Silvilaser 2019 - Poster Presentations »
Forest growth monitoring using bitemporal areal LIDAR data: case study of national park Czech Switzerland
Growth from remote sensing can be assessed based on repeated measurements (separated by an interval of time) of the same tree or the same stand. The temporal differences in the remotely sensed data usually directly or in some other manner indicate the growth. Such studies make sense particularly in protected areas, i.e. areas where there is no human intervention or management and temporal changes are driven only by natural development of the forest. Growth dynamics assessment from remotely sensed data has been addressed by many authors (Maltamo et al. 2007; Næsset and Gobakken 2005; Tompalski et al. 2016; Yu et al. 2008), who evaluated the use of remote sensing data to estimate changes in stand characteristics such as: height, diameter, basal area, but also the volume or site index. The aim of the research is to evaluate the growth dynamics from bitemporal three-dimensional data obtained with repeated laser scanning and photogrammetric methods from aerial images. In the research are used historical data from the National Park Czech Switzerland with area approximately 80 km2, which was acquired in April 2005 by the Technical University of Dresden, Germany. The data were acquired using aerial laser system Falcon II. The average flight altitude was 1200 m above ground with 50% overlap. The final data represent the first and the last reflection of each pulse. The average density of the last reflection points was 8.5 points / m2 (Trommler & Csaplovics, 2006). The repeated data collection was carried in April 2019. LiDAR Leica ALS70-CM and multispectral camera Leica RCD30 Series was used in our data acquisitions. Flying height was from 579 to 1069 m above ground level (AGL) and laser pulse rate was used 231 kHz. The point density of the laser scan was 5 points / m2, FOV angle is no more than 50 degrees. The flights were made with longitudinal overlap of at least 80%, transverse overlap of at least 50% and average image resolution of 12.5 cm. As large-scale verification data covering the whole area, we used forest management plan, which includes the repeated stand inventory. Moreover, detailed ground verification data are collected in the field, including the cores for diameter growth estimation. This study presents the methodology for growth dynamics assessment and describes the advantage and disadvantage of each type of the data. Resulting increment maps and overall summary estimation for dominant species are presented.