Use of Remote Sensing in Developing a Forest Fire Occurrence Management System

在加拿大，森林火灾是一种严重的干扰。在过去的25年里，平均每年发生8300起火灾。每年平均有230万公顷的森林被烧毁。虽然火在丰富土壤养分、杀死昆虫和疾病、再生森林方面带来了一些优势;但是，它也造成了大量的有害影响。其中包括：（i）火警地点附近的树木及财物被烧毁，引致经济损失;（ii）增加大气中的二氧化碳含量;（iii）空气质素恶化，危害健康;及（iv）扑灭大型火警时可能引致人命损失等。此外，在过去十年中，加拿大平均每年花费5亿至10亿加元用于灭火。因此，开发一个有效的森林火灾发生管理系统是至关重要的，以制定可持续的火灾管理战略，以抵消不利影响。在这方面，目的是利用能够定期收集大面积地理区域信息的卫星平台所获得的遥感数据（以数字图像的形式）。本研究计画的总体目标是开发一个森林火灾发生管理系统。拟议研究的首要内容是利用遥感数据：* ㈠加强最近开发的基于遥感的森林火灾危险预报系统;* ㈡以更高的时间分辨率确定陆地表面的物候阶段（即，* ㈢结合前两个要素的结果，模拟火灾发生;* ㈣利用第一个要素的结果和遥感得出的叶面积指数数据，制定燃料管理战略，在发生火灾时降低火灾的严重程度。这项研究计划的成功执行将导致一个原型森林火灾发生管理系统。它将在森林火灾管理、森林资源的可持续性和经济学以及指导火灾处理的研究和守则等方面非常有用。此外，作为拟议研究方案的一个副产品，打算开发一个基于开放源码平台的中分辨率成像光谱仪图像处理软件包;该软件包将向公众提供。这类行动将使遥感数据在加拿大和世界其他地方的自然资源和自然危害/灾害管理活动中的使用激增。

In Canada, forest fire is a critical disturbance. In the last 25 years, there have been on average 8300 fires annually. Each year on average 2.3 million hectares of forest have been burned. Though fire brings some advantages in enriching soil nutrients, killing insects and diseases, regenerating forests; however, it causes a significant amount of detrimental impacts. Those include: (i) economic loss due to burning of the trees and properties nearby the fire locations; (ii) increasing the carbon dioxide loading in the atmosphere; (iii) health hazard through air quality deterioration; and (iv) possible loss of human life in fighting large fires, among others. Also, every year on average Canada has spent CAD $500 million to $1 billion in fighting fires during the last decade. Thus, development of an efficient forest fire occurrence management system is crucial to develop sustainable fire management strategies in order to offset the adverse impacts. In this context, the intent is to use remote sensing data (in the form of digital imagery) acquired by the satellite platforms capable of gathering information over a large geographic area periodically.*******The overall goal of this proposed research program is to develop a forest fire occurrence management system. The foremost elements of the proposed research are the employment of remote sensing data to:****(i) enhance the recently developed remote sensing-based forest fire danger forecasting system;****(ii) determine land surface phenological stages at an enhanced temporal resolution (i.e., 2-day), and their relationships with fire occurrences;****(iii) model fire occurrences by combining the outcomes of the first two elements; and****(iv) develop fuel management strategies to reduce fire severity in the event of fire occurrences by using the outcomes of the first element and remote sensing-derived leaf area index data.*******The successful execution of this proposed research program will result in a prototype forest fire occurrence management system. It will be very useful in forest fire management; in sustainability and economics of forest resources; and in guiding research and codes in handling fire instances, among others. In addition, as a by-product of the proposed research program, the intent is to develop a Moderate Resolution Imaging Spectroradiometer (MODIS) image processing software package based on an open source platform; which will be made publicly available. Such action will proliferate the use of remote sensing data in natural resource and natural hazard/disaster management activities in Canada and elsewhere in the world.***