Liu Duosen, Cai Weiqi
In order to identify hydrothermal Condition for distribution of vegetation,nine indices were taken from 586 meteorological stations in China,i.e.temperature effect of Chemical reactions in nature (T),mean air temperature in thecoldest tenday period (t1),number of days with daily mean air temperature≥10℃(Dt),moisture effect (M),effect of saturated water vapour pressure (V),relative humidity (f),numbcr of days with daily precipitation≥10mm (Dp),altitude(A),and latitude (L).The 586×9 original data mentioned above were treated by the method of principal component analysis.No correlation is found among different principalComponents,each component reflecting a different and independent factor.Fromeigenvectors of the first two principal components,it will be seen that the firstprincipal component reflects thermal intensity in a region,and the second reflects moisture intensity in a region.Cumulative variance of first and second principal components to 84.6%.From principal component analysis,we have obtained sets of points of firstand second principal components for 13 vegetation types of China.It is betterto use these polntsets to identify hydrothcrmal condition of different typesof vegetation in China.These vegetation types are:1.savanna,2.tropical bush,3.warm evergreen and deciduous bush,4.warm forest,5.warm deciduousbush,6.temperate dry deciduous bush,7.temperatemoist deciduous bush,8.temperate forest,9.desert,10.desert steppe,11.steppe,12.alpine steppe,and13.alpine meadow.