2001 Volume 65 Issue 5 Pages 563-568
this study's aim is the prediction of potential natural vegetation (PNV). We built date base of the natural vegetation with remains from 862 stand descriptions that was investigated whole area of Kanagawa from 1971 to 2000, and We created environmental data of climate (warm index, cold index, mean tempurature, min-temprature, mean precipitation), topography (elevation, slope, aspect of slope, unevenness index, distance from sea), soil, geology, by using GIS. All natural vegetation data was classified 14 types by using TWINSPAN and the mathed of Phytosociology. The environment value of the feild's sampling points was extracted by using GIS, environment factor of each vegetation types was grasped quantitatively by using logistic regression analysis. PNV of Investigation area were estimated by the warm index, cold index, mean precipitation, elevation, slope, aspect of slope, unevenness index, distance from sea, 2 soil types and 4 geology types. The proprtion of right answers of the made model shows 98%. The result of the simulation of PNV by using the model correspond to potential natural vegatatin maps based on phytosociology. The environment which each natural vegetation existed in became obvious by this research. The mapping of PNV by the objective method became possible.