Line profile reconstruction: validation and comparison of reconstruction methods

Citation
My. Tsai et al., Line profile reconstruction: validation and comparison of reconstruction methods, ATMOS ENVIR, 35(28), 2001, pp. 4791-4799
Citations number
18
Categorie Soggetti
Environment/Ecology,"Earth Sciences
Journal title
ATMOSPHERIC ENVIRONMENT
ISSN journal
13522310 → ACNP
Volume
35
Issue
28
Year of publication
2001
Pages
4791 - 4799
Database
ISI
SICI code
1352-2310(200110)35:28<4791:LPRVAC>2.0.ZU;2-8
Abstract
Currently, open path Fourier transform infrared (OP-FTIR) spectrometers hav e been applied in some fenceline monitoring, but their use has been limited because path-integrated concentration measurements typically only provide an estimate of the average concentration. We present a series of experiment s that further explore the use of path-integrated measurements to reconstru ct various pollutant distributions along a linear path. Our experiments wer e conducted in a ventilation chamber using an OP-FTIR instrument to monitor a tracer-gas release over a fenceline configuration. These experiments val idate a line profile method (1-D reconstruction). Additionally, we expand c urrent reconstruction techniques by applying the Bootstrap to our measureme nts. We compared our reconstruction results to our point samplers using the concordance correlation factor (CCF). Of the four different release types, three were successfully reconstructed with CCFs greater than 0.9. The-diff icult reconstruction involved a narrow release where the pollutant was limi ted to one segment of the segmented beampath. In general, of the three reco nstruction methods employed, the average of the bootstrapped reconstruction s was found to have the highest CCFs when compared to the point samplers. F urthermore, the bootstrap method was the most flexible and allowed a determ ination of the uncertainty surrounding our reconstructions. (C) 2001 Elsevi er Science Ltd. All rights reserved.