Study On Regional Forest Cover Change And Aboveground Biomass Estimation

The Three-North Shelterbelt Program(TNSP)zone is the largest ecological restoration project in the world.Yulin in Shaanxi province,located at the border between the region of the loess plateau and the Inner Mongolia plateau,is the ecological fragile area in TNSP zone.The study of the dynamic variation of forest cover and forest biomass estimation in Yulin is of great significance to environment monitoring and ecological evaluation.Landsat data has long-term data accumulation and high spatial resolution.This paper mainly uses time series Landsat data to quantitatively study the land use change,forest cover change,afforestation and deforestation time retrieval in Yulin six counties,and combines the age of afforestation to estimate the forest aboveground biomass in the study area,which can evaluate ecological construction achievements in Yulin since 1974.In this paper,the land use/cover change,single tree biomass allometric growth equations,inversion of forest cover change time,and regional forest biomass estimation in Yulin were investigated,and the following conclusions were obtained:(1)In this paper,long time series Landsat data were quantified by atmospheric correction,relative radiation normalization,terrain radiation correction,and cloud removal to obtain the time series surface reflectance data of the study area.Then,we used the CART classification method to classified land use/cover types of study area in 1987,1993,2000,2007,and 2014.The land use/cover type of the study area was divided into crop land,forest,grassland,water,urban construction land,sandy land,and bare land..In the CART classification method,EVI was introduced as a feature band to increase the accuracy of feature classification.The overall classification accuracy reached 88.91%,Kappa coefficient of 0.87,has higher classification accuracy.(2)According to the dominant tree species(Chinese scholar tree,poplar,Chinese pine) in the study area,different types of allometric growth equations were constructed using measured data in 2012 such as DBH,tree height,and tree age.The results showed that the effect of the allometric growth equations with the two parameters combining the DBH and tree height was the bes.The allometric growth equation based on the power function of DBH,which fitting effect was good and only one parameter of breast diameter is needed,compared to the allometric growth requiring breast diameter and tree height,with better applicability.The allometric growth equations of Chinese scholar tree,poplar,Chinese pine based on DBH were established to estimate the above-ground biomass of all single trees in each plot and obtain the total forest biomass in the plot.(3)Using Landsat time series surface reflectance data from 1972 to 2012 of the study area,Integrated Forest Z-Score(IFZ)was constructed,and a vegetation change tracker(VCT)algorithm was designed and established to invert forest cover changes and estimate forest afforestation and deforestation time.After the actual test area test,the overall accuracy of the VCT algorithm classification was 90.95%,the Kappa coefficient was 0.88,and the forest cover change time error(less than or equal to 3 years)reached 83.72%,which has a good inversion effect.The research area presented two afforestation peaks in1977 and 2003,reaching 279.52 km~2 and 1249.26 km~2,respectively.The deforestation area reached two peaks of deforestation in 1987 and 2007,which were 11.27 km~2 and10.61 km~2,respectively.The results show that using the time-series IFZ data to construct the VCT algorithm and to invert the forest cover change time can effectively invert regional forest afforestation and deforestation time.(4)In this paper,we established a regional forest aboveground biomass estimation model based on the afforestation age in the pixel scale.A regional forest biomass estimation model was constructed using NDVI*age as a variable and the estimated equation was y=150.6x+332.31,the coefficient of determination R~2 reached 0.62.In 2012,the forest biomass of study area was estimated.The total forest aboveground biomass of Shenmu County was the highest,reaching 4.76×109kg.The area with the lowest total forest aboveground biomass is Mizhi County(7.94×10~8kg),followed by Xiangyang District(1.66×10~9kg),Jia County(1.40×10~9kg),Zizhou County(1.35×10~9kg)and Hengshan County(1.16×10~9kg).Combining NDVI and pixel-scale afforestation ages,we effectively estimate regional forest biomass and realize monitoring of forest biomass and distribution in various districts and counties in Yulin District.It is of great significance to explore the carbon sequestration capacity of forests in the Yulin and the ecological restoration effectiveness of the TNSP.