Air pollution significantly impacts health. Various natural and anthropogenic factors influence particulate matter (PM) concentration, exhibiting different cyclic patterns and temporal distributions. Understanding these factors is crucial for planning mitigation actions and predicting pollution levels. In moderate climates, meteorological factors vary in importance between warm and cold periods, often influenced by human activities and energy use. This study utilizes data from PM sensors in Krakow, Poland, a city that pioneered legal changes to protect air quality. For long-term analyses, reference measurements from nine years were used, while short-term analyses with spatial considerations employed data from 52 low-cost sensors. The ensemble methods with the Boosted Regression Trees were used for feature analysis to overcome challenges with rare high-emission peaks. The research demonstrated the importance of legislation in achieving long-term reductions in PM concentrations. The most important short-term factors are: surface pressure, wind speed, and soil moisture.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Big Data Analysis of Long-Term Anthropogenic and Short-Term Natural Factors Influencing Air Pollution in Moderate Climate Zones: Implications for Sustainable Development

  • Mateusz Zareba,
  • Tomasz Danek

摘要

Air pollution significantly impacts health. Various natural and anthropogenic factors influence particulate matter (PM) concentration, exhibiting different cyclic patterns and temporal distributions. Understanding these factors is crucial for planning mitigation actions and predicting pollution levels. In moderate climates, meteorological factors vary in importance between warm and cold periods, often influenced by human activities and energy use. This study utilizes data from PM sensors in Krakow, Poland, a city that pioneered legal changes to protect air quality. For long-term analyses, reference measurements from nine years were used, while short-term analyses with spatial considerations employed data from 52 low-cost sensors. The ensemble methods with the Boosted Regression Trees were used for feature analysis to overcome challenges with rare high-emission peaks. The research demonstrated the importance of legislation in achieving long-term reductions in PM concentrations. The most important short-term factors are: surface pressure, wind speed, and soil moisture.