| The forest biomass is one of important factors to monitor the global change,but the remote sensing technique can fast,and not have the destruction to estimate the forest biomass.The Westem Sichuan Subalpine forest is a present relative preserved good natural forest in the west of our country,and is also the most sensitive areas of global climate change.In this paper,taking Daofu County's Subalpine forest as an example,based on sampling plots(broad leaf forest is 44 and coniferous forest is 68) of forest biomass determination and the Landsat TM data analysis remote sensing information and the actual forest biomass correlation dependence,has established the best forest biomass remote sensing model of the Western Sichuan Subalpine coniferous forest and broad leaf forest based on the remote sensing information,has estimated the forest biomass and the carbon storage of Daofu,and has carded on the spatial analysis to the forest carbon storage.(1) The remote sensing data of topographic correction was obviously higher than uncorrected image,the results showed that the remote sensing data of topographic correction was more suitable to use in establishing the forest biomass remote sensing model.(2) Based on the relevant analysis of the forest biomass and the remote sensing single wave data(TM1～7),vegetation index(DVI,RVI and NDVI and so on) and the topographical data(DEM,SLOPE and ASPECT) indicated that:①Besides TM6,other wave bands were correlated with the biomass at 0.01 confident level,the biomass was most obviously correlated with TM1,the correlation coefficients were respectively 0.620 (coniferous forest) and 0.514(broad leaf forest);②The coniferous forest and the broad leaf forest biomass were correlated with vegetation index PVI and BVI at 0.01 confident level. The coniferous forest biomass was correlated with vegetation index NDVI at 0.05 confident level;③The coniferous forest biomass were correlated with atitude and slope at 0.01 confident level,but the broad leaf forest biomass was only correlated with aspect at 0.01 confident level.The results showed that variable's relevant difference between the coniferous forest and the broad leaf forest biomass with the remote sensing information was more obviously,the overall performance was the broad leaf forest correlated with the remote sensing higher than the coniferous forest biomass.This means that the forest type has the tremendous influence to the remote sensing biomass model precision.(3) Based on sampling plots of forest biomass determination and single-band T M,18 variables were selected to build the estimation regression model to predict th e biomass of the subalpine coniferous forest and broad leaf forest in westem Sichua n province separately.In the establishment of the monadic linear regression,the non -linear regression and the multivariate linear regression model of forest biomass,the results showed that multiple linear regression model was more accurate.The most superior regression model of coniferous forest biomass is Y=2.511TM1-1.366TM3+0. 057DEM-1.021SLOPE-239.107(R=0.700),and the most superior model of broad leaf f orest biomass is Y=19.103+0.027ASP-0.863TM5+1.093TM7+0.485BVI(R=0.592).(4) Based on the establishment coniferous forest and the broad leaf forest most superior model studied the region forest total biomass was 56.243 Tg,the forest total carbon storage was 26.434 TgC,the average carbon density was 87.153 t·hm-2.And the coniferous forest biomass was 45.842 Tg,accounted for 81.507%of total forest biomass in the study area,the carbon storage was 21.546 TgC,accounted for 88.180%of total forest carbon storage in the study area,the carbon density was 98.564 t·hm-2;The broad leaf forest biomass was 10.401 Tg,accounted for 18.493%of total forest biomass in the study area,the carbon storage was 4.888 TgC,accounted for 11.820%of total forest carbon storage in the study area,the carbon density was 75.724 t·hm-2,the result indicated that the coniferous forest was the biggest forest plant carbon pool of this region.(5) Studied forest plant carbon storage of the region for as the center decrease s progressively along with the elevation change performance took elevation3800-420 0m to the two-pole,the elevation 3800-4800m forest carbon storage was the biggest, accounts for 39.328%of total forest carbon storage in the study area;The forest ca rbon storage along with slope change performance for steep slope＞anxious slope＞pit ch slop＞slow slope＞dangerous slope＞even slope;The forest cover plant carbon stora ge along with the slope to the change performance is half shady side＞half sunny sl ope＞sunny slope＞shady side＞no slope to approach.
|
PDF Full Text Request