使用机器学习评估西伯利亚森林野火的范围
Ivan P Malashin1, Igor Masich2, Vladimir Nelyub2
1Bauman Moscow State Technical University, 105005, Moscow, Russia. ivan.p.malashin@gmail.com.
Scientific reports
|September 25, 2025
概括
机器学习模型可以使用天气和森林数据预测西伯利亚森林野火的大小. XGBoost获得了88.8%的准确性,确定城市附近和干燥的条件是大火的关键因素.
科学领域:
- 林业林业 林业 林业 林业
- 生态生态学 生态生态学
- 计算机科学 计算机科学
背景情况:
- 野火对全球生态系统和森林管理构成重大威胁.
- 西伯利亚的森林,特别是克拉斯诺亚尔斯克州的森林,容易发生大规模的火灾事件.
- 准确的野火规模估计对于有效的管理和生态影响评估至关重要.
研究的目的:
- 开发和评估一种机器学习 (ML) 框架,用于估计西伯利亚克拉斯诺亚尔斯克州地区野火的大小.
- 确定影响野火规模的关键环境和气象因素.
- 为了比较各种ML模型的性能,用于野火大小分类.
主要方法:
- 使用集成气象变量 (温度,湿度,风,降水),森林组成,检测数据和历史火灾记录的数据集.
- 训练并比较多个ML模型:XGBoost,随机森林,K-最近邻居,后勤回归和决策树.
- 采用SHAP (夏普利添加式解释) 分析来解释模型预测和特征的重要性.
主要成果:
- XGBoost 显示了最高的分类准确度,达到 88.8%,超过了其他测试的 ML 模型.
- 功能重要性分析显示,靠近城市地区,风力模式和影响燃料湿度的气象条件显著影响火灾规模.
- SHAP分析表明,局部天气条件与较小的火灾相关,而长时间的干旱时期与较大的火灾事件有关.
结论:
- 开发的ML框架显示,在研究的西伯利亚地区,野火大小分类的潜力很大.
- 这些发现强调了气象因素和人类接近在确定野火范围方面的关键作用.
- 该框架目前是探索性和区域特定的,需要对更广泛的应用进行本地数据校准.
更多相关视频
08:16Collecting and Processing Drone-based Remotely Sensed Data for Use in Forest Recovery Monitoring
Published on: October 24, 2025
488
12:26Integrating Remote Sensing with Species Distribution Models; Mapping Tamarisk Invasions Using the Software for Assisted Habitat Modeling SAHM
Published on: October 11, 2016
13.8K
相关概念视频
Steps in Outbreak Investigation
492
In the ever-evolving field of public health, statistical analysis serves as a cornerstone for understanding and managing disease outbreaks. By leveraging various statistical tools, health professionals can predict potential outbreaks, analyze ongoing situations, and devise effective responses to mitigate impact. For that to happen, there are a few possible stages of the analysis:
492
Applications of GIS: Disaster Management and Emergency Response
471
Geographic Information System (GIS) technology is essential for risk identification, action prioritization, and resource optimization in critical situations like flooding and earthquakes. By integrating spatial and demographic data, GIS provides a comprehensive framework for emergency response.GIS integrates data layers, like rainfall intensity, topography, elevation profiles, and river levels, to model high-risk flood zones. These layers assess areas susceptible to flooding based on their...
471
