Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Estimating Forest Fire Risk Using Simple Fuzzy Modelling in Dalarna, Sweden (2018)

Kowalczyk, Anna Maria LU (2025) In Student thesis series INES NGEK01 20251
Dept of Physical Geography and Ecosystem Science
Abstract
In recent years, forest fires have become more frequent and severe worldwide. Sweden experienced an extreme fire season in 2018, resulting in extensive burnt areas and challenges in fire suppression. With climate change such extreme fire seasons are likely to become more common, highlighting the urgent need for effective fire risk assessment tools. This study aimed to develop a simplified forest fire risk model based on 2018 fire season in Dalarna County, Sweden, using fuzzy logic and the Analytic Hierarchy Process (AHP). The model incorporated nine factors: temperature, precipitation, relative humidity, wind, slope, aspect, forest type, and distances to roads and settlements. Two aggregation methods of OR operator and Weighted Linear... (More)
In recent years, forest fires have become more frequent and severe worldwide. Sweden experienced an extreme fire season in 2018, resulting in extensive burnt areas and challenges in fire suppression. With climate change such extreme fire seasons are likely to become more common, highlighting the urgent need for effective fire risk assessment tools. This study aimed to develop a simplified forest fire risk model based on 2018 fire season in Dalarna County, Sweden, using fuzzy logic and the Analytic Hierarchy Process (AHP). The model incorporated nine factors: temperature, precipitation, relative humidity, wind, slope, aspect, forest type, and distances to roads and settlements. Two aggregation methods of OR operator and Weighted Linear Combination (WLC) were applied and compared to determine the most effective approach. Model evaluation was conducted using historical fire points, burnt areas, and Kappa statistics. Resulting weights from AHP combined with literature showed that climatic factors and forest type had the greatest influence on fire risk estimations, while topographic weights were the least significant. Evaluation from overlay showed that 83% of fire points and 93% of burnt areas fell within medium to very high-risk zones, indicating good spatial correlation. Even though Kappa values were low (0.02 for OR and 0.1 for WLC), they showed that WLC outperformed the OR method, suggesting its better suitability for this study. Additionally, it highlighted the need for larger sample size and consideration of both evaluation methods to more accurately assess model performance. Despite its limitations, the simple fuzzy model developed in this thesis can already serve as a useful decision-support tool for identifying high-risk areas and improving forest fire management strategies, especially important in a changing climate. (Less)
Please use this url to cite or link to this publication:
author
Kowalczyk, Anna Maria LU
supervisor
organization
course
NGEK01 20251
year
type
M2 - Bachelor Degree
subject
keywords
Forest fire, Sweden, 2018, fire season, factors, AHP, fuzzy
publication/series
Student thesis series INES
report number
697
language
English
id
9196536
date added to LUP
2025-06-11 11:51:24
date last changed
2025-06-11 11:51:24
@misc{9196536,
  abstract     = {{In recent years, forest fires have become more frequent and severe worldwide. Sweden experienced an extreme fire season in 2018, resulting in extensive burnt areas and challenges in fire suppression. With climate change such extreme fire seasons are likely to become more common, highlighting the urgent need for effective fire risk assessment tools. This study aimed to develop a simplified forest fire risk model based on 2018 fire season in Dalarna County, Sweden, using fuzzy logic and the Analytic Hierarchy Process (AHP). The model incorporated nine factors: temperature, precipitation, relative humidity, wind, slope, aspect, forest type, and distances to roads and settlements. Two aggregation methods of OR operator and Weighted Linear Combination (WLC) were applied and compared to determine the most effective approach. Model evaluation was conducted using historical fire points, burnt areas, and Kappa statistics. Resulting weights from AHP combined with literature showed that climatic factors and forest type had the greatest influence on fire risk estimations, while topographic weights were the least significant. Evaluation from overlay showed that 83% of fire points and 93% of burnt areas fell within medium to very high-risk zones, indicating good spatial correlation. Even though Kappa values were low (0.02 for OR and 0.1 for WLC), they showed that WLC outperformed the OR method, suggesting its better suitability for this study. Additionally, it highlighted the need for larger sample size and consideration of both evaluation methods to more accurately assess model performance. Despite its limitations, the simple fuzzy model developed in this thesis can already serve as a useful decision-support tool for identifying high-risk areas and improving forest fire management strategies, especially important in a changing climate.}},
  author       = {{Kowalczyk, Anna Maria}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Student thesis series INES}},
  title        = {{Estimating Forest Fire Risk Using Simple Fuzzy Modelling in Dalarna, Sweden (2018)}},
  year         = {{2025}},
}