Processing terrestrial lidar data to produce vertical profiles of fire fuel for use in the Canadian Forest Fire Danger Rating System

野火在加拿大是一种重要的灾害，每年都会威胁到资产和生计，预计这种情况会随着气候变化的影响而恶化。在过去的30年里，加拿大省级和地区消防管理机构一直在使用加拿大森林火灾危险等级系统（CFFDRS），该系统由加拿大森林服务局（CFS）开发。各机构每天使用该系统作为其主要决策工具，以支持火灾危险和行为评估。随着科学技术和消防管理环境的发展，消防中心的应急管理战略（EMS）计划优先发展下一代CFFDRS （NG-CFFDRS）。NG-CFFDRS的火灾行为模型将使用燃料结构属性的新输入，这些属性很难在现场直接测量，也不能被传统的森林清单很好地捕获。为了取代耗时耗力的直接测量和破坏性采样，地面激光扫描仪（TLS）已经被确定为提供真实的三维树和林分信息，用于结构表征和一些新燃料结构属性的间接、非破坏性估计。然而，由此产生的共同注册TLS点云的处理仍然很复杂，目前的软件工具通常需要对底层算法有深入的了解，这阻碍了这些工具向林业和火灾管理领域的非专业用户传播。因此，该项目旨在开发新工具，在NG-CFFDRS的操作实施背景下使用TLS来估计燃料结构属性。明确地说，我们希望开发一个TLS数据处理管道，可以从激光雷达点云中提取燃料结构属性的估计。

Wildfire is a hazard of importance in Canada, threatening assets and livelihood on a yearly basis, a situation that is expected to worsen with the impact of climate change. For the last 30 years, Canadian provincial and territorial fire management agencies have been using the Canadian Forest Fire Danger Rating System (CFFDRS), developed by the Canadian Forest Service (CFS). Agencies use this system daily as their main decision-making tool to support fire danger and behaviour assessments. With the evolution of science, technologies, and of the fire management environment, the Emergency Management Strategy (EMS) Program of CFS has prioritized the development of the Next-Generation of the CFFDRS (NG-CFFDRS). The fire behaviour models of the NG-CFFDRS will use new inputs of fuel structural attributes that are hardly directly measurable in the field and that are not well captured by classical forest inventories. To replace time and resource consuming direct measurements and destructive sampling, Terrestrial Laser Scanner (TLS) has been identified to provide realistic 3D tree and stand information for structural characterization and indirect, non-destructive estimation of some of the new fuel structural attributes. However, the processing of the resulting co-registered TLS point clouds remains complex and current software tools often require a deep understanding of the underlying algorithms, preventing the dissemination of these tools to non-expert users in forestry and fire management. This project thus aims at developing new tools to use of TLS to estimate fuel structural attributes in the context of the operational implementation of the NG-CFFDRS. Explicitly we want to develop a TLS data processing pipeline that can extract estimates of fuel structural attributes from lidar point clouds.