A Study On Distribution Area Prediction Of The Main Types Of Vegetation In Yunnan

In combination with vegetation classification from remote sensing, we divided vegetation of Yunnan Province into27types and compiled the current map of vegetation. The data source was remote-sensing imageries of ALOS from2008to2011. On this basis, we analyzed the climatic and spatial distribution characteristics of main vegetation types by six climate factors. Choosing four relatively higher climate factors as predictor variables and vegetation data obtained by remote sensing classification as response variable, we used the classification-tree model to construct the predictive distribution model of vegetation to predict the potential distribution of main vegetation in Yunnan Province. We also systematically analyzed the main climate factors impacting on vegetation distribution of research area, which can provide theoretical basis when estimating the potential influence of climate change on distribution of vegetation in research area and even the whole southwest region of China. The result shows:(1) We choose five pairs climate indicators which are the average temperature (TMA) and annual precipitation(PRA), summer temperature (TMS) and summer precipitation (PRS), winter temperature (TMW) and winter precipitation (PRW), winter temperature and summer precipitation, summer precipitation and winter precipitation. We use those to construct the spatial distribution of seven main kinds of vegetation. By analyzing, various types of vegetation form are relatively concentrated in distribution of climate space and it shows that the acquired results of the vegetation classification are accurate through interpreting the image interpretation.(2) By comparing of each factor’s DWS,we can conclude that winter temperature significant influence in vegetation distribution and the contribution rate of the model average over thirty percent. The contribution rate of winter temperature for temperate-cool coniferous forests, cold-temperate coniferous forests exceed eighty percent. The second is the summer precipitation and winter precipitation, the two precipitation index have certain influence on the distribution of various vegetation types, and it shows that the water condition also is one of important impacting factor of the distribution of vegetation. The contribution rate is minimal in summer temperature, but it plays a decisive role in the distribution of warm-temperate coniferous forests.(3) The classification and regression trees (CART) predict show that the biological climate factors have different influence on distribution of vegetation. The winter temperature has the deceive effect on semi-humid evergreen broad-leaved forests, mountainous humid evergreen broad-leaved forests, temperate-cool coniferous forests and cold-temperate coniferous forests; The distribution of monsoon evergreen broad-leaved forests and warm-hot coniferous forests codetermined by summer precipitation, winter precipitation and winter temperature, and winter temperature and winter precipitation have common impact on the distribution of semi-humid evergreen broad-leaved forests and temperate-cool coniferous forests; summer temperature has great influence on distribution of warm-temperate coniferous forests; winter temperature and winter precipitation determine the distribution of the mountainous humid evergreen broad-leaved forests, as well as the distribution of main influence factors of cold-temperate coniferous forests, but cold-temperate coniferous forests have certain requirements for summer temperature.