| As the main body of renewable natural resources on the earth,the forest is the largest terrestrial ecosystem,which has the most abundant material resources and the most complicated structure,playing an important role in maintaining ecological processes and ecological balance,and promoting the global environmental protection.The Greater Hinggan Mountains is an important forestry base in northern China,it's also the largest coniferous forest in China.What's more,wildfire has been an influential disturbance and occurs frequently in the Greater Hinggan Mountains with a large burned area,which will pose serious threats to the ecological environment of the entire northern area,therefore it's very necessary to conduct scientific and effective post-fire recovery measures.Traditional forest resource survey methods used to obtain forest change information,forest structure and forest parameters,but the access to information timely and the accuracy of information are not allowed usually.In recent years,with the application of remote sensing technology increasingly developed,there are many successful examples made by using RS for forest types identification,classification,and the inversion of forest vegetation structure,ecological parameters,in particular,there are substantial studies focusing on the large-scale forest biomass estimation and extraction.Moreover the utilization of RS to quantify the forest post-fire recovery's monitoring makes great progress,contributing to the forest regeneration,improving China's forestry ecological environment construction system and digital forestry construction.Hence,it's of great significance to study the forest post-fire recovery and forest environment of burned areas in the Greater Hinggan Mountains,maintaining forest ecological balance,and promoting forest economy and forestry sustainable development.Based on the catastrophic fire in May 1987 in the Greater Hinggan Mountains,the paper using mult-source remote sensing technology to study the natural recovery process of the post-fire forests in burned areas of the Greater Hinggan Mountains,mainly focusing on these aspects,involved the changes in stock volume and the composition of forest tree species.The main contents and results of the study were as follows:(1)Classification mapping of forest tree species based on optical remote sensing.According to the spectral characteristics of different plants in different growth periods,the multi-period remote sensing images in growing season were selected,and some relevant spectral indices such as NDVI and GVI were extracted.Combined with Google earth and 2010 forest inventory data,decision tree classifier was used to classify the present forest tree species in the study area,and the accuracy of the classification results were verified with field measurements.The accuracy of the classification results was86%,the Kappa coefficient was 0.8,indicting the classification accuracy was relatively high.(2)Estimation mapping of forest stock volume via multi-band and multi-polarized radar images.First,established the fitting function between the radar backscattering coefficient and the average volume of forest inventory data after processing.Second,compared the correlation between the radar backscattering coefficients with multi-band and multi-polarized radar images and the average forest stock volume,the results showed that L-band HV polarization method was more suitable for the estimation of forest stock volume,and the estimation map of forest stock volume in the study area was made,and the RMSE of the estimated forest average volume was verified by field data.The results of the estimation of xing'an larch,mongolia and white birch were16.76 m~3/hm~2,16.55 m~3/hm~2 and 15.27 m~3/hm~2 respectively,with an whole average of 16 m~3/hm~2,indicating the estimation the average forest volume was ideal.(3)Analysis of forest recovery in burned areas in Tuqiang Forestry Bureau.The recovery process of the forest in the burned area of the study area was analyzed by means of the method of space instead of time,combining the tree species classification map and the forest volume estimation map of Tuqiang Forestry Bureau.The results showed that there was a certain recovery in the composition of the tree species in the burned area of the study area and the tree species were mainly white birch,so there were still some gaps compared with that of pre-fire.The average forest volume was also at a low level belonged to the early stage of succession.(4)Mapping fire intensity in May 1987 combining post-fire TM images,analyzed recovery factors of forests in burned areas in the study area.Based on the tree species classification map of Tuqiang Forestry Bureau,the proportion of tree species distribution in areas with different fire intensities was studied.The results showed that the distribution ratio of Xing'an larch and Mongolia increased with the rise of fire level,but the recovery of forest tree species did not reach the pre-fire level.Based on the estimation of forest stock volume of Tuqiang Forestry Bureau,the recovery of forest average volume in each fire area was analyzed in detail,indicting there was some recovery in forest stock volume,but it was not completely recovered.The influence of terrain factors on the burned forest recovery showed that the zero slope and the gentle slope were favorable for the recovery of the white birch,while the steep slope,where there were some seed trees so as to protect the provenance,was conducive to the recovery of Xing'an larch and Mongolia.The effect of slope aspect on forest recovery was that the distribution of tree species and the average stock volume of forest in sunny slope were better than those of shady slope.(5)Extraction of the typical burned area and the simulation of the landscape succession of 100a and 200a by using the spatially explicit landscape model(LANDIS),and the results of succession were analyzed.The results showed that the direction of forest succession was the transition from white birch to Mongolia with time,and finally the forest succession was Xing'an larch.And the ratio of tree species was also close to the hypothesis of space replacement in this paper,and the simulation results were better. |
PDF Full Text Request