Statistical and simulation modelling of boreal fire regimes

闪电引起的野火在加拿大北方森林是一种自然现象，经常发生。在许多领域，它们是构建这个生态系统的最重要的过程之一。“火灾制度”是根据特定地区火灾的数量、规模、严重程度和一年中的时间来定义的。本研究的主要目标是描述加拿大北部地区的火灾制度如何变化，然后量化这种变化的原因。当然，大致的轮廓是众所周知的：例如，气候是火灾的主要驱动因素。大火发生在温暖干燥的条件下。因此，在气候变暖的情况下，预计北方许多地区每年平均被烧毁的面积将增加。然而，越来越清楚的是，植被也起着重要的作用。不同类型和年龄的北方森林有不同的火灾发生或蔓延的可能性。例如，如果年轻的森林能抑制火，那么火一定在某种程度上是自我限制的。同样，由于森林管理现在操纵了北方大部分地区的年龄结构和物种组成，它也可能影响未来的火灾状态，并且正如我最近的一项研究所表明的那样，减轻了一些与气候变暖有关的火灾影响。我们最近关于北方火灾的大部分数据都是由消防管理机构收集的，他们的任务是扑灭这些火灾。他们成功的程度尚不清楚，这本身就很有趣；然而，要了解气候和植被的影响，我们还需要了解火灾管理的影响，这反映在我们现有的数据和我们今天看到的森林中。因此，本研究的具体目标是开发预测统计模型，量化所有这些因素对火灾制度的两个方面的影响：火灾数量和火灾规模分布。我还寻求将这些统计模型直接与空间模拟工具联系起来的新方法，并将其应用于森林管理和保护规划。这项工作的直接受益者将是对加拿大北部森林和湿地的可持续管理和保护负有责任或感兴趣的政府机构、工业和环境组织。

Wildfires caused by lightning are a natural and regular occurrence in Canada's boreal forests. In many areas, they are one of the most important processes structuring this ecosystem. A "fire regime" is defined by the number, size, severity and time-of-year of the fires characteristic of a given region. The main goals of this research are to describe how fire regimes vary across the Canadian boreal and then to quantify the reasons for this variation. Of course, the broad outlines are well known: climate, for example is a main driver of fire. Large fires occur under warm, dry conditions. Thus, under climate warming, the average area burned per year is expected to increase in many parts of the boreal. However, it is becoming clear that vegetation also plays an important role. Different types and age of boreal forest have different potentials for fires to start or spread. If young forest, for example, inhibits fire, then fire must be to some extent self-limiting. Similarly, because forest management now manipulates the age-structure and species composition of large parts of the boreal, it too could influence future fire regime and, as one of my recent studies suggests, mitigate some fire-related effects of climate warming. Much of our recent data on boreal fires is collected by fire management agencies, whose mission is to put these fires out. The extent to which they have been successful remains unclear, and is of considerable interest in itself; however, to understand the effects of climate and vegetation, we need to understand also the effects of fire management, as reflected in the data we have and the forests we see today. Thus, the specific objectives of this research are to develop predictive statistical models that quantify the effects of all these factors on two aspects of fire regime: the number of fires, and the fire size distribution. I also seek new methods for linking these statistical models directly to spatial simulation tools, with applications to forest management and conservation planning. The direct beneficiaries of this work will be government agencies, industries and environmental organisations having responsibilities for or interests in the sustainable management and conservation of Canada's northern forests and wetlands.