The Hannigan Lab has just received news that they can support two CU Engineering undergraduate students on NSF Research Experience for Undergraduate paid positions for the summer. Below is the detail about the projects, which might sound a little dry ... but, the critical piece of information for students interested is that you would get to work with a research team to better understand how air sensors can be used to improve human health. Students will be looking at lots of sensor data and using Matlab to better understand what the sensors are telling us. If you are interested, please send your resume to Professor Mike Hannigan.
The two students will be tasked with developing and implementing sensor data quality control and assurance plans for both ozone and particulate matter air pollution data that has previously been collected in South Los Angeles. The students will work with current project graduate students (Ashley Collier-Oxandale, Duanfeng Gao, Tony Zhang, Evan Coffey) and the project PIs (Mike Hannigan, Qin Lv, Daven Henze, and Rob Dick) to ensure that the output from their work could be used in publications planned for Fall 2018. The students will be folded into the existing CU College of Engineering and Applied Sciences Summer Program for Undergraduate Research (SPUR) for additional mentoring as well as a platform for presenting their project results. Specific details for each project are below. We do anticipate that each project will have a lead REU student but that the two REUs student will work as a team and play off each other’s strengths. Ideally, these REU students will have completed 2+ years of course work in Engineering. Students
Ozone Sensor Project. Our CyberSEES research team has developed an approach for down-scaling ground level ozone concentrations generated with an Atmospheric Chemical Transport Model (CTM) to a spatial and temporal scale that is more appropriate for people than the model can accurately generate. The approach for this down scaling is to use open and available cyber data like traffic counts, population demographics, road density, key industrial source locations, temperature, humidity and wind in a statistical model to generate time and space variability patterns around a concentration mean that was generated from the CTM. We have developed the statistical model using regulatory site data from the South Coast Air Quality Management District. We would like to refine this statistical model using ozone concentration data from a finer spatial scale. One of the project co-PIs (Mike Hannigan) placed Pod sensor systems at an improved spatial scale (20 in a neighborhood) for a different project but in the region of interest in Los Angeles. The goal of that deployment was the assessment of the methane concentration but an ozone sensor was including in the Pod sensing system. The REU student will be tasked with moving the ozone sensor data from raw signal to ambient concentration with associated uncertainty estimates. This work would be similar to the ozone sensor work recently published as part of the NSF grant (Sadighi et al, Atmospheric Measurement Techniques, 2018). As such, the REU student will be able to start with the process used for this past work. We would anticipate the the REU student will also make advances in automation of the ozone sensor data processing to enable future ozone field campaigns to more effectively employ low cost sensors in experimental design.
PM Sensor Project. Our CyberSEES team has recently started evaluating our ability to down scale a second, and maybe even more important for public health, pollutant … particulate matter (PM). Over the past year, there have been a proliferation of low cost PM sensors deployed by citizens across the globe. The most notable might be the Purple Air sensors which have an open data platform that is showing live PM readings from all Purple Air sensors. At present, there are more than 100 Purple Air sensors in the Los Angeles area. The PM2.5 readings from this sensors have not undergone quality control and assurance. There are a dozen regulatory PM2.5 monitors in the same region. These two data streams need to be merged to improve the utility of the PM2.5 sensor data. In addition, our research team is interested in developing down scaling using visibility degradation from on-line web cameras as well. The REU student will be first tasked with merging the sensor data with the regulatory data and exploring algorithms and process for merging those two. The REU student will then work with the project graduate student to enable the development of a visibility to PM2.5 down-scaling approach, which will involve computer vision/image processing.