In this study, using 3D scanning Light Detection and Ranging (LiDAR) systems, which are the non-destructive methods, to investigate and monitor individual trees in artificial forest sample areas. Using ground-based LiDAR and traditional field survey data to obtain individual tree characteristic values which included tree height, diameter at breast height, canopy height, tree location map and evaluated the feasibility of ground-based LiDAR data in artificial forest areas. Results showed that the accurate information from the ground-based LiDAR system in permanent sampling plots, and different scanning stations could decrease occlusions. By using high density point cloud data, the height and diameter at different height of the tree could also be measured directly by the LiDAR systems. Different number of point cloud affected the accuracy of diameter. Results also showed that the systems could establish tree volume model non-destructively with tree height and DBH data obtained from ground-based LiDAR. Using threshold value to classify gap fraction, results showed that were accurate. Therefore, the results could be used to analyze the horizontal and vertical canopy structure, it was beneficial for the monitoring of plot dynamics in forest areas. Ground-based LiDAR could be used to determine tree measurements, tree competition and canopy gap which solve the problems of quantification and spatial research and also help with the stimulation of thinning operations. Results of this study were beneficial to artificial forest management and improved the forest resource inventory methods.

The forest canopy affects the laser penetration rate of airborne LiDAR system (ALS). The denser the forest canopy, the higher the laser penetration rate, and vice versa. For an ALS survey project, with the goal to acquire sufficient ground surface points, the flight planning should be conducted according to the distribution forest canopy thickness. The information of forest canopy can be easily accessed via the satellite-derived vegetation index. In this study, we examined the correction relationship between the satellite-derived vegetation index and laser penetration index (LPI). The study area was 21 km2 located in Tseng-Wen Reservoir and 36 km2 located in Jin-Shan volcanic region. The ALS data were collected with an Optech HD400 instrument. The LPI is the ratio between the total number of laser pulses reaching at the ground surface and the total number of transmitted laser pulses. The vegetation indices were calculated from Formsat-2, SPOT-5, WorldView2 and GeoEye, all of which were acquired near the time of ALS data acquisition. This study calculated four vegetation indices, i.e., NDVI, RVI, PVI, and SAVI. The effect of atmospheric correction (conducted by ATCOR-3) were also discussed. The results show high linear correlation between LPI and LAI.

Extreme weather events have induced more frequent geological hazards in Taiwan. The heavy rainfall brought by the Typhoon Morakot has triggered a large amount of landslides including the Xiaolin village which was demolished by a catastrophic Landslides. The study of such catastrophic landslides is urgently needed. It is considered that the creep phase of a landslides is a preparatory stage of progressive failure and gives enough signals before turning into a catastrophic landslides. This study used high-resolution airborne LiDAR-derived DEM data from the Central Geological Survey, MOEA. We used different geomorphometric analyses to process the high resolution and high accuracy DEM data including the hillshade, aspect, slope, eigenvalue ratio & openness. Among these geomorphometric analyses, which to converged openness, slope and hillshade method of the results, the capacity to discuss the possible cause and the influence analysis of the catastrophic Landslides in view of the slight terrain features has increased drastically through DEM processing. Our results indicate that over hundreds potential catastrophic landslides may present in southern Taiwan after the Typhoon Morakot event. The quantitative methods used in this study highlight the terrain features of the creep phase of catastrophic landslides and is helpful for landslide feature interpretation and hazard assessment.

Building boundary is one of the important components for the mapping of 2D digital topographic maps and the modeling of 3D city buildings. Photogrammetry is currently the common technique applied for building boundary generation,which are labor intensive. Airborne LiDAR data provides abundant3Dinformation of the scanned objects.The characteristics of objects are implicitly contained in the data set. Usually top surfaces, such as roofs, may have densely distributed points, but vertical surfaces, such as walls, usually have sparsely distributed points or even no points. Building boundaries, referring to the intersections of roof and wall planes are, therefore, not clearly defined in point clouds.To overcome this problem, this paper develops an algorithm to acquire building boundary from airborne LiDAR data. Threemajor process steps are included in the algorithm. Firstly the point clouds are classified as building points and non-building points. Then, octree-based split-and-merge segmentation is implemented to extract plane features.Second, those building points and coplanar points are used to trace the boundary points by concave-hull algorithm. Boundary points of coplanar point group and building points and the first and intermediate echo points of multi-return scan are selected as candidates of building boundary points. Finally, methods of the Hough transform, line fitting and line segmentation are applied to find line segments belonging to building boundaries. The experiment results show the effectiveness of the proposed method for automatic building boundary extraction from airborne LiDAR data, and that combining the information of the first and intermediate echo points of multi-return and the boundary points increases the completeness of boundaries. And, it is promising to use the extracted boundaries for 3D building modeling in the future.

Guishan Island is located at north eastern Taiwan belongs to Yilan County. According to geophysical and geochemical studies, it is an active volcano, and the latest eruption occurred in the Holocene (7 ka). The active hydrothermal vents and earthquake can be considered as manifestations of volcanic activity. There are some hydrothermal vents at eastern offshore, and the fluids from the vents are mostly made up of sulfur, which cause the discoloration of the ocean around Guishan Island. The discolored area is called plume which can be easily detected by satellite image. The shape and size of plume are disturbed by ocean currents or winds. In this study, we detect the plume area by three steps: (1) remove land area; (2) remove cloud area; (3) classification. First, ENVI is used to build a mask and remove land area. Followed by a fully constrained least squares approach, it is a mixed pixel classification method for concentration estimation in each pixel. Finally, the area of plume is calculated from the result of least square approach, and then compare with the number of earthquake events to find the relationship between them. In our experiment, 36 SPOT images in 2008 are adopted, and the result shows that the larger area of plume was occurred around the large cluster of events. The area of plume is related to earthquake but there isn't clear. After the correlation coefficient is calculated, the highest is in ±1day. It means maybe the earthquakes can cause the area of plume become larger.