Remote Sensing-based Forest Fire Occurrence and Management System

在过去的25年里，加拿大平均每年发生7300起火灾，每年大约有250万公顷的森林被毁。仅在过去的十年里，加拿大人每年在灭火上的花费就在8亿到15亿美元之间。虽然火灾是自然循环的一部分，可以使土壤恢复活力，杀死害虫并导致新的生长，但不受控制的火灾也会造成重大的有害影响，例如(i)由于失去成熟树木而造成的经济损失；（ii）增加进入大气的二氧化碳负荷；（iii）火灾期间及火灾后空气质素恶化；（四）人员生命财产损失。因此，开发有效的森林火灾发生和管理系统对于抵消这些不利影响至关重要。这项拟议的研究计划旨在利用各种卫星平台获得的遥感数据（以数字图像的形式），收集大地理区域的信息，以帮助防止不受控制的火灾造成的破坏。该项目的长期目标/愿景是开发一个主要利用遥感但有其他辅助数据支持的可靠的可操作森林火灾发生预测和管理系统。建议研究的首要要素是：(i)利用河流流量和土壤湿度的关系预测森林火险条件，将这些数据整合到我们最近开发的基于遥感的森林火险预测系统的框架中；确定历史上森林火灾发生的趋势及其与大尺度大气环流和气候变量的关系；利用天基平台估计森林火灾造成的二氧化碳排放量；（iv）制定燃料管理战略，通过使用(i)的结果和遥感衍生的叶面积指数数据（森林冠层的测量）来降低火灾的严重程度。该研究计划的成功完成将产生一个森林火灾发生管理系统的原型。它将在森林火灾管理方面具有开创性；森林资源的可持续性和经济性；指导研究和规范处理火灾事件。此外，基于遥感的森林火灾二氧化碳排放量的快速可靠估算将有助于决策者了解进入大气的非人为二氧化碳的数量。这种知识将对推进加拿大年度温室气体清单以及泛加拿大清洁增长和气候变化框架产生重大影响，以实现“健康的环境和健康的经济”。此外，拟议的研究计划将培训四名高素质的专业人员（hqp），他们在更广泛的遥感、自然灾害和资源管理领域拥有独特的市场技能，这些领域是加拿大经济可持续增长和进一步发展所需要的。

Over the past 25 years, Canada has had an average of 7300 fires annually - each year about 2.5 million hectares of forest are destroyed. Over the past decade alone, Canadians have spent between $800 million and $1.5 billion per annum on fighting fires. Though fires are a part of the natural cycle and rejuvenate the soil, kill pests and lead to new growth, uncontrolled fires also cause significant detrimental impacts, such as (i) economic damage due to loss of mature trees; (ii) increase the CO2 loading into the atmosphere; (iii) air quality deterioration during and immediately after the fires; and (iv) loss of human life and property. Thus, the development of an efficient forest fire occurrence and management system is crucial to offset such adverse impacts. This proposed research program aims to use remote sensing data (in the form of digital imagery) acquired by various satellite platforms that gather information over large geographic areas to help prevent the devastation of uncontrolled fires. The long-term goal/vision of this program is to develop a robust operational forest fire occurrence prediction and management system primarily using remote sensing but supported by other ancillary data. The foremost elements of the proposed research are to (i)predict forest fire danger conditions using the relationships of river flow regimes and soil moisture, integrating this data in the framework of our recently developed remote sensing-based forest fire danger forecasting system; (ii)determine trends in the occurrences of historical forest fires and their relationships to large-scale atmospheric circulations and climate variables; (iii) estimate CO2 emissions from forest fires using space-borne platforms; and (iv) develop fuel management strategies to reduce fire severity by using the outcomes of (i) and remote sensing-derived leaf area index data, a measurement of forest canopy. The successful completion of the proposed research program will result in a prototype forest fire occurrence management system. It will be ground-breaking in forest fire management; in sustainability and economics of forest resources; and in guiding research and codes in handling fire instances. Furthermore, remote sensing-based quick and reliable estimation of CO2 emissions from the forest fires will help policy makers understand the amount of non-anthropogenic CO2 entering the atmosphere. Such knowledge will be highly impactful in advancing Canada's annual greenhouse gas inventory as well as the Pan-Canadian Framework on Clean Growth and Climate Change to achieve "A Healthy Environment and a Healthy Economy." Additionally, the proposed research program will train four highly qualified professionals (HQPs) with unique, market-ready skills in the broader field of remote sensing, natural hazard and resource management needed for sustainable growth and further development of the Canadian economy.