.In particular,national forest resource inventory data can be used to combine the advantages of remote sensing data,and its spatial characteristics,with the reliability of detailed information from the ground to produce reliable statistical information reflecting the surface characteristics.This paper is based on the sixth China forest inventory dataset,a vegetation map of the People′s Republic of China(1∶1000000),and the spatially explicit Net Primary Production(NPP) datasets derived from the Moderate-resolution Imaging Spectroradiometer(MODIS) Gross Primary Production(GPP)/NPP products.We quantitatively estimated the spatial distribution of forest biomass(1km resolution) using the spatial downscaling technique.The results provide four finding.(1) The downscaling technique can effectively combine the advantages of both remote sensing and forest inventory data and will be useful in mapping forest biomass at the regional scale.In this study,the average errors in the calculated total biomass and average biomass are 1.4% and 1.6%,respectively,which is comparable to other studies on a national scale.In this study,average error is 6% for the estimated biomass density on the provincial scale.In addition,the total biomass error is 37% for Yunnan Province,while other provincial scales averaged an error level of 10%.(2) China′s biomass in young forests,middle-aged forests,nearly mature forests,mature forests and over mature forests show an increasing trend in biomass,and the overall trend appears reasonable.Young to mature forest stages,which are gradually increasing in age,have shown a large increase in biomass.Mature forests to old growth forests have experienced a reduced rate of increase in some areas,with old growth forest biomass even decreasing.(3) Forest biomass in China has obvious spatial distribution patterns,with widely distributed forests in eastern China and a scattered distribution of forests in western China.China can be divided into five main regional forest divisions based on biomass density,listed here in descending order: the high mountains of Xinjiang Uyghur Autonomous Region,northeastern Inner Mongolia,the Hainan Island tropical region,southwestern China,and southern China.(4) The total stock of forest biomass in China is 11.0 Pg with an average biomass of 74.8Mg/hm2.China′s biomass is primarily found in the Da Xing′an(Greater Khinghan),Xiao Xing′an(Lesser Khinghan) and Changbai mountains of the northeast,Xinjiang Mountain and the Hengduan Mountains of the southwest and in the Wuyi Mountains of the southeast.