i

Author@):

Submitted to:

Roger R. Petrin CST- 1 John R. Quagliano CST-6 Douglas H. Nelson CST-1 Mark J. Schmitt XCM Charles R. Quick CST-1 Robert K. Sander CST-4

Mike Whitehead CST-6 Joe J. Tiee CST-6

Proceedings of the 1996 international Geoscience and Remote Sensing Symposium, Lincoln, NE May 27-31

Los Alamos National Laboratory. an affirmatbe adionlequal opportunity errpbyer. is operaled by the Univerdty of California for the U.S. Department d Energy under contract W-7405-ENG36. By acceptance of this article, the publisher recognizes that the U.S. Government retains a nonexdusive, IDyaftY-fW UCeflSe to publish or reproduce the published form of this contribution. or to allow others to do so, for U.S. Government purposes. The L w Marnos National Laboratory requests that the publisher Menttfy this article as work performed under the auspices of the US. Department

FormNo.836R5 ST 2629 lag1

DISCLAIMER

Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.

.

Atmospheric Effects on CO, Differential Absorption Lidar Performance

Roger R. Petrin, Doug H. Nelson, John R. Quagliano, Mark J. Schmitt, Charles R. Quick, Robert K. Sander, Joe J. Tiee, and Michael Whitehead

Los Alamos National Laboratory MS E543

Los Alamos, NM 87545 Phone: (505) 667-4573 Fax: (505) 665-4267 E-mail: rrp@lanl.gov

Abstract -- CO, differential absorption lidar (DIAL) performance can be adversely affected by the ambient atmosphere between the laser transmitter and the target through a number of different processes. This work addresses two sources of atmospheric interference with multi- spectral C02 DIAL measurements: effects due to beam propagation through atmospheric turbulence and extinction due to absorption by atmospheric gases. We compare measurements of the effective beam size after propagation to predictions from a beam propagation model that includes turbulence effects such as beam steering and beam spreading. We also compare the experimental measurements of atmospheric extinction to those predicted by both a standard atmospheric transmission model (FASCODE) and a chemometric analysis.

INTRODUCTION

COZ differential absorption lidar is an important remote sensing technique and has been used for a variety of applications including chemical effluent detection, geophysical structure identification, and monitoring of meteorological phenomena. Various studies have utilized ground and airborne based systems. Three primary factors contribute to the popularity of COz laser based LIDAR systems: (1) commercial availability, (2) operation in a spectral "atmospheric window", and (3) operation in a "spectral fingerprint" region for many chemicals. The relatively mature laser technology available from commercial COz systems is a major advantage for using COz laser based lidar. A variety of output formats (CW-IOOkHz, kW average power, 100's kW peak powers) are commercially available, unlike many other laser technologies. A second advantage is the relative transparency of the atmosphere in the CO2 spectral region allowing long range operation with modest laser energy. This reduces power and size requirements as compared to other laser technologies. The third important factor is the broad tunability (9-1 1 pm) available in a spectral region where numerous materials exhibit characteristic

This work was supported by the U.S. Department of Energy.

"fingerprint" spectral signatures. This becomes especially important in applications where identification of numerous components is important.

A number of issues must be considered as part of the design and characterization of a COz LIDAR system for any application. Here we will address issues associated with atmospheric effects that influence both any COz DIAL application and those specific to applications involving multi- spectral DIAL from hard targets. In multi-spectral DIAL, several different wavelengths are used in the measurement rather than just two as in typical DIAL systems. The spectral fingerprint of the material of interest, either the ground in geophysical measurements or the chemicals present in a pollution monitoring system, will be imprinted in the spectral characteristics of the signal return. A large number of wavelengths are necessary for systems required to perform a multiple component chemometric analysis. This need introduces additional system complexity. It is necessary to identify and investigate atmospheric effects that could adversely effect its performance before such a system can be uti1 ized effectively .

MODEL

The COz DIAL model used here is described in detail in [ 13. Only a brief overview is given here. The complete mod,el contains the entire DIAL system: transmitter, ambient atmosphere, hard target, and receiver. In the transmitter portion of the model, information about the spatial, spectral, and temporal characteristics of the laser and optical system is used to model the characteristics of the energy output of the system. The receiver portion uses similar information about the detector and receiving optics to predict the signal return strength. A model of the field conditions which includes the effects of atmospheric turbulence, molecular and aerosol absorption and scattering, target spectral response, and speckle on the signal return connect the transmitter and receiver hardware sections of the model. The final output of the model is reported as a signal-to-noise ratio (SNR) for each wavelength used. Detailed descriptions of the sections of the model that predict atmospheric propagation through turbulence can be found in [2,3] while a description of the standard codes (FASCODE and HITRAN) for calculating

spectral variation in atmospheric transmission can be found in

In addition to the standard codes for modeling atmospheric transmission, we have also used a partial least squares chemometric analysis to "fit" the transmission data. This technique is often used to identify and quantify components in a complex mixture. A commercial software package, Grams386/PLSPlus from Galactic Industries, was used. The model was trained with spectra for carbon dioxide, ozone, and water vapor generated using FASCODE. In our analysis, the important result is the spectral residual not the concentrations of the atmospheric constituents. The ability to correct for atmospheric transmission using a partial least squares type chemometric analysis would be extremely useful in situations where background transmission spectra cannot be collected for reference purposes.

[4-81.

RESULTS AND DISCUSSION

A detailed description of the DIAL system used for these measurements can be found in [9]. The measurements were conducted over a 4 week period and performed over terrain consisting of both desert scrub and dry lake bed. The atmospheric temperature varied from 32-40 C and the relative humidity from 3- 15%.

Atmospheric Turbulence Effects The relative increase in beam diameter with turbulence for

two initial beam divergences are shown in Fig. 1. As predicted by the model, smaller initial beam divergences are affected much more by atmospheric turbulence than larger initial beam divergences. While the larger divergence beam shows little variation in relative size with turbulence, the smaller initial divergence beam shows a >50% increase at higher levels of turbulence.

Concentrating on the results with the smallest initial beam divergence used, 0.190 mad, we can examine how well the measurements correspond to model predictions. The best comparison between the model and experiment results when using the Gaussian beam limit for calculating the transverse coherence length. This is not surprising considering the geometry of our system and the general Gaussian shape of our beam. We have calculated the Rayleigh range for our system to be approximately 1.5 km. At 3.39 km to the target, we are at a range of approximately twice the Rayleigh range. This is an intermediate range where neither the plane-wave or spherical-wave approximations are accurate. From the model results, it is apparent that at longer ranges or higher turbulence levels, the results of the Gaussian and spherical wave forms for the transverse coherence length converge, while at shorter ranges or lower turbulence levels, the Gaussian and plane wave forms converge.

An important caveat to these results, however, is the uncertainty in c,' measurements. There are large uncertainties

Gaussian Beam

Gaussian Beam

High Divergence: 1.2 mr

1 0 4 5 1044 1043 1 0-l2

c,' (m"3) Figure 1. Comparison of propagation measurements and model predictions for different divergences - 3.39 km.

associated with any measurement of C,'. Both point probe and optical scintillometer measurements have limitations. The point probes were located near the trailer. They accurately measure Cn2 (+/- 25%) but give only localized measurements. Any spatial variations of Cn2 along the beam are not monitored. Even though the Lockheed scintillometer does perform a spatially averaged measurement of the optical turbulence, it only samples a portion of the beam path. When the uncertainty in these measurements and the similarity in the results of the three different methods of calculating the transverse coherence length are considered, i t is not possible to specify which method of calculating the transverse coherence length is most accurate.

Atmospheric Transmission Fig. 2 shows results of one of our measurements of the

relative absorption coefficient vs. wavelength to those predicted by FASCODE using HITRAN and by the chemometric analysis. The data has been normalized such that the results are relative to the absorption coefficient of the 10P28 line and do not represent absolute values of the

n D FASCODWHITRAN S

For an initially small beam divergence, the data closely matches the model prediction. Our atmospheric transmission results also agree with the model. The relative absorption coefficients for atmospheric transmission are simulated quite accurately (to better than 0.01 km-') over a broad spectral range. The results of the chemometric analysis indicate that this method performs as well as the full FASCODE modeling and may lead to a method for correcting for atmospheric transmission without requiring a prior background scan.

ACKNOWLEDGMENTS

9.0 9.5 10.0 10.5 11.0

Wavelength(pm) Figure 2. Comparison of FASCODE and chemometric fitting to measured relative atmospheric absorption coefficients.

absorption coefficient. Good agreement is found in the 9-1 1 pm region with generally less than a 0.02 km-' difference between the predicted and measured values (slightly larger errors, up to 0.04 km-', appear in some measurements).

Ignoring the 10R20 and the 9R12 lines (lines with strong water vapor absorption and known discrepancies with the HJTRAN database [lo]) the standard deviation of the difference between measured and predicted values across the spectrum for the FASCODE model is 5.0 x km-' while for the chemometric analysis it is 5.8 x l o 3 km-'. The spectral residual for the chemometric analysis is as small as the that from the FASCODE run. This implies that the three constituents used in the analysis, carbon dioxide, ozone and water vapor, provide adequate information to determine the atmospheric transmission. Additionally, this result indicates that it may be possible to use these three constituents as additional factors in a full chemometric analysis of spectral data with additional information from a target imprinted on it to correct for the atmospheric transmission without using a prior background scan. The spectral residual is a measure of the sensitivity of a multi-wavelength DIAL system. If a separate background scan cannot be obtained, only those variations in absorption coefficient larger than the spectral residual can be attributed to effects other than atmospheric absorption

CONCLUSIONS

We have shown reasonable agreement between our COz DIAL model and our experimental results. The model accurately simulates measurements of beam size after propagating through high levels of atmospheric turbulence.

The authors would like io acknowledge technical assistance from William M. Porch, John J. Jolin and Charles Fite.

REFERENCES

[ I ] M. Schmitt, B. Cooke, and G. Busch, "A comprehensive system model for C02 DIAL, overview," Los Alamos National Laboratory Unclassified Report 95-2988, 1995.

[2] R.R. Beland, "Propagation through atmospheric turbulence," The Infrared Electro-optical Sysrenis Handbook, Vol. 2, SPIE, Bellingham, WA., pp. 157-232, 1993.

[3] R.L. Fante, "Electromagnetic beam propagation in a turbulent media," Proc. IEEE, Vol. 63., pp. 1669- 1692, December 1975.

[4] M.E. Thomas and D.D. Duncan, "Atmospheric Transmission," The Infrared Electro-optical Systems Handbook, Vol. 2, SPIE, Bellingham, WA., pp. 1-156, 1993.

[SI V.N. ArefYev, "Molecular absorption in the 8-13 pm atmospheric window," Izv. Atmos. and Ocean. Phys., Vol.

[6] L.S. Rothman et al, "The HITRAN database: 1986 - edition," Appl. Opt., Vol. 26, pp. 4058-4097, 1987. [7] W.L. Ridgway, R.A. Moose, and A.C. Cogley,

"Atmospheric transmittancekadiance computer code FASCOD2," Rep. AFGL-TR-82-0392, U.S. Air Force Geophysics Laboratory, Hanscom Air Force Base, Mass. 1982.

[8] S.A. Clough, F.X. Kneizys, E.P. Shettle, and G.P. Anderson, "Atmospheric radiance and transmittance: FASCOD2," Proceedings of the Sixth Conference on

Atmospheric Radiation, p. 14 1 , American Meteorological Society, Boston, MA. 1986.

[9] R.R. Petrin, D.H. Nelson, M.J. Schmitt, C.R. Quick, J.J. Tiee, M. Whitehead, "Atmospheric effects on COz differential absortion lidar sensitivity," SPIE Vol. 2702, in press.

[ 101 W.B. Grant, "Water vapor absorption coefficients in the 8-13-pm spectral region: a critical review," Appl. Opt., Vol. 29, pp. 451-462, February 1990.

21, pp. 863-891, 1991.