-

One of the principels of forest management is recognization of forest parameters such as forest classification. This reaserch was carried out for ground truth mapping of forest types in order to study possibility of forest classification using satellite data (ETM+). Until now, forest typing has only been done based on species frequency. As thick trees have high dominance and share in forest canopies, 100 thick trees in each plot were selected for forest typing in this research,. The study area was 2170 ha in the Patom, Namkhaneh and Gorazbon districts, Kheyroudkenar watershed. In sampling method, 193 plots of one hectare area were selected. Each plot was sub divided into four sub-plots and data were collected individually. Percent of species frequency were computed for each sub-plot and plot and were then classified. Concerning to aware of low stratifiction of special types through satellite data, two typing methods were used, i.e. general and special types. In the general typing, three types were considered: pure, dominant species and mixture types. In the special types, those types were componened with two dominant species, were subdivided to details
types. Six general and fifteen special types were finally recognized. Comparison of general types of total trees frequency method and 100 thick trees frequency showed that they were similar about 71 percent. Also, the best sample size for sample forest typing was found 0.5 ha for such forests