The shifting frontiers of air pollution emission sources contribute to stagnation or reversal of air quality gains across the United States (US). The frequency and possible duration of Exceptional Events - driven primarily by wildfires and dust storms - have significantly increased in the US over the past decade. Combined with the US Environmental Protection Agency (EPA) final rule strengthening primary annual National Ambient Air Quality Standards (NAAQS) for PM2.5 by 25%, communities will need to reevaluate domestic and international sources of PM2.5.This study applies the Isolation Forest methodology to Exceptional Event demonstrations to flag and evaluate sources of anomalies in large PM2.5 measurement datasets. Focusing on a decade of hourly PM2.5 data measured in seven regions across Texas from 2012 to 2021 (>3 million data points), we present methods to efficiently flag hourly PM2.5 anomalies with compute times of ~minutes and characterize their spatial impacts as local or (multi-) regional; subsequent evaluation of potential sources of the increase can then be conducted more efficiently in a targeted manner. For a subset of anomalies, we incorporate air mass back trajectories, surface influences, and positive matrix factorization to evaluate potential sources.Our anomaly characterization method separated statistically normal PM2.5 data and enabled differentiation of localized versus larger-scale PM2.5 sources. In addition, our method successfully characterized the Summer 2020 severe Saharan dust intrusions into Texas, as well as the influence of international smoke from Mexico on El Paso's regional air quality.This anomaly flagging and characterization method is promising for assessing the relative importance of sources to anomalies in PM2.5 and other criteria air pollutants for multiple purposes; while this work focuses on its capacity for exceptional event demonstrations, the applicability includes long-term trend analyses from environmental justice analyses of air pollutant exposure to air quality attainment demonstrations.Implications: The shifting frontiers of air pollution emission sources contribute to stagnation or reversal of air quality gains across the United States (US). The frequency and possible duration of Exceptional Events - driven primarily by wildfires and dust storms - have significantly increased in the US over the past decade. Combined with the US Environmental Protection Agency (EPA) final rule strengthening primary annual National Ambient Air Quality Standards (NAAQS) for PM2.5 by 25%, communities will need to reevaluate domestic and international sources of PM2.5. This study presents a robust methodology to rapidly flag and evaluate sources of anomalies in PM2.5 measurements. This anomaly flagging and characterization method is promising for assessing the relative importance of sources to anomalies in PM2.5 and other criteria air pollutants for multiple purposes; while this work focuses on its capacity for exceptional event demonstrations, the applicability includes long-term trend analyses from environmental justice analyses of air pollutant exposure to air quality attainment demonstrations.