| Summary: | 碩士 === 國立屏東科技大學 === 森林系 === 93 === Large-scale forest inventory maps typically show the outline of forest stands and provide information on the tree species, tree height, density, crown closure, age, health, and other characteristics for each stand. The need for such forest information has increased during recent years. Users now demand up-to-date digital information on such things as wood volume, habitat type, or potential biodiversity. It is, however, recognized that large scale official forest inventory maps are often outdated and vary in quality, sometimes showing high error levels. New methods used to create forest information that like forest succession, ecological function which is from high resolution satellite image and digital image processing techniques has shown great capacity in reconnaissance surveys. Nanjenshan forest ecosystem is a hot spot of tropical forest that is a site of long team logical research in Taiwan, it is an important work to simulate the structure of forest canopy. In this study, we use the high resolution IKONOS satellite image as material and use the semi-variograms statistic methods to estimate the spatial variation and the effect mechanism of canopy structure in Nanjenshan secondary forest ecosystem.
In this study a total of 500 10m × 10m plots allocated in the Nanjenshan secondary forest stands. The ecological values were analysis from the survey data. Then we used the Kriging semi-variogram model to estimate the distribution of stand ecological values of the Nanjenshan secondary forest stands, to build the whole stand ecological values model. We found the exponential model is the best in this study. The exponential model can present the distribution of stand ecological values and offered the basic data of the Nanjenshan secondary forest stands structure. We also use the IKONOS satellite image include the red and green bands to estimate the relationship amount 3 parameters and stand ecological values by semi-variogram analysis. The results showed that the relationship is significantly between red band sill index and stand density. The relationship is also significantly between green band nugget index and stand height, diameter at breast height and basal area per hectare. The sill index showed the best result, so it is the best to model the stand structure for the semi-variogram model of green band. For finding out the effects of northeast monsoon in the secondary forest stand structure of spatial variability, we divided the secondary forest into windward forest type and leeward forest type. We used t-test to check up if there was difference between the stand ecological values and 2 parameters (sill and nugget) of green band, and the result showed that the difference was significantly in all stand ecological values beside the volume of wood.
|
|---|
