Complex Relationships of the Effects of Topographic Characteristics and Susceptible Tree Cover on Burn Severity

• Main Authors: Hyun-Joo Lee, Yun Eui Choi, Sang-Woo Lee
• Format: Article
• Language: English
• Published: MDPI AG 2018-01-01
• Series: Sustainability
• Subjects: burn severity; GAM; Japanese red pine; nonlinear relationship; topography
• Online Access: http://www.mdpi.com/2071-1050/10/2/295

Forest fires and burn severity mosaics have profound impacts on the post-fire dynamics and complexity of forest ecosystems. Numerous studies have investigated the relationship between topographic variables and susceptible tree covers with regard to burn severity. However, these relationships have not been fully elucidated, because most studies have assumed linearity in these relationships. Therefore, we examined the linearity and the nonlinearity in the relationships between topographic variables and susceptible tree covers with burn severity by comparing linear and nonlinear models. The site of the Samcheok fire, the largest recorded forest fire in Korea, was used as the study area. We generated 802 grid cells with a 500-m resolution that encompassed the entire study area and collected a dataset that included the topographic variables and percentage of red pine trees, which are the most susceptible tree cover types in Korea. We used conventional linear models and generalized additive models to estimate the linear and the nonlinear models based on topographic variables and Japanese red pine trees. The results revealed that the percentage of red pine trees had linear effects on burn severity, reinforcing the importance of silviculture and forest management to lower burn severity. Meanwhile, the topographic variables had nonlinear effects on burn severity. Among the topographic variables, elevation had the strongest nonlinear effect on burn severity, possibly by overriding the effects of susceptible fuels over elevation effects or due to the nonlinear effects of topographic characteristics on pre-fire fuel conditions, including the spatial distribution and availability of susceptible tree cover. To validate and generalize the nonlinear effects of elevation and other topographic variables, additional research is required at different fire sites with different tree cover types in different geographic locations.