The data and reports which we received for February '98 can be characterized as being fairly uneventful, with generally less recorded emissions activity and generally better instrument performance than we have seen in preceeding months.
Both FTIR units operated well throughout the month. The "on-stream efficiency" for the North FTIR was better than 96%. The on stream percentage for the South unit was only 89% due to a problem with an older computer's hard drive.
Reported detection-limit figures are higher than they have been in the preceeding months. Because we have been unable to update our "background reference" spectra in the analytical routines, we expect to see detection limits erode like this over time. We do have indications, however, that not all of the problem with detection limits can be attributed to our dated background files. Overall, the quality of the raw spectral data being produced by the two FTIR monitors is good for the month of February '98, although clearly not as good as we had seen in previous months.
Background reference spectra
On the north fenceline, spectral file bkg00307 (from 9/6/97 2:04pm) was used as the background reference for all the absorbance files. On the South fenceline, spectral file # bks 00727 (11/13/97 10:01am) was used. We have been using these same files for several months now, and considering their vinatge, they seem to files have held up very well. To date, we have yet to find a replacement or updated background reference spectrum which will outperform them over the period of a month.
As we try to do with each month's data set, several of the more promising single-beam spectral files from February were converted for background use in attempts to find a newer background reference spectrum to replace the older background files. Again in February, none of the "background candidates" which were considered was found to be capable of producing absorbance spectra of high enough quality to provide for an improvement detection limits over what we have been seeing with our existing background reference files.
It should be stated that while we did examine several of what seemed to be very promising background reference "candidate" spectra, we did not undergo the type of systematic and time-consuming search for an appropriate background file that we have in past months. It is therefore very possible that there exists within this February '98 data set one or more FTIR spectral files which - if used as a background reference file for the rest of the month - could provide for lower overall detection limits than what we are reporting here.
Even if there isn't any one FTIR spectrum which could effectively replace our existing background file, there is little doubt that the detection limits which we are reporting could be improved upon to some degree by utilizing a string of contemporary "background" files from within the data set.
For the time being, we plan to continue to use these more or less "tried and true" background files, even though the detection limits are beginning to erode, unitl we can find something better to replace them with.
It is interesting and encouraging to note that there was no significant weather-related downtime for either FTIR unit in February, and this seems to be the case month after month. While these FTIR units continue to have problems with the cryocoolers and various software glitches, it is very encouraging to see that these units are capable of providing continuous real-time monitoring even during periods of heavy fog and rain.