Urban smoke exposure events from large wildfires have become increasingly common in California and throughout the western United States. The ability to study the impacts of high smoke aerosol exposures from these events on the public is limited by the availability of high-quality, spatially resolved estimates of aerosol concentrations. Methods for assigning aerosol exposure often employ multiple data sets that are time-consuming to create and difficult to reproduce. As these events have gone from occasional to nearly annual in frequency, the need for rapid smoke exposure assessments has increased. The rapidfire (relatively accurate particulate information derived from inputs retrieved easily) R package (version 0.1.3) provides a suite of tools for developing exposure assignments using data sets that are routinely generated and publicly available within a month of the event. Specifically, rapidfire harvests official air quality monitoring, satellite observations, meteorological modeling, operational predictive smoke modeling, and low-cost sensor networks. A machine learning approach, random forest (RF) regression, is used to fuse the different data sets. Using rapidfire, we produced estimates of ground-level 24 h average particulate matter for several large wildfire smoke events in California from 2017-2021. These estimates show excellent agreement with independent measures from filter-based networks.
