About

Michael Hannigan is a professor in the Paul M. Rady Mechanical Engineering department at the University of Colorado Boulder. His research interests include air pollution and air quality. He is associated with the College of Engineering and Applied Science and has a laboratory at ECSL 119. His contact email is michael.hannigan@colorado.edu, and his office is located at ECOT 235. The page indicates his involvement in research related to air quality and pollution, but does not provide further details about his background, key contributions, or specific projects.

Research topics

• Computer Science
• Chemistry
• Environmental chemistry
• Environmental science
• Photochemistry
• Engineering
• Geology
• Waste management
• Organic chemistry
• Materials science
• Meteorology
• Atmospheric sciences

Selected publications

• Low-cost sensor data for VOC and BTEX calibration

  Mendeley Data · 2026-03-27

  datasetOpen accessSenior author

  Uncalibrated and calibrated sensor data used in Frischmon, et. al. "Evaluating machine learning model performance in a two-step colocation process for TVOC and BTEX sensor calibration." The uncalibrated data contains all sensor inputs used to build the calibration models. The calibrated data contains the concentrations predicted from the calibration models. Please note the data was time-averaged (median) to 15-min intervals prior to publication. The first hour of data collection was also removed prior to publication, as the metal oxide sensors require a warm-up period to reach operating temperatures. File name key: {modelname}_colocation_calibrated: The predicted concentrations by a given calibration model during colocation colocation_raw: The raw secondary standard sensor data collected during colocation harmonization_secondary-standard: The raw secondary standard sensor data collected during harmonization harmonization_harmonized: The harmonized sensor data collected during harmonization by field sensors harmonized_raw: The raw sensor data collected during harmonization by field sensors {model_name}_field_harmonized-and-calibrated: The predicted field concentrations by a given calibration model field_raw: The raw sensor data collected in the field Units: VOC concentration- ppb BTEX concentration- ppb NO2 concentration- ppb Uncalibrated sensor data does not have units, as it is an uncalibrated sensor signal.

  PublisherDOI

• Using Low-Cost Sensors for Fenceline Monitoring to Measure Emissions from Prescribed Fires

  Sensors · 2026-01-22

  articleOpen accessSenior author

  Prescribed burning is a highly effective way to reduce wildfire risk; however, prescribed fires release harmful pollutants. Quantifying emissions from prescribed fires is valuable for atmospheric modeling and understanding impacts on nearby communities. Emissions are commonly reported as emission factors, which are traditionally calculated cumulatively over an entire combustion event. However, cumulative emission factors do not capture variability in emissions throughout a combustion event. Reliable emission factor calculations require knowledge of the state of the plume, which is unavailable when equipment is deployed for multiple days. In this study, we evaluated two different methods used to detect prescribed fire plumes: the event detection algorithm and a random forest model. Results show that the random forest model outperformed the event detection algorithm, with a detection accuracy of 61% and a 3% false positive rate, compared to 51% accuracy and a 31% false positive rate for the event detection algorithm. Overall, the random forest model provides more robust emission factor calculations and a promising framework for plume detection on future prescribed fires. This work provides a unique approach to fenceline monitoring, as it is one of the only projects to our knowledge using fenceline monitoring to measure emissions from prescribed fire plumes.

  PublisherOA PDFDOI

• Comment on egusphere-2025-4697

  2026-03-05

  peer-reviewOpen accessSenior author

  <strong class="journal-contentHeaderColor">Abstract.</strong> Calibration of low-cost air quality sensors (LCSs) for total volatile organic compound (TVOC) and benzene, toluene, ethylbenzene, and xylenes (BTEX) quantification remains challenging due to the sensors' cross-sensitivity to temperature and humidity and their tendency to drift over time. In this study, we aimed to improve TVOC and BTEX metal oxide sensor calibration using a two-step colocation strategy. This strategy made it possible to develop the calibration model under environmental conditions closely matching those of the field, which is essential for model transferability from colocation to field conditions. The approach also addressed intra-sensor variability and drift in the harmonization step. In addition to TVOC and BTEX, we applied the two-step colocation process to nitrogen dioxide (NO₂) electrochemical sensors to demonstrate the broader applicability of our approach beyond TVOC and BTEX quantification. Next, we compared the performance of multiple machine learning models, including ridge, lasso, random forest, gradient boosting, extreme gradient boosting, support vector regression, and linear regression, to investigate the optimal model choice for calibration. We found that no single model performed best across all pollutants. For example, gradient boosting excelled at capturing peak TVOC concentrations, while linear regression performed best for BTEX. Conversely, linear regression was the worst-performing model for NO₂. Overall, the models showed satisfactory RMSE around 40&ndash;50 ppb for TVOC, 1.25&ndash;1.75 ppb for BTEX, and 4&ndash;6 ppb for NO₂. However, all models also overestimated baseline concentrations and underestimated peaks. The severity of this bias depended on the reference concentration distribution, with the most severe peak underestimation occurring in the more heavily skewed TVOC and BTEX data. The systematic bias at baseline and peak concentrations was not evident in the overall mean bias error, which was near zero for all pollutants. This result underscores the need to evaluate model performance across the entire concentration distribution. Finally, we found that calibration performance was sensitive to the choice of training and testing data split. Future research could seek to optimize the training and testing split to ensure robust model transferability to field data.

  PublisherDOI

• Low-cost sensor data for VOC and BTEX calibration

  Mendeley Data · 2026-03-27

  datasetOpen accessSenior author

  Uncalibrated and calibrated sensor data used in Frischmon, et. al. "Evaluating machine learning model performance in a two-step colocation process for TVOC and BTEX sensor calibration." The uncalibrated data contains all sensor inputs used to build the calibration models. The calibrated data contains the concentrations predicted from the calibration models. Please note the data was time-averaged (median) to 15-min intervals prior to publication. The first hour of data collection was also removed prior to publication, as the metal oxide sensors require a warm-up period to reach operating temperatures. File name key: {modelname}_colocation_calibrated: The predicted concentrations by a given calibration model during colocation colocation_raw: The raw secondary standard sensor data collected during colocation harmonization_secondary-standard: The raw secondary standard sensor data collected during harmonization harmonization_harmonized: The harmonized sensor data collected during harmonization by field sensors harmonized_raw: The raw sensor data collected during harmonization by field sensors {model_name}_field_harmonized-and-calibrated: The predicted field concentrations by a given calibration model field_raw: The raw sensor data collected in the field Units: VOC concentration- ppb BTEX concentration- ppb NO2 concentration- ppb Uncalibrated sensor data does not have units, as it is an uncalibrated sensor signal.

  PublisherDOI

• Mentorship and Community in Engineering: a Case Study of Undergraduate Engagement in Research

  2025-04-22

  article

  Engaging undergraduate students in research is a critical responsibility of institutions of higher learning; a respon-sibility which has innumerable positive impacts on students and faculty alike. To effectively fulfill this responsibility, it is critical for experienced researchers to consider ways to engage under-graduate students in research fields. One established method of increasing undergraduate engagement is by pairing research mentors with undergraduate mentees. However, this approach requires a substantial time commitment from both the mentor and mentee and often impacts only one student at a time. A review of current literature suggests that to combat these pitfalls, experienced researchers could foster a cohort of undergraduate students in a research setting, as this may also effectively increase undergraduate's involvement in research. To further examine both of the aforementioned approaches, this work outlines a framework comprised of a model of mentorship and a set of adoptable practices for experienced research mentors to use when mentoring undergraduates. Additionally, this work examines a preliminary case study in which a graduate student employed this framework in the setting of an engineering research lab. In this instance, a graduate student acts as a mentor to a cohort of three undergraduate student mentees. To conduct preliminary assessment, autoethnographic reflections were collected from participants to be coded and analyzed for themes. In addition, any resources associated with implementation of the mentorship framework was also saved for illustration of the proposed prac-tices associated with the framework. This assessment has yielded promising results, as all three undergraduate student mentees have remained in the lab as active and effective team members. Furthermore, all three students felt that this experience positively impacted their personal and professional growth. In analyzing student reflections, these undergraduate students have reported an increased level of confidence in their technical abilities and a recognition of their importance to the community of engineering research.

  PublisherDOI

• Supplementary material to "Evaluating machine learning model performance in a two-step colocation process for TVOC and BTEX sensor calibration"

  2025-10-29

  articleOpen accessSenior author
  PublisherDOI

• Improving the quantification of peak concentrations for air quality sensors via data weighting

  Atmospheric measurement techniques · 2025-07-15

  articleOpen accessSenior authorCorresponding

  Abstract. Traditional calibration models for low-cost air quality sensors have demonstrated a tendency to underpredict peak concentrations. We assessed the utility of adding data weights to low-cost sensor colocation data to improve the quantification of peak concentrations when the majority of colocation data is at a baseline concentration and varies due to intermittent, transient events. Specifically, we explore the effects of data weighting on three different pollutant colocation datasets: total volatile organic compounds (VOCs), carbon monoxide (CO), and methane (CH4). Leveraging two different weighting functions, a sigmoidal and a piecewise weighting regime, we explored the impacts of the base model choice (multilinear regression, MLR, vs. random forest, RF, models), the sensitivity of weighting functions, and the ability of data weighting to improve high-concentration pollution measurements. When compared to unweighted colocation data, we demonstrate significant reductions in both error (root mean square error, RMSE) and bias (mean bias error, MBE) for pollutant peaks across all three datasets when data weighting is employed. For the top percentile of data, we observe an average of 23 % reduction in RMSE and a 35 % reduction in MBE when optimal weights are employed. More significant reductions occurred in the 95th–99th percentile of data, where MBE was reduced by an average of 70 %. RMSE in the 95th-99th percentile was reduced by an average of 26 %. However, data weighting can also generate larger errors at baseline pollutant concentrations. Data weighting regimes were sensitive to input parameters, and input weighting functions may be tuned to better predict peak concentration data without significant reductions in the fidelity of baseline pollutant predictions.

  PublisherOA PDFDOI

• Using terrestrial laser scanning to estimate mass of hand-built slash piles following hazardous fuels treatments

  Frontiers in Forests and Global Change · 2025-10-17

  articleOpen accessSenior author

  Pile burning is increasingly used in many forest and woodland ecosystems to reduce hazardous fuel loads following fuel hazard reduction or forest restoration efforts. Pile burning is often linked to thinning practices where residual fuel is piled and subsequently burned; the burning is typically done in winter months when conditions reduce the risk of unwanted fire behavior such as escapes. A key aspect of pile burning is estimating the amount of pile biomass and the amount of fuel consumed during burning as these two variables are critical for estimating treatment efficacy and smoke emissions. Methods to estimate pile masses have been studied and developed previously, however, they are time consuming and require extensive user training. Terrestrial laser scanning (TLS) is a remote sensing tool that has been successfully used on broadcast burning for fuel characterization and has the potential to estimate pile masses at prescribed burning sites. TLS reduces measurement error, requires less extensive user training, and eliminates observer bias in measurements. A total of 16 pile masses were measured across Colorado, United States, using a previously developed pile measurement methodology, using TLS, and by taking apart the pile and weighing the contents of the pile, to determine if TLS would be an adequate method for predicting pile masses. Individually, TLS did not do a good job predicting pile masses, however, when comparing across all 15 piles, using three TLS scans of a pile to estimate pile mass had the lowest median percent error across all piles.

  PublisherOA PDFDOI

• Detecting air pollution episodes and exploring their impacts using low-cost sensor data and simultaneous community symptom and odor reports

  Environmental Research Letters · 2025-03-19 · 4 citations

  articleOpen access

  Abstract As a team of community organizers and academic researchers, we conducted a community-based participatory exploration of industrial pollution impacts in Cherokee Forest, a fenceline community adjacent to an industrial park in Pascagoula, Mississippi. Using a derivative-based episode detection algorithm with low-cost uncalibrated sensor signal data sensitive to VOCs, ammonia, amine-series, and sulfurous odors, we identified frequent and intense pollution episodes within the community. According to wind data, these episodes came from the direction of the industrial park and often correlated with increased symptom and odor reports. Additionally, metals biomarker toenail sampling revealed elevated nickel levels in a subset of resident children, which is an industrial pollutant of concern in this community. The findings have supported Cherokee Concerned Citizens’ advocacy efforts to mobilize the community and engage with regulatory agencies. Our work demonstrates a transferable methodology for using low-cost sensors and community reports to document industrial pollution impacts in fenceline communities.

  PublisherDOI

• Calibration and Validation of an Autonomous, Novel, Low-Cost, Dynamic Flux Chamber for Measuring Landfill Methane Emissions

  Sensors · 2025-10-28

  articleOpen access

  A low-cost, dynamic flux chamber optimized for landfill emissions measurement was designed, fabricated, calibrated, and validated for measurements of methane flux ranging from 0 to 150 g/m2-day. A centrifugal blower fan and a flow meter were plumbed in series to draw a bypass flow through the flux chamber. Both ambient and chamber methane concentrations were measured using the arrays of four low-cost metal oxide sensors. Leveraging the sensors’ overlapping sensitivity to changes in methane concentration, temperature, and humidity, multiple linear regressions were trained on laboratory data and combined into a piecewise methane calibration function. An algorithm was developed to select the most useful interaction terms among all sensor responses to optimize the predictors in each model. The piecewise regions for methane measurement were 0–100 ppm, 100–1500 ppm, and 1500–12,000 ppm. The root mean squared errors for each piecewise region were 3.1 ppm, 21 ppm, and 307 ppm, respectively. Controlled quantities of methane were delivered to the flux chamber in a laboratory setting for validation. Measurements yielded good agreement with an RMSE and MBE of 7.3 g m−2 d−1 and 2.2 g m−2 d−1, respectively. The flux chamber was tested at a closed landfill to validate its ability to autonomously and continuously operate in the field.

  PublisherDOI

Recent grants

• Collaborative Research: Impact of Metals on Photochemical Aging of Water Soluble Organic Carbon in Atmospheric Particulate Matter: A Combined Lab and Field Study

  NSF · $254k · 2016–2019

• CPS: TTP Option: Synergy: Collaborative Research: Calibration of Personal Air Quality Sensors in the Field - Coping with Noise and Extending Capabilities

  NSF · $290k · 2015–2019

Frequent coauthors

• Ricardo Piedrahita

  Berkeley Air Monitoring Group (United States)

  36 shared

• Jana B. Milford

  University of Colorado Boulder

  35 shared

• Christine Wiedinmyer

  Cooperative Institute for Research in Environmental Sciences

  35 shared

• Steven J. Dutton

  33 shared

• Evan Coffey

  University of Colorado Boulder

  31 shared

• Katherine L. Dickinson

  29 shared

• Mingjie Xie

  Nanjing University of Information Science and Technology

  24 shared

• Sarah G. Riddle

  University of California, Davis

  23 shared

Education

• PhD, Environmental Engineering Science

  California Institute of Technology

  1997

• MS, Environmental Engineering Science

  California Institute of Technology

  1991

• BS, Civil Engineering

  Southern Methodist University

  1990