Prescribed fires (fires intentionally set for mitigation purposes) produce pollutants, which have negative effects on human and animal health. One of the pollutants produced from fires is fine particulate matter (PM_2.5). PM_2.5 can penetrate deep into the lungs and harm cardiovascular and respiratory systems. The Flint Hills region of Kansas experiences extensive prescribed burning each spring (March - May). Smoke from prescribed fires is often understudied due to a lack of monitoring in the rural regions where prescribed burning occurs, as well as the short duration and small size of the fires. Our goal was to attribute PM_2.5 concentrations to the prescribed burning in the Flint Hills. To determine PM_2.5 increases from local burning, we used low-cost PM_2.5 sensors (PurpleAir) and satellite observations. The Flint Hills were also affected by smoke transported from fires in other regions during 2022. We separated the transported smoke from smoke from fires in eastern Kansas. Based on data from the PurpleAir sensors, we found the 24-hour median PM_2.5 increased by 5.2 µg m^-3 on days impacted by smoke from fires in the eastern Kansas region compared to days unimpacted by smoke. We found the Flint Hills to be the most smoke PM_2.5 impacted region compared to other regions of Kansas, as observed in satellite products and in situ measurements.

We recruited participants to deploy 38 PurpleAir monitors throughout eastern Kansas to study smoke from prescribed burning from March - May 2022. PurpleAir are low-cost sensors (~$300 USD per unit) that estimate PM_2.5 concentrations every two minutes. We performed a quality check on the raw PM_2.5 concentrations from PurpleAir. We took 10-minute averages of all measurements. We then removed data with the following conditions: (1) temperature > 65 ^oC (0.005 % of observations), (2) relative humidity > 100% (0.0001 % of observations), (3) channel disagreement > 10% from the average of the two channels or 10 µg m^-3 in the absolute difference between the channels (3.3% of observations), and (4) measurements > 500 µg m^-3 (0.0017% of observations). We applied the Barkjohn et al. (2021) correction factor to the quality checked PurpleAir field measurements. This correction factor reduces bias in PurpleAir PM_2.5 by scaling based on the concentrations and relative humidity (https://doi.org/10.5194/amt-14-4617-2021). Due to continuously erroneous humidity observations for one sensor (PurpleAir ​​1361083), we used humidity observations from a nearby PurpleAir sensor to calculate the PM_2.5 correction factor. After we applied the correction factor, some observations with high humidity and low concentrations became negative. All negative concentrations after the correction factor were set to 0 µg m^-3 (1.87% of observations). There were two sensors that we could not recover the PM_2.5 measurement from. We are reporting hourly measurements in this dataset.