This test has significantly improved our understanding, confidence and capability with this unique, system. Clearly, the System's performance was much improved over that demonstrated during the first test in Ft. Morgan, CO, as we continue to significantly further its development. In addition, the March 99 test included the first very long path measurements for propane. Moreover, it included the first detection of atmospheric ethane by this instrument. (That is, ethane not contained within a sampling tube or induced by known source). Finally, as a demonstration of the System's spectral performance, the March 99 test included the first detection of an unknown trace gas. Further research will be required to uniquely identify this trace gas, but its spectral location near that of ethane suggests that it may he a hydrocarbon.

Summary

Chimera Geophysics and OPHIR Corporation performed the second field test of the Hydrocarbon Differential Absorption Lidar (DIAL) System between T through 14 March 1999. Two different test sites were explored, both near Rockdale Texas. Both sites included multiple trailer-to-reflector configurations. One site included two trailer locations, The test included a total of nine unique trailer/reflector geometries. The most significant findings are summarized below:

1.	The trailer-mounted system was transported from Littleton CO to Rockdale Texas without damaging any system components. Moreover, the trailer system was dissassembled and moved on three different incidences without damage to an, component. The trailer was moved through field terrain on all three occasions. The System is comprised of expensive and custom-designed electronics and optics. For some items, the procurement lead"time is very long (manymonths). These activities have demonstrated that the System can be readily transported, given a reasonable amount of precaution for traveling and safe-handling. In turn, these activities have demonstrated that the shock-isolation system currently used on (and designed explicitly for) the Hydrocarbon DIAL System functions well. The successful demonstration of the portability of the System Is a significant project accomplishment.
2.	Multiple test-geometries and multiple trailer-locations were performed in a single day. Trailer-to-reflector ranges of approximately 1 mile (2 mile optical path) were readily accomplished. System set-up after trailer relocation was demonstrated to be relatively simple and rapid. Realignment to new reflector locations was surprisingly simple, rapid and stable through time. The Test Crew had minimal previous opportunities to evaluate or practice multiple-geometry set-ups. The use of a hand-held (and independent) GPS locator proved to be invaluable to those operating the retro-reflector, Still, the first demonstration of multiple reflector geometries and multiple trailer locations was very successful. The Interim Effort activities to improve the laser alignment stability improved the system significantly. During the March 99 Field test, laser alignment never became a major required activity beyond daily maintenance. The Improved laser alignment stability and the demonstration of multiple test geometries within a 24-hour period are significant project accomplishments.
3.	No component failures were encountered. Although the test was at a short duration, the absences of hardware and system failures are a solid sign of the system's reliability. Further use and testing will be required to fully assess reliability issues. Still, these activities establish a base-level for reliability performance.
4.	Methane, ethane and propane concentrations were monitored on a routine basis. Measurements of atmospheric background methane concentrations were demonstrated for a second time. Using a contained sampling chamber (test tube) located nearly a mile from the trailer, demonstrations of the System's capability to measure ethane were completed for a second time, at a level roughly equivalent to a 1 ppm background, (First demonstrations occurred during the field test in Ft. Morgan CO). These tests demonstrated that the Hydrocarbon DIAL hardware was performing at (or above) its anticipated performance level.
5.	The quality of the spectroscopic data was quite high, even in moderate wind and relatively high humidity (water absorption interference). The laser wavelength stability and spectral purity were good, Shot-to-shot laser spectral purity was markedly better than during the Ft. Morgan field trials, but further improvements in this arena would be a significant benefit. As an indication of the System's spectral purity, the identification of an unknown trace-gas species is currently being limited by the available spectral databases (rather than the laser hardware itself). Known spectral features trave slightly different spectral width, depth and location when compared to HITRAN and the HANST databases. The DIAL's spectral performance is sufficiently good to identify inconsistencies in these databases. As a second indication of the System's spectral performance, an unknown trace species was detected along with atmospheric ethane. A significant effort has been invested to identify this competitive trace species. Through investigations, we are convinced the species is not water vapor, nor any species currently cataloged in the U.S. Air Force Phillips Laboratories' HITRAN database. The HANST database does not possess sufficient spectral accuracy or spectral quality to positively identify this species. The most-likely candidate species would be another hydrocarbon. However, positive identification of this trace gas cannot be completed without additional research. One potential method would be to develop an independent database using the DIAL itself. The System's demonstrated spectral purity, along with the separation of an unknown trace gas from atmospheric ethene, is a significant project accomplishment.
6.	On two days (11 and 13 March), at site T3R2 (trailer site 3, retro-reflector site 2), atmospheric ethane was detected. The measured concentration averaged approximately 40 ppm on 11 March, and 19 ppb on 13 March. The 1976 U.S. Standard Atmosphere background concentration for ethane is 2 ppb. 2 ppb ethane is too small to be measured by the Hydrocarbon DIAL System operating at a 1 mile trailer-to-reflector range. 40 ppm ethane is very close to the DIAL System's current detection limit threshold (roughly 10 to 20 ppb ethane under the current configuration and ideal meteorological conditions). Data analysis performed subsequent to the collection of the field data have verified that, indeed, ethane was detected, although in minute concentrations. An unknown trace gas was also detected, as discussed in item 5 above. The Syetem's demonstrated capability to identify atmospheric ethane at average Concentrations of approximately 40 ppm is a significant project accomplishment.
7.	Atmospheric propane was never detected at any site during the March 99 Field Test. The MDC for propane is approximately 1 ppm, limited by spectral interferences with methane and water.
8.	Due to numerous factors, the Minimum Detectable Concentration (MDC) varies between measurements. The primary meteorological influences on the MDC are wind and atmospheric turbulence. The MDC also varies with optical pathlength and sampling time, Finally,!he MDC changes significantly with the absorption feature and the trace species under study. Due to these varying factors encountered on each day, the MDC for ethane varied from roughly 10 to 130 ppb, significantly more sensitive than the design goal of 1 ppm. For this reason, ethane may have existed at other sites as well. The MDC was quite low at T3R3 (less than 11 ppb), and no ethane concentration above this level was detected, in contrast to T3R2 where 18 to 77 ppb were detected.
9.	Methane was measured at each site. The ambient background level of methane for a 1976 U.S. Standard Atmosphere is 1.7 ppm. The average values measured with the DIAL System- throughout the Field Test was 1.89 ppm, indicating good agreement with a background concentration. Measurement uncertainty is below the background level, and approximately' 1 ppm. The measured atmospheric methane concentration varied markedly at each test site and from day-to-day. The mechanism that results in this variability i's currently unknown. Potential mechanisms include variable vertical migration rates, variable atmospheric conditions, variable foliage and soil types from site-to-site, and variable microbial activity. It is interesting to note that the smallest methane concentrations were all measured after a large rainstorm, which occurred on 12 March.
10.	A summary of concentration measurements is presented in Table 1.

The Hydrocarbon System was shown to be very reliable and capable of being readily moved and set up again. Moving the trailer required approximately one-half day to break-down and re-set up. Moving the retro-reflector required approximately one hour The Laser, Data Acquisition, Laser Control, Optical Train, and Health Checks were found to work well.

The alignment instabilities experienced during the June 1998 field demonstration had been completely eliminated. The single-ended telescope alignment procedure developed in response to the June (Ft. Morgan) 1998 field test was found to work very well. No freezing weather was experienced, however the thermostat-controlled propane heater, which was installed in December 1998, maintained the trailer temperature through the night and probably contributed to the improved laser alignment stability. We were able to search for propane in the field because of calibration and testing performed at OPHIR in 1-2/1999.

High humidity levels caused the laser to form condensation on the laser rod faces. This will increase thermal stress on the laser rods possibly reducing their lifetime and/or inducing optical damage. To prevent this in the future the laser rod region will have to be nitrogen purged if operated while the dewpoint is above 18°C. To be able to purge will require that we implement the "ramp and fire" of the laser seeding (see below) to ensure good seeding.

High wind gusts (approximately >15 mph) were found to introduce significant amplitude noise in the data due to flexing of the trailer/goniometer mirror moving the laser beam off of the retro-reflector. The system operated properly but the data quality decreased significantly due to these meteorological effects.

The system is still operating at a prototype level of development with no spares. Equipment failure of some key components would shutdown field operations for weeks awaiting repair or replacement. Debugging of hardware and data acquisition and control software is still occurring, as well as data analysis software. Operating procedures are still being developed and changes will have to be made in response to results of the data analysis. Different hydrocarbon lines will be investigated for better interference rejection. Larger numbers of spectral scars will be done at each site in order to be able to reduce the noise and ensure there are no unknown interfering gases affecting the results. The reference cell will have to be mounted so that reference cell scans can be taken simultaneously with air scans for all data. This will also help in rejecting interferences.

The trailer GPS system worked properly and gave accurate readings of the differential path length from trailer to retro-reflector. The absolute loca'.ion readings were accurate in latitude/longitude. Communication to with the trailer to get a reading to provide for data recording functioned as anticipated. A handheld GPS unit was purchased while in the field and found to be irreplaceable, due to the speed at which it enabled the retro-reflector operator to locate the test point position.

Although much improved over all previous systems (and sufficient to illuminate inconsistencies with the HITRAN and HANST databases), the DIAL System's spectral purity of the laser should, be improved on. This will help for a number of reasons. First, automated concentration measurements will be more readily performed without the need for operator intervention. This should increase the amount of "good" data and so will require less averaging time to achieve a good measurement. This will assist in rejecting interferences from other gases. Improving on the laser spectral purity should be accomplished with the injection seeded laser using the "ramp-and-fire" method. In this the diode laser wavelength is swept slightly (<1 GHz) as the laser is Q-switched. This ensures that there is a maximum possible amount of diode laser seed power in the oscillator cavity when the oscillator fires. This will cause no significant smearing of the laser lineshape but instead should produce a cleaner more reproducible lineshape. To do this will require a fairly simple circuit be constructed and tested on the system, in order to optimize the amplitude and speed of the sweep and ultimately the laser results. This will also allow us to begin purging the laser with dry nitrogen to eliminate humidity.

The Minimum Detectable Concentration (MDC) is a way to characterize our ability to detect very low levels of a gas( e.g. ethane and propane). It is not useful for gases which have a large ambient level (e.g. methane) where a different measure is needed. To understand what 'actors affect the MDC a functional relationship is given by:

This is not meant to be exact but to illustrate what is important. It shows that we can detect lower concentrations of gases, in general, if we 1) use longer path lengths (assuming uniform gas concentrations), 2) use stronger absorption lines of those available that do not saturate, and 3) average for longer times which reduces the effect of noise. Noise in the system, for example due to laser spectral purity problems or amplitude noise, detector noise, or turbulence noise, make us less sensitive therefore increasing the MDC.

If hydrocarbons are pulsing out of the ground and dissipating quickly this would be very difficult for the DIAL System to distinguish from turbulence (including wind gusts), depending on the exact time scale of the pulsing. This is because we cannot distinguish between large intermittent gas absorption and laser beam wander produced by turbulence or wind causing the beam to partially (Or completely) miss the retro-reflector. In addition, our ability to improve our sensitivity (MDC) would be degraded because pulsing gas concentration measurements would not improve by averaging longer. There are possible ways around this, but a better understanding at the time scales involved is necessary.

Careful evaluation of the field data allowed us to identity which of the logged data parameters were most useful for flagging and throwing out bad data points. The most useful collected currently on every shot was found to be the normalization channel amplitude. When this amplitude is below approximately 60 percent of its mean value, it is likely due to poor seeding or a laser misfire and so that data should be excluded. In the future, having the revere ice cell data for every shot will enhance our ability to throw out bad data significantly. This data flagging of bad data is crucial to achieve successful automation of the data analysis algorithms.

Click images below to view larger image

©1999, Chimera Geophysical Corporation, all rights reserved