Multi-scale assessment of river-floodplains

河流生态系统是高度动态和异质性的生态系统，由于自然和人为的力量。沿海河岸森林的异质性使其成为温带森林中生物多样性最丰富的森林之一，但在广阔的地区绘制地图也具有挑战性。沿海洪泛区森林内的各种生态重要结构导致不同尺度的空间格局。其中许多模式包括结构、空间和光谱特性，这些特性有助于利用高空间分辨率卫星图像进行区分。精细比例尺图像（亚米分辨率）用于河岸测绘的潜力特别引人注目，因为使用较粗糙的图像辨别这些地区一直是一项挑战。这项拟议的研究计划的目标是严格量化的层次模式的自然和人为变化的河流洪泛区使用一套新的遥感工具和地面实况信息。我们强调使用，扩展和测试在我们的试点项目中开发的技术，以在Clayoquot Sound内新的更大的洪泛区，重点是稀有的保护栖息地以及自然和人为干扰的模式。我们将分类高空间分辨率图像沿着三条河流在沿海不列颠哥伦比亚省使用国家的最先进的基于对象的分类软件，以及使用粗尺度图像和景观格局指数评估大尺度的变化。使用这种工具的组合，以了解河流洪泛平原的变化，在精细和广泛的空间尺度，将有助于解开河流洪泛平原生态系统的变化的人为和自然因素。作为我们长期目标的一部分，我们还考察了海达加勒和沿着斯基纳河的其他地点，其中包括原始地区以及经过历史森林管理改造的地区。这些河流系统也将在历史框架内进行研究，通过使用历史重建和航空照片量化过去的变化。这些不同的方法将为我们了解BC省一些生物多样性最丰富、管理最严格的河流洪泛区系统增加一个关键维度。

Riverine ecosystems are highly dynamic and heterogeneous ecosystems, due to both natural and anthropogenic forces. The heterogeneity of coastal riparian forests makes them among the most biodiverse of temperate forests, but also challenging to map over broad areas. A variety of ecologically important structures within coastal floodplain forests result in spatial patterns at different scales. Many of these patterns include textural, spatial and spectral properties which aid in their discrimination using high spatial resolution satellite imagery. The potential of fine-scale imagery (with sub-metre resolution) for use in riparian mapping is particularly compelling, as such areas have been challenging to discern using coarser imagery. The goal of this proposed research program is to rigorously quantify hierarchical patterns of natural and anthropogenic variability in river-floodplains using a suite of new remote sensing tools and ground-truth information. We emphasize the use, extension, and testing of techniques developed in our pilot site to new larger floodplains within Clayoquot Sound, with an emphasis on rare habitats of conservation concern and patterns of natural and anthropogenic disturbance. We will classify high spatial resolution imagery along three rivers in coastal British Columbia using state-of-the-art object-based classification software, as well as assess broad-scale variability using coarser-scale imagery and landscape pattern indices. Using this combination of tools to understand the variability within river-floodplains, at both fine and broad spatial scales, will help disentangle the anthropogenic and natural agents of change in river-floodplain ecosystems. As part of our long-term objectives, we also examine additional sites in Haida Gwaii and along the Skeena River, which include pristine areas as well as those modified by historic forest management. These same river systems will also be examined within an historical framework, by quantifying past variability using historic reconstruction and aerial photos. These different approaches will add a critical dimension to our understanding of some of the most biodiverse and heavily managed river-floodplain systems in BC.