Predicting dust storm susceptibility: exploring control strategies with XGBoost models
(2024) In Student Thesis series INES NGEM01 20231Dept of Physical Geography and Ecosystem Science
- Abstract
- Dust storms are meteorological phenomena that occur in arid and semi-arid regions, causing harm to the environment and human health. The frequency and intensity of dust storms have increased in recent years due to climate change and human activities. In response, various dust storm mitigation strategies were introduced, including land use management, afforestation, and mulching. The existing research highlights various strategies for addressing the issue, however, there is a lack of specific regional modelling to evaluate their effectiveness.
This study simulates five dust storm mitigation strategies in three different regions in the Tigris-Euphrates River Basin (TERB) in the Middle East, and measures their effectiveness based on the... (More) - Dust storms are meteorological phenomena that occur in arid and semi-arid regions, causing harm to the environment and human health. The frequency and intensity of dust storms have increased in recent years due to climate change and human activities. In response, various dust storm mitigation strategies were introduced, including land use management, afforestation, and mulching. The existing research highlights various strategies for addressing the issue, however, there is a lack of specific regional modelling to evaluate their effectiveness.
This study simulates five dust storm mitigation strategies in three different regions in the Tigris-Euphrates River Basin (TERB) in the Middle East, and measures their effectiveness based on the decrease in dust storm susceptibility of the area. The strategies included increasing vegetation cover through expanding cropland and/or natural vegetation and expanding lakes through the partial opening of upstream dams. Extreme Gradient Boosting (XGBoost) machine learning algorithm was utilized for dust source susceptibility mapping. One model was created at pixel-level, which was the default for all simulations analysis. A second model was built to ascertain spatial influence on local areas following the simulations.
The study found no effect of the Spatial Model on the neighbouring pixels following the application of the mitigation strategies.
The study’s main finding was that all simulations decreased the areas’ overall susceptibility to dust storms. The study revealed a spatial pattern where susceptibility increased and the response to mitigation strategies diminished progressively from the northern to the southern part of the TERB region. Further detailed analysis is recommended to consider region-specific factors in the region’s response to mitigation strategies. Overall, this study provided insights into the dust storm control strategies’ effectiveness. It has the potential to serve as a starting point for further research into modelling the effects and to aid in constructive and scientifically based mitigation planning. (Less) - Popular Abstract
- Dust storms are increasingly prevalent in arid and semi-arid regions, posing environmental and health risks. To address this issue, various dust storm mitigation strategies were proposed, including land use management, afforestation, and mulching. However, there is a lack of specific regional modelling to assess their effectiveness. This study conducts simulations of five dust storm mitigation strategies in the Tigris-Euphrates River Basin (TERB) in the Middle East. Using a popular machine learning algorithm (XGBoost), their impact on dust storm susceptibility was evaluated.
The results showed that all simulations led to a reduction in the overall susceptibility to dust storms in the study area. The strategies included expansion of... (More) - Dust storms are increasingly prevalent in arid and semi-arid regions, posing environmental and health risks. To address this issue, various dust storm mitigation strategies were proposed, including land use management, afforestation, and mulching. However, there is a lack of specific regional modelling to assess their effectiveness. This study conducts simulations of five dust storm mitigation strategies in the Tigris-Euphrates River Basin (TERB) in the Middle East. Using a popular machine learning algorithm (XGBoost), their impact on dust storm susceptibility was evaluated.
The results showed that all simulations led to a reduction in the overall susceptibility to dust storms in the study area. The strategies included expansion of lakes and increase in vegetation cover through expanding cropland and natural vegetation. Notably, the effectiveness of these strategies showed differences depending on the region analysed. The northern part of the TERB region showed more significant improvement compared to the southern areas. Across the tested areas, expanding the lake size was the best performing strategy. Land use change strategies worked well in the northern regions, but not in the south.
This study emphasizes the need to consider region-specific factors when implementing dust storm mitigation measures. It provides valuable insights into the effectiveness of the selected strategies, and the additional usage of susceptibility mapping as a tool. The study can serve as a basis for further research, and informed mitigation planning in the future. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9137019
- author
- Musur, Malgorzata Anna LU
- supervisor
-
- Pengxiang Zhao LU
- Ali Mansourian LU
- organization
- course
- NGEM01 20231
- year
- 2024
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Physical Geography, Ecosystem Analysis, Dust storms, Susceptibility mapping, Mitigation strategies, Tigris-Euphrates River Basin, Machine learning, XGBoost
- publication/series
- Student Thesis series INES
- report number
- 626
- language
- English
- id
- 9137019
- date added to LUP
- 2023-09-11 10:27:57
- date last changed
- 2024-09-08 23:13:28
@misc{9137019, abstract = {{Dust storms are meteorological phenomena that occur in arid and semi-arid regions, causing harm to the environment and human health. The frequency and intensity of dust storms have increased in recent years due to climate change and human activities. In response, various dust storm mitigation strategies were introduced, including land use management, afforestation, and mulching. The existing research highlights various strategies for addressing the issue, however, there is a lack of specific regional modelling to evaluate their effectiveness. This study simulates five dust storm mitigation strategies in three different regions in the Tigris-Euphrates River Basin (TERB) in the Middle East, and measures their effectiveness based on the decrease in dust storm susceptibility of the area. The strategies included increasing vegetation cover through expanding cropland and/or natural vegetation and expanding lakes through the partial opening of upstream dams. Extreme Gradient Boosting (XGBoost) machine learning algorithm was utilized for dust source susceptibility mapping. One model was created at pixel-level, which was the default for all simulations analysis. A second model was built to ascertain spatial influence on local areas following the simulations. The study found no effect of the Spatial Model on the neighbouring pixels following the application of the mitigation strategies. The study’s main finding was that all simulations decreased the areas’ overall susceptibility to dust storms. The study revealed a spatial pattern where susceptibility increased and the response to mitigation strategies diminished progressively from the northern to the southern part of the TERB region. Further detailed analysis is recommended to consider region-specific factors in the region’s response to mitigation strategies. Overall, this study provided insights into the dust storm control strategies’ effectiveness. It has the potential to serve as a starting point for further research into modelling the effects and to aid in constructive and scientifically based mitigation planning.}}, author = {{Musur, Malgorzata Anna}}, language = {{eng}}, note = {{Student Paper}}, series = {{Student Thesis series INES}}, title = {{Predicting dust storm susceptibility: exploring control strategies with XGBoost models}}, year = {{2024}}, }