Detecting Fluvial Wood in Forested Watersheds using LiDAR Data: A Methodological Assessment
Detecting Fluvial Wood in Forested Watersheds using LiDAR Data: A Methodological Assessment
复制标题
使用激光雷达数据检测森林流域中的河流木材：方法评估
DOI：
--
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
J. Dietrich
中科院分区：
文献类型：
--
作者：
J. Atha;J. Dietrich
人们早就知道，河流中的大片树木增加了河道的复杂性，是太平洋西北部森林覆盖的山脉溪流地貌变化的主要驱动力。分析河流相木材的存在和分布的研究往往局限于场地或范围尺度的空间范围，因为全面的木材调查需要密集的实地工作。遥感技术开始使研究人员能够评估河流相木材在盆地或区域尺度上的动态和分布。我们使用了2009年的高分辨率光探测和测距(LiDAR)点云数据来检测和量化俄勒冈海岸范围内五个森林流域内的木材。我们对LiDAR数据进行了过滤，去除了溪流通道上的森林树冠，并根据河流中河流木材的独特形状对其进行了目视清点。我们得出了几个木材和溪流形态测量变量，以检验与溪流下游木材丰度和位置相关的理论。我们能够有把握地探测到河流相木材；然而，由于木材随时间的动态和移动特性，在研究中很难用地面真实数据来验证结果。我们在五个研究流域内绘制了总共163个单一原木和55个堵塞的地图。我们没有发现与坡度有关的单个木片和堵塞位置之间的统计上的显著差异；然而，所调查的木材通常在河流功率较低的地区被发现。这项研究表明，在森林茂密的流域，利用高度滤波的激光雷达探测流入流中的木材是可能的，并具有在未来广泛的空间尺度上进行木材研究的潜力。版权所有©2015 John Wiley&Sons，Ltd.
It has long been known that large wood in rivers increases channel complexity and is a primary driver of geomorphic change in forested mountain streams in the Pacific Northwest. Studies analyzing the presence and distribution of fluvial wood are often limited in their spatial extents to the site or reach scales because of the intensive fieldwork required for comprehensive wood surveys. Remote sensing techniques are beginning to allow researchers to assess fluvial wood dynamics and distributions on a basin or regional scale. We used 2009 high‐resolution light detection and ranging (LiDAR) point cloud data to detect and quantify wood within five forested watersheds in the Oregon Coast Range. We filtered the LiDAR data to remove the forest canopy over the stream channel and visually inventoried fluvial wood based on its distinct shape within the channels. We derived several wood and stream morphometric variables to test theories relating to wood abundance and positioning in the lower reaches of streams. We were able to detect fluvial wood with confidence; however, validation of results with ground‐truth data was difficult in the study due to the dynamic and mobile nature of wood through time. We mapped a total of 163 single logs and 55 logjams within the five study watersheds. We did not find statistically significant differences between individual pieces and jam positioning in relation to slope; however, the surveyed wood was often found in areas of lower stream power. This research shows that it is possible to use height‐filtered LiDAR to detect in‐stream wood in densely forested watersheds and has the potential to be employed in future wood studies across broad spatial scales. Copyright © 2015 John Wiley & Sons, Ltd.