| 英文摘要 |
To understand the relationship between overall weather patterns and air quality, this project integrates daily weather patterns and continuously expands a historical database from 2022 until now. It also includes the addition of pollution trend charts and optimized query functionalities. Furthermore, the project has completed the "Air Quality Data Statistical Query" feature, which reduces the time for forecasters to search and organize relevant information and provides decision-making references for determining air quality forecasts based on specific weather patterns.
In addition, the project has compiled data on the distribution of typhoon locations during the peripheral circulation and impact periods from 2009 to 2021. It analyzes the concentrations of PM2.5, PM10, O3, and O3, 8hr at grid points and monitoring stations, and calculates the number and frequency of pollution event days in each grid. By considering the historical typhoon locations and their accompanying flow patterns, the project provides "Typhoon Air Quality Forecasts" and has also completed the "High-Low Pressure Query" feature, which allows forecasters to select different time periods and pressure locations to obtain an overview of air quality in different regions.
Additionally, the project assists in providing a 7-day air quality outlook and supports the analysis of air quality pollution events by the EPA. It offers simulation results using the AERMOD dispersion model and the Hysplit trajectory model. To explore the relationship between ozone air pollution events and weather patterns, the project analyzes the factors influencing ozone fluctuations on specific ozone event days. Through quantile regression analysis and decision tree algorithms, it establishes decision tree models for individual monitoring stations in various air quality zones, providing references for ozone forecasting. The project has also produced two popular science graphics and texts to convey accurate information about air quality and pollution causes, aiding public understanding. Additionally, an educational training session on the AERMOD dispersion model has been conducted to enhance forecasters' familiarity with the model, enabling its application in emergency event analysis and related support work in the future.
|