Are Past Wildfire Activities Correlated with the Frequency of Different Types of Post-Wildfire Geomorphic Events such as Landslides and Debris Flows

Around the world, the frequency and intensity of wildfire events are rapidly increasing. Such a trend increasingly exposes some communities to the risk of the secondary hazards of wildfire, such as post-wildfire geomorphic events like a landslide and debris flow. To better manage the risk imposed by the secondary hazards of wildfire, a better understanding of the relationship between wildfire activities and the geomorphic events related therewith is necessary. In this paper, the temporal relationship between past wildfire events and the frequency of two different types of geomorphic events (landslides and debris flow) was studied statistically through the conduction of a time series analysis. The analysis conducted using the data created by the construction of a geomorphic event inventory derived from a set of Landsat based land cover classification of the region of the Rocky and Omineca Mountains in the Canadian province of British Columbia ranging from 1986 to 2019, and the wildfire boundary data provided by the Federal Government of Canada. Both a simple linear regression and a Poisson regression were conducted to analyze the relationship between time and the frequency of geomorphic events, with the application of a slope-based stratification. The study has three key findings. The first finding is that overall, there exists a strong negative temporal correlation between the time elapsed since the most recent event of a wildfire and the frequency of post-wildfire geomorphic events. Second, slope appears to be an influential factor for the relationship between wildfire and the frequency of post-wildfire geomorphic events. Third, despite the presence of a strong correlation, time is a very poor variable for explaining the dynamics of the frequency of post-wildfire geomorphic events.

全球范围内，野火事件的发生频率与强度正迅速上升。这种趋势正日益将部分社区置于野火次生灾害的风险之下，例如滑坡、泥石流等野火后地貌事件（geomorphic events）。为更好地管理野火次生灾害带来的风险，有必要深入理解野火活动与相关地貌事件之间的关系。本文通过时间序列分析（time series analysis）的方法，对历史野火事件与两类不同地貌事件（滑坡和泥石流）发生频率之间的时间关系进行了统计研究。该分析所使用的数据包括：基于Landsat的1986-2019年加拿大不列颠哥伦比亚省落基山脉与奥米尼卡山脉区域土地覆盖分类构建的地貌事件清单，以及加拿大联邦政府提供的野火边界数据。研究应用基于坡度的分层方法，分别采用简单线性回归（simple linear regression）和泊松回归（Poisson regression）分析了时间与地貌事件发生频率之间的关系。本研究得出三项关键结论：首先，总体而言，最近一次野火发生后的时间间隔与野火后地貌事件的发生频率之间存在强烈的负时间相关性；其次，坡度是影响野火与野火后地貌事件发生频率之间关系的重要因素；第三，尽管存在强烈的相关性，时间变量对解释野火后地貌事件发生频率的动态变化效果极差。