This study examines PM2.5 dynamics in five European capitals, Madrid, Berlin, Paris, Rome, and Warsaw, over January 2020 to January 2024. We rely on validated, open data accessed via OpenAQ, which aggregates measurements reported to the European Environment Agency. Our aims are twofold: (i) to describe seasonal and spatial patterns across contrasting urban contexts; and (ii) to assess the value of those patterns for more adaptive, evidence-informed policy. The workflow covers cleaning, monthly aggregation, and comparative time-series exploration, complemented by a hybrid forecaster that pairs Prophet with a Long Short-Term Memory network. Results show clear winter peaks and marked cross-city differences in both levels and variability. These contrasts reflect not only meteorology and geography but also choices in transport, energy, and regulation. We conclude that high-resolution open data are essential for transparent, context-sensitive decision-making.

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

AI-Powered Assessment of Air Pollution in Europe: OpenAQ Insights 2020–2024

  • Jesús Cáceres-Tello,
  • Ziwei Shu,
  • Jose Javier Galán-Hernández

摘要

This study examines PM2.5 dynamics in five European capitals, Madrid, Berlin, Paris, Rome, and Warsaw, over January 2020 to January 2024. We rely on validated, open data accessed via OpenAQ, which aggregates measurements reported to the European Environment Agency. Our aims are twofold: (i) to describe seasonal and spatial patterns across contrasting urban contexts; and (ii) to assess the value of those patterns for more adaptive, evidence-informed policy. The workflow covers cleaning, monthly aggregation, and comparative time-series exploration, complemented by a hybrid forecaster that pairs Prophet with a Long Short-Term Memory network. Results show clear winter peaks and marked cross-city differences in both levels and variability. These contrasts reflect not only meteorology and geography but also choices in transport, energy, and regulation. We conclude that high-resolution open data are essential for transparent, context-sensitive decision-making.