Aaron Weiner

Abstract

The remote sensing gas detection problem is one with no straightforward solution. While success has been achieved in detecting and identifying gases released from industrial stacks and other large plumes, the fugituve gas detection problem is far more complex. Fugitive gas represents a far smaller target and may be generated by leaking pipes, vents, or small scale chemical production. The nature of fugitive gas emission is such that one has no foreknowledge of the location, quantity, or transient rate of the targeted effluent which requires one to cover a broad area with high sensitivity. In such a scenario, a mobile airborne platform would be a likely candidate. Further, the spectrometer used for gas detection should be capable of rapid scan rates to prevent spatial and spectral smearing, while maintaining high resolution to aid in species identification. Often, insufficient signal to noise (SNR) prevents spectrometers from delivering useful results under such conditions. While common dispersive element spectrometers (DES) suffer from decreasing SNR with increasing spectral dispersion, Fourier Transform Spectrometers (FTS) generally do not and would seemingly be an ideal choice for such an application. FTS are ubiquitous in chemical laboratories and in use as ground based spectrometers, but have not become as pervasive in mobile applications. While FTS spectrometers would otherwise be ideal for high resolution rapid scanning in search of gaseous effluents, when conducted via a mobile platform the process of optical interferogram formation to form spectra is corrupted when the input signal is temporally unstable. This work seeks to explore the tradespace of an airborne Michelson based FTS in terms of modeling and characterizing the performance degradation over a variety of environmental and optical parameters. The major variables modeled and examined include: maximum optical path distance (resolution), scan rate, platform velocity, altitude, atmospheric and background emissivity variability, gas target parameters such as temperature, concentration-pathlength, confuser gas presence, and optical effects including apodization effects, single and double-sided interferograms, internal mirror positional accuracy errors, and primary mirror jitter effects. It is through an understanding of how each of the aforementioned variables impacts the gas detection performance that one can constrain design parameters in developing and engineering an FTS suitable to the airborne environment. The instrument model was compared to output from ground-based FTS instruments as well as airborne data taken from the Airborne Hyperspectral Imager (AHI) and found to be in good agreement. Monte Carlo studies were used to map the impact of the performance variables and unique detection algorithms, based on common detection scores, were used to quantify performance degradation. Scene-based scenarios were employed to evaluate performance of a scanning FTS under variable and complex conditions. It was found that despite critical sampling errors and rapidly varying radiance signals, while losing the ability to reproduce a radiometrically accurate spectrum, an FTS offered the unique ability to reproduce spectral evidence of a gas in scenarios where a dispersive element spectrometer (DES) might not.

Library of Congress Subject Headings

Spectrometer--Design and construction--Evaluation; Fourier transform infrared specroscopy; Gas-detectors--Evaluation; Remote sensing--Equipment and supplies--Evaluation; Imaging systems--Image quality--Evaluation

Publication Date

7-1-2010

Document Type

Dissertation

Student Type

Graduate

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)

Advisor

Messinger, David

Comments

Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works. Physical copy available through RIT's The Wallace Library at: QC373.S7 W45 2010

Campus

RIT – Main Campus

Download

Share

COinS