Exploring the Potential of Low-Cost Air Quality Sensors and Comparing HRRR Forecast vs Analysis Inputs in a High-Resolution Model for Prescribed Burn Applications
Nigro, E., Shams, S., Williams, M., Piña, A.. (2026). Exploring the Potential of Low-Cost Air Quality Sensors and Comparing HRRR Forecast vs Analysis Inputs in a High-Resolution Model for Prescribed Burn Applications. , doi:https://doi.org/10.5065/6mx6-km36
| Title | Exploring the Potential of Low-Cost Air Quality Sensors and Comparing HRRR Forecast vs Analysis Inputs in a High-Resolution Model for Prescribed Burn Applications |
|---|---|
| Genre | Manuscript |
| Author(s) | Emily Nigro, Shima Shams, M. Williams, A. Piña |
| Abstract | Prescribed burns are implemented by land managers to promote ecosystem resilience and reduce wildfire risk, but the smoke they produce can pose health risks to fire personnel and nearby communities. Accurate prediction of smoke dispersion is therefore critical for planning and management decisions aimed at minimizing exposure. Here, we investigated the application of low-cost air quality sensors and a high-resolution regional weather model to improve prescribed burn smoke detection and forecasting. We deployed three 3D-Printed Automatic Weather Stations (3D-PAWS) in the Oconee National Forest, GA, in spring 2026 to measure meteorological conditions and air quality during prescribed burns. Additionally, we used High-Resolution Rapid Refresh (HRRR) model data to investigate how wind conditions influenced sensor measurements and smoke transport. Finally, we analyzed high-resolution large-eddy simulations within the Weather Research and Forecasting Model coupled with chemistry (WRF-LES-Chem) of the prescribed burn on April 9, 2026, using either HRRR analysis data or HRRR forecast data as the meteorological input. We found that 3D-PAWS air quality sensors captured smoke enhancement and smoldering when deployed adjacent to a burn unit. HRRR captured the evolving wind conditions that controlled smoke transport, and WRF-LES-Chem simulations, based on HRRR analysis data, reasonably reproduced the observed timing and spatial extent of smoke dispersion. Simulations based on HRRR 48-hour forecasts produced somewhat similar pollutant magnitudes but did not capture the same spatial patterns; differences between the two model experiments highlight the sensitivity of smoke projections to meteorological forecasts. This work demonstrates the potential of using low-cost sensor observations and a high-resolution regional model as complementary tools to improve understanding of prescribed fire smoke behavior and ultimately support more effective smoke management. |
| Publication Title | |
| Publication Date | Aug 1, 2026 |
| Publisher's Version of Record | https://doi.org/10.5065/6mx6-km36 |
| OpenSky Citable URL | https://n2t.net/ark:/85065/d7sq9500 |
| OpenSky Listing | View on OpenSky |
| RAL Affiliations | NSAP |