高分一号遥感影像下崂山林场林分生物量反演估算研究_曾晶.pdfVIP

36 1 Vol. 36 No. 1 2016 1 Journal of Central South University of Forestry Technology Jan. 2016 Doi:10.14067/ki.1673-923x.2016.01.008 http: // 100083 19 80.75% 2013 402 485.44 t S771.8S718.55+6 A 1673-923X(2016)01-0046-06 Laoshan forest biomass estimation based on GF-1 images with inversion algorithm ZENG Jing, ZHANG Xiao-li (Key Lab. for Silviculture and Conservation Co-constructed by China Ministry of Education and Beijing, Beijing Forestry University, Beijing 100083, China) Abstract: Based on GF-1 images as data sources, combined with fi eld actual measured data, a total of 19 factors, including vegetation index factors, band combination factors and topographical factors and other factors, were extracted, then Laoshan forest biomass multivariate linear inversion model were established, and further the analytical accuracy, the advantages and disadvantages of t he inversion model were analyzed, and the biomass of Laoshan Forest Farm was estimated. The results show that the biomass model set by multiple linear regression had average retrieval accuracy of 80.75%. Based on the biomass estimation model by inversion, the Laoshan forest biomass in 2013 was 402 485.44 tons. Key words: forest biomass; remote sensing image by GF-1; inversion algorithm; inversion model; multiple linear regression [3] 90%[1] [2]