Ozone trends and their relationship to characteristic weather patterns

Area of Science:

• Environmental Science
• Atmospheric Chemistry
• Public Health

Background:

• Ozone concentration trends can vary locally due to meteorological conditions.
• Trends at ozone distribution extremes, crucial for health, are often unreported.

Purpose of the Study:

• To categorize daily weather patterns and analyze their impact on ozone concentration trends.
• To investigate if ozone trends differ at the extremes (5th and 95th percentiles) versus the median.
• To assess the interaction between local weather patterns and ozone concentration changes over time.

Main Methods:

• K-means clustering was used to group 16 years of observational data into five distinct weather categories based on temperature, wind speed, water vapor pressure, and barometric pressure.
• Poisson regression with penalized splines and quantile regression were employed to model trends in cluster frequency and ozone concentrations.
• Analysis focused on changes in mean, median, and extreme ozone percentiles within each weather cluster.

Main Results:

• Five characteristic weather groupings were identified, with a decrease in days of high temperature variability and an increase in warmer days with stable temperatures.
• Ozone concentration trends varied significantly across the different weather clusters.
• Rates of ozone change at the 95th and 5th percentiles differed significantly from median trends in several weather categories.

Conclusions:

• Local ozone concentration trends are demonstrably influenced by characteristic weather patterns.
• The interaction between changing weather conditions and ozone levels is significant, highlighting the need to consider weather in air quality management.
• Ozone trends at distribution extremes present different patterns than median trends, with potential implications for public health assessments.