Thesis
Detection of volatile organic compounds using passive Fourier transform infrared spectrometry and short interferogram segments
University of Iowa
Master of Science (MS), University of Iowa
Autumn 2019
DOI: 10.17077/etd.005215
Abstract
The United States Environmental Protection Agency (USEPA) Airborne Spectral Photometric Environmental Collection Technology (APSECT) program is the nation’s only airborne emergency response program with remote chemical and radiological detection capability. The goal of the ASPECT program is to provide real-time or near-real-time hazard information to first responders so they can respond appropriately to the specific hazard present. The instrumentation in the ASPECT aircraft includes a passive Fourier transform infrared (FTIR) spectrometer that is used for the detection of many volatile organic compounds (VOCs) that can be hazardous to the environment and human health. The high data acquisition rate required for the aircraft-based measurement and the need to provide timely information to first responders dictate that the data interpretation must be automated. This is accomplished through the application of numerical pattern recognition methods to the acquired data. The effectiveness of detection is based upon our ability to develop effective classification models that are specific to each analyte of interest.
In the field, there is constantly changing background information in each scan. Unlike in traditional laboratory-based spectroscopy, we do not have the luxury of taking representative background spectra to use to remove background information from field-collected spectra. For this reason, analysis is performed directly on the FTIR interferograms, rather than on spectra. Selecting a segment from an optimized location offset from the interferogram centerburst is a strategy employed here to remove background information.
This thesis work is focused on the improvement of ground resolution in the FTIR application. Because scans are taken as the aircraft flies, each interferogram point is representative of a slightly different spot on the ground. The ASPECT program has traditionally v implemented classifiers based upon using one segment from one side of each interferogram. Not only does this discard potentially useful information, but it also holds the assumption that what was viewed by the spectrometer at that time was representative of the area viewed throughout the entire scan. This research explores the possibility of using more than one segment from a single interferogram to improve the ground resolution of passive FTIR data.
The work here focused on three analytes: methanol, ethanol, and ammonia. Laboratory data of these three analytes collected to mimic field conditions were used to complete studies of classification error of segments of lengths 25, 50, 75, 100, and 150 points spanning the entire length of each interferogram. This gave us a profile of where information valuable for classification resided in the length of the interferogram for each analyte. Exploration of how the signal-to-noise ratio affects classification error was also explored using these data. The research performed here shows the feasibility of improving ground resolution of FTIR data by using multiple interferogram segments. This methodology, once fully developed, could provide more useful information to first responders in their response to and mitigation of hazardous VOC releases.
Details
- Title: Subtitle
- Detection of volatile organic compounds using passive Fourier transform infrared spectrometry and short interferogram segments
- Creators
- Rebecca Rhea Meredith
- Contributors
- Gary W. Small (Advisor)Renee Cole (Committee Member)Elizabeth Stone (Committee Member)
- Resource Type
- Thesis
- Degree Awarded
- Master of Science (MS), University of Iowa
- Degree in
- Chemistry
- Date degree season
- Autumn 2019
- DOI
- 10.17077/etd.005215
- Publisher
- University of Iowa
- Number of pages
- x, 60 pages
- Copyright
- Copyright 2019 Rebecca Rhea Meredith
- Language
- English
- Description illustrations
- color illustrations
- Description bibliographic
- Includes bibliographical references (pages 58-60).
- Public Abstract (ETD)
The United Stated Environmental Protection Agency (USEPA) Airborne Spectral Photometric Environmental Collection Technology (ASPECT) program aims to provide real-time chemical and radiological hazard information to first-responders so that they may tailor their response to the specific hazards present. To detect volatile organic compounds (VOCs), the ASPECT aircraft is equipped with a downward-facing Fourier transform infrared (FTIR) spectrometer and detects characteristic frequencies being absorbed or emitted by analytes of interest. Typically, this FTIR raw data would be Fourier transformed to obtain spectra, but in this application, it is advantageous to work directly with the interferogram data in its raw form. This provides a convenient way to remove information about the spectral background and offers the potential to preserve the ground location information inherent to the scans.
This thesis presents research that aims to improve the ground resolution of passive FTIR measurements made from an aircraft platform. Since interferograms are digitized time-domain data collected as the aircraft flies, each point of the interferogram is representative of a slightly different point on the ground. Therefore, if multiple segments of a single interferogram are used, better ground resolution can be obtained than if one segment were used to represent an entire interferogram. This better ground resolution would allow for more accurate representation of where the hazard is present, which would be useful to first-responders.
- Academic Unit
- Chemistry
- Record Identifier
- 9983779698502771
Metrics
34 File views/ downloads
68 Record Views