Open Access
Add to BibSonomyAbstract
Methane is an efficient absorber of infrared radiation and a potent greenhouse gas with a warming potential 72 times greater than carbon dioxide on a per molecule basis. Development of methane active remote sensing capability using the differential absorption lidar (DIAL) technique enables scientific assessments of the gas emission and impacts on the climate. A performance evaluation of a pulsed DIAL system for monitoring atmospheric methane is presented. This system leverages a robust injection-seeded pulsed Nd:YAG pumped Optical Parametric Oscillator (OPO) laser technology operating in the 1.645 µm spectral band. The system also leverages an efficient low noise, commercially available, InGaAs avalanche photo-detector (APD). Lidar signals and error budget are analyzed for system operation on ground in the range-resolved DIAL mode and from airborne platforms in the integrated path DIAL (IPDA) mode. Results indicate system capability of measuring methane concentration profiles with <1.0% total error up to 4.5 km range with 5 minute averaging from ground. For airborne IPDA, the total error in the column dry mixing ratio is less than 0.3% with 0.1 sec average using ground returns. This system has a unique capability of combining signals from the atmospheric scattering from layers above the surface with ground return signals, which provides methane column measurement between the atmospheric scattering layer and the ground directly. In such case 0.5% and 1.2% total errors are achieved with 10 sec average from airborne platforms at 8 km and 15.24 km altitudes, respectively. Due to the pulsed nature of the transmitter, the system is relatively insensitive to aerosol and cloud interferences. Such DIAL system would be ideal for investigating high latitude methane releases over polar ice sheets, permafrost regions, wetlands, and over ocean during day and night. This system would have commercial potential for fossil fuel leaks detection and industrial monitoring applications.
© 2013 Optical Society of America
1. Introduction
Development of a new remote sensing capability to better understand the distributions and variability of atmospheric methane (CH4) is required to address a number of important science and societal issues. CH4 is cited as an important atmospheric variable by several panel reports in the 2007 National Research Council Decadal Survey (DS) [1]. These variables include human health and security, climate variability and change, land-use change, ecosystem dynamics, and biodiversity. Consequently the DS recommended adding CH4 sensing capabilities, provided technological maturity and cost-effectiveness of remote sensors could be realized. The plan for a climate-centric architecture by NASA recognizes the importance of CH4 and discusses the potential for the gas sensing capability on the follow-on to OCO-2 [2]. A U.S. Carbon Cycle Science Plan (CCSP), currently under development, recognizes the importance of CH4. CCSP emphasizes the need of an integrated system to collect and maintain essential data that drive scientific understanding of the gas processes [3]. Internationally, science leaders recognize the need for global scale CH4 measurements with active remote sensing. A joint German-French initiative to develop a space-based CH4 DIAL system, MERLIN, is under development [4].
A standard practice is to project global warming over a 100-year time horizon. In this case, carbon dioxide (CO2) has higher influence than CH4 in global warming. However, CH4 has stronger infrared (IR) absorption features that are unsaturated and thus is a potent greenhouse gas. Uncertainty in climate prediction over a 20-year horizon is dominated by uncertainty from CH4 radiative effects because, on a per molecule basis, CH4 warming influence is 72 times more than that of CO2 [5]. A 20 year time horizon is a realistic time scale for development of mitigation strategies to counteract global warming trends. The DS mission ASCENDS (Active Sensing of CO2 Emission over Nights, Days, and Seasons) Workshop Report [6] defines and emphasizes CH4 observation requirements and its relevance in support of the mission core science. Specifically, CH4 observations are needed to assess global climate change more accurately, particularly over the northern high latitudes where the ecosystem is more prone to climate change and a faster response time has been observed [7]. In addition, potentially large unknowns include climate feedbacks associated with rising temperatures in the atmosphere and bodies of water in the northern latitudes, which could destabilize vast terrestrial (e.g., permafrost) and marine (e.g., permafrost and methane hydrate) CH4 reservoirs. A U.S. Geological Survey report [8] estimates that worldwide the amount of carbon bound in gas hydrates is about twice the total amount of carbon in all known fossil fuels on Earth. Therefore, monitoring and identifying new CH4 gas release sources could become as important as oil exploration in the next few decades, while extraction of CH4 from hydrates could prove to be a sustainable source of energy in the future. In addition, CH4 is chemically active in the atmosphere and influences the production and distribution of tropospheric CO2 and ozone (O3). Due to its abundance, CH4 can be the dominant hydroxide (OH) sink in some regions, significantly determining the oxidation capacity of the atmosphere [9]. The subsequent CH4 oxidation chemistry further influences tropospheric O3, a greenhouse gas and pollutant associated with poor air quality [10].
Although current satellite observations are an important step towards fulfilling the need for global scale CH4 monitoring, they do not satisfy all the requirements articulated by the scientific community [1]. For instance, AIRS observations of thermal IR emission from CH4 are sensitive to the mid-troposphere and lack sensitivity near the surface where sources induce the strongest gradients [11]. SCIAMACHY, GOSAT, and the scheduled CarbonSat observations use reflected solar radiation to achieve better sensitivity in the lower troposphere, but have large footprints and are challenged over water due to low reflectivity [12–14]. Global CH4 observations from GOSAT produce useful retrievals, with limited coverage at high latitude, over oceans, in the tropics, and at night, for sources that tend to vary dramatically [15]. Therefore, new technology is needed to enable the mapping of near-surface CH4 to delineate boundary layer influences from total column at high spatial resolution for both marine and terrestrial source regions.
Developing new active remote sensing technology for monitoring atmospheric CH4 using the differential absorption lidar (DIAL) technique enables new capability for the gas retrieval that will overcome some of the deficiencies of passive remote sensing techniques [16, 17]. In this paper, the performance analysis and evaluation of a CH4 DIAL system is presented. This DIAL system could operate in a range-resolved (RR-DIAL) mode from ground or in the integrated path differential absorption (IPDA) mode from an airborne platform. A byproduct of a ground-based CH4 RR-DIAL system is the ability to retrieve simultaneous measurements of aerosol and cloud distributions, as well as the inherent ranging capability. Airborne CH4 IPDA would provide measurements over regional scales with high resolution and accuracy using simple retrieval techniques. This feasibility study is based on realistic system parameters that are scalable to a space-based instrument. This provides an investigation of a new class of Earth observing remote sensing capability which is applicable to NASA sub-orbital satellite validation and venture class programs. Such system would have commercial potential in many applications such as fossil fuel leak detection and industrial monitoring applications.
2. DIAL system for atmospheric methane
In the DIAL remote-sensing technique, two laser pulses are transmitted through the atmosphere to monitor a specific gas molecule. One laser pulse is tuned to a spectral location that is strongly absorbed by the molecule of interest (on-line) and the other pulse is tuned to a nearby, less absorbing spectral location for the same molecule (off-line). The backscattered radiation from the atmosphere or the reflected radiation of the surface from both pulses are collected with a telescope and imaged onto photo-detectors. In principle, gas concentration profiles can be retrieved using the on-line and off-line lidar signal ratios and the knowledge of the differential absorption cross-section, Δσ [18]. DIAL has been applied successfully and accurately in monitoring different atmospheric gases such as water vapor (H2O), O3, CO2 [19–24].
There are numerous advantages for monitoring atmospheric CH4 using the DIAL technique. These advantages include direct inversion of absolute CH4 concentration profiles without a need for initial guesses or additional calibration. DIAL measurements are optimized by the selection of the on and off-line wavelengths to provide high sensitivity to the target molecule while minimizing interference from other absorbing molecules. In addition, DIAL systems provide measurements during day and night. The by-products of RR-DIAL measurements are simultaneous aerosol profile and cloud distributions. Further, the pulsed system eliminates interferences from aerosols and clouds. Additionally, pulsed lasers provide higher signal-to-noise ratios because of the ability to overcome background and detector noise over the short pulse duration. Parameters for the key components of the CH4 DIAL system are listed in Table 1.
Table 1. Instrument parameters list applied for analysis based on existing technology.
2.1 Laser Transmitter
One critical and technically challenging component in achieving the CH4 DIAL capability is the development of the pulsed laser transmitter. Such transmitter, being developed at Fibertek, Inc., is based on an innovative tunable, pulsed laser system that operates in the 1.645 µm spectral region [25]. The CH4 DIAL transmitter uses an injection seeded Nd:YAG pumped Potassium Titanyl Phosphate (KTP) based optical parametric oscillator (OPO). The OPO cavity is pumped at 1 kHz pulse repetition frequency (PRF). To achieve narrow line-width operation out of the OPO cavity, injection seeding is employed via two CW fiber coupled tunable distributed feedback (DFB) lasers. The CW lasers are stabilized and operated at the on-line and off-line wavelengths of the CH4 R6 absorption line. By combining a single frequency pump laser and an injection seeded KTP OPO, 3-5 mJ of output energy over a 10 ns full width at half maximum (FWHM) pulse duration has been achieved at 1.645 μm. At 1 kHz PRF, the measured optical conversion efficiency is expected to be greater than 20%. The residual pump at 1.064 μm is frequency doubled and used as a transmitter for high spectral resolution lidar (HSRL) measurements of aerosol properties. For comparison, the residual energy used for this HSRL measurement is a factor of 2-3 higher than that of the laser used in the NASA Langley Research Center (LaRC) HSRL-I instrument with 5 times higher PRF [26]. This additional capability aids in identifying the sampled aerosol types [27], boundary layer heights [28] and liquid water at the surface [29]. This significantly helps in distinguishing between natural and anthropogenic CH4 sources and how these sources mix within the boundary layer and free troposphere. Performance specifications of the laser transmitter are provided in Table 1.
Stabilization of the OPO cavity is implemented by locking the laser cavity length referenced to the DFB seed laser using the Pound-Drever-Hall (PDH) technique [30]. The seed lasers provide broad wavelength tunability in the vicinity of the CH4 absorption line which allows for optimization of the on-line and off-line locations. Three seed lasers are implemented to control the on-line (or side-line) and off-line wavelengths. First, a reference DFB seed laser is locked to a CH4 absorption cell using the PDH technique. Second, offset locking of two separate DFB seed lasers is employed to set the desired operating wavelengths relative to the stabilized reference laser [31]. This allows for optimization of the on-line and off-line operation wavelengths to achieve the desired weighting function. Injection seeding of the OPO cavity between the two cw DFB laser sources is achieved using an electro-optic switch operating at a 1000 Hz PRF with a rise and fall time of less than 10 µs. Frequency stabilization of the OPO laser cavity to the selected injection seeding wavelength occurs on a shot-by-shot basis, allowing for near simultaneous sampling of the atmospheric volume and surface foot print required for the DIAL measurements. Sufficient beam expansion of the output beam sets the laser divergence to 300 µrad (full angle) which decreases the spot size at the surface to reduce background radiation.
2.2 Receiver
The CH4 DIAL receiver design consists of a large area telescope and aft-optics that focus the radiation onto a small area detector. The receiver utilizes a 40-cm diameter all aluminum Cassegrain telescope with integrated field stop and all reflective collimating optics. The telescope FOV is set to 500 μ-radians using a pinhole located at the focal plane before the collimating mirrors. The FOV setting was driven to achieve a small focus area of 180-μm diameter at the detector. This allows for coupling the collected radiation onto the 200-μm-diameter low-noise InGaAs avalanche photodiodes (APD) that are implemented. The aft-optics also includes a narrowband, 1-nm, interference filter and broadband blocking filters to reject background radiation. The selected APD (Voxtel Siletz SCM-APD) has high quantum efficiency (80%) and a gain of 10 with low noise-equivalent-power (NEP). The APD is integrated to the lidar detection system which includes the detector conditioning in terms of the operating bias voltage and temperature and signal conditioning electronics. Signal conditioning consists of a low noise current-to-voltage converter amplifier (Femto DHPCA-100). To extend the dynamic range of the return signals, the preamplifier output is split into a low and high gain channels. The high gain channel is achieved by introducing another variable gain amplifier after the preamplifier. The signal from each channel applied to a 14-bit, 50-MS/s waveform digitizer (National Instrument NI-5751). Single shot profiles are stored in the system computer (National Instruments PXI-chassis) for further analysis. Table 1 list the receiver parameters assumed for this analysis.
3. Methane DIAL system analysis
Specifications for evaluating the CH4 DIAL system performance are presented in this section. The specifications are based on the 1.645 μm pulsed laser development in progress at Fibertek, Inc., and the receiver and detection system designed and developed at NASA LaRC, as discussed in the previous section. The CH4 DIAL can be operated in the RR-DIAL mode from ground, or in the IPDA mode from an airborne platform. The principal products of CH4 RR-DIAL measurements are the gas concentrations from the ratio of the on to off-line signals, and aerosol backscatter profiles from the off-line channel. The principle product of the IPDA measurement from an airborne platform is the spatial and temporal distribution of the gas column weighted average volume mixing ratio. A key selection affecting the DIAL instrument sensitivity is the choice of the on-line and off-line wavelength locations. Such locations influence the measurement sensitivity and weighting to a specific altitude region. CH4 has strong absorption features around the 1.6 μm wavelength region that are possible to target due to several critical technological advances in recent years including the tunable seed lasers, robust and high power OPO lasers and high performance APDs.
3.1 Spectroscopy and line selection
For 1.65 µm CH4 DIAL, the R6 line at 1.6456 μm provides optimum absorption cross sections for tropospheric measurements. Unique characteristics of this location include low temperature sensitivity and low interference from higher abundant H2O and CO2 molecules [4, 32]. Figure 1 compares the composite absorption spectra for CH4 with H2O and CO2. The spectra were derived using the HITRAN 2008 database for line parameters assuming Voigt line profile at three altitudes [33, 34]. Altitude specification was based on the CH4 DIAL operating platform; 0 km for the ground-based evaluation, 8 km for small aircraft such as the NASA B-200, and 15.24 km for unmanned aerial vehicle (UAV) such as the Global Hawk, as listed in Table 2. The selection of the on-line wavelength is similar to Kiemle et al. [4]. The off-line wavelength was selected on the shorter wavelength side of the line center. This new off-line location provides less sensitivity to H2O and CO2 and reduces the on-to-off line separation (179.4 pm). A reduced on-to-off line separation minimizes systematic errors due to differential aerosol scattering and extinction for RR-DIAL and differential albedo effects for IPDA DIAL. Table 1 lists the spectral locations used in this study.
Fig. 1 Comparison of the CH4, CO2 and H2O absorption spectra derived using the HITRAN 2008 database for line parameters and Voigt line profile at three altitudes. Altitude specification was based on the CH4 DIAL operating platforms from; ground (0 km), small aircraft (8 km) and UAV (15.24 km). Metrological data were obtained from the US Standard model. Vertical lines mark the instrument operating wavelengths.
Table 2. Platform and environmental conditions assumed in this study.
Radiative transfer calculations were conducted using US Standard model for meteorological profiles and gas mixing ratios as shown in Fig. 2 [35]. CH4 and CO2 mixing ratio models were modified based on current nominal values at ground of 1.8 and 390 ppm, respectively. A standard model was used to drive the aerosol extinction coefficient profile at 1.6 µm wavelength as shown in Fig. 2 [36]. Aerosol backscatter coefficient was calculated from the extinction assuming a lidar ratio (extinction-to-backscatter ratio) of 35. An enhanced aerosol backscatter layer was introduced to the aerosol profile near the top of the atmospheric boundary layer. This layer assumes a Gaussian profile with 2.4 × 10−6 m−1⋅sr−1 peak backscatter coefficient occurring at 1.4 km altitude and 200 m width.
Fig. 2 Temperature and pressure profiles as well as CH4, CO2 and H2O dry mixing ratio profiles used in the calculations. The profiles were obtained from the US Standard model with nominal values of CH4 and CO2 mixing ratios of 1.8 and 390 ppm, respectively, at ground [21]. Aerosol extinction profile at 1.6 µm wavelength was obtained from [24]. An enhanced backscatter layer was introduced to the model assuming a Gaussian distribution versus altitude with a peak backscatter coefficient of 2.4 × 10−6 m−1⋅sr−1 occurring at 1.4 km altitude and 200 m width. Horizontal dash lines mark the airborne platform altitude.
Generally, for an airborne DIAL system, the optical depth of any of the atmospheric gases, ODgas, is calculated from
where σgas and χgas are the gas absorption cross section and dry air volume mixing ratio profiles, respectively, and ndry is the dry air number density. The integration is performed with respect to the range, r, with the limits RG and RA representing ground and airborne altitudes, respectively. The optical depth is calculated for both on-line and off-line wavelength locations, λ. Applying the Ideal Gas Law, the dry air number density is obtained using the equationwhere P and Τ are the atmospheric pressure and temperature profiles, respectively, k is the Boltzmann’s constant and χwv is the water vapor dry volume mixing ratio.Figure 3 shows the double-path differential optical depth calculations for CH4, H2O and CO2. The elevation-dependent double-path differential optical depth, dODgas, was calculated from
and applied for both aircraft and UAV platforms. The figure indicates the advantage of the selected off-line location in minimizing H2O differential optical depth near the surface, where the gas amounts are largest and most variable. This minimizes the interference from H2O variability on the CH4 measurements. The figure also shows the aerosol double-path optical depth calculated by integrating the aerosol extinction coefficient obtained from the model [33]. Table 2 lists the platform and environmental conditions used for these calculations. The on-line and off-line wavelengths are selected to achieve optimum sensitivity to near surface altitude for column CH4 measurements. Computation of the pressure-based weighting function of CH4 is a prerequisite for the retrieval of the average volume dry mixing ratios from the DIAL data. Figure 4 shows the CH4 pressure-based, peak-normalized, weighting function at the selected spectral positions for the nadir IPDA measurement from an airborne platform versus altitude. The wavelength tunability of the laser line location enables the selection of an optimum weighting either in the free troposphere for transport measurements, in the boundary layer (selected) or near surface for sources and sinks identification. However, IPDA operation near the surface could results in higher error budget. The pressure-based CH4 weighting function is calculated fromwhere σmn is the methane absorption cross section, g is the Earth’s gravitational constant and mdry and mwv are the average masses of dry air and water vapor molecules, respectively [32].Fig. 3 Integrated double-path differential optical depth for atmospheric CH4, H2O and CO2 molecules and aerosol from an aircraft (dashed) and UAV (solid) at 8km and 15.24 km altitudes, respectively.
Fig. 4 CH4 pressure-based, peak-normalized weighting function at the selected spectral positions for nadir IPDA measurements. Tuning the on-line wavelength 8 pm and 35 pm away from the selected location would optimize the IPDA measurement to the free troposphere or ground surface, respectively.
3.2 Signal and noise analysis
The CH4 DIAL system can be operated in the RR-DIAL mode in open atmosphere or in the IPDA mode using a hard target. Performance projections of the CH4 DIAL system were conducted using methodology developed earlier [19, 37, 38] for RR-DIAL measurements in the boundary layer and column CH4 IPDA from aircraft and UAV platforms. For airborne operation, the IPDA mode would be used for nadir measurements using backscatter returns from ground, enhanced aerosol features or thin clouds.
For horizontal ground operation, the on-line and off-line return signal profiles are calculated using the backscatter lidar equation. For lidar signals, the number of detected photoelectrons, Ns, per sampling period, τ, is given by
where λ is the operating wavelength, E is the transmitted laser pulse energy, h is the Plank’s constant, A is the receiver area, ηd is the quantum efficiency of the detector, ηr is the receiver efficiency including the overlap function, β is the aerosol backscatter coefficient, OD is the total double-path optical depth. An analytical expression was used to include the influence of the beam overlap, which affects the near-field signal [18]. Similarly, for nadir airborne operation, the on-line and off-line return signal are calculated from the hard target lidar equation given (in photoelectrons) bywhere ρ is the surface reflectivity of the hard target (ground) and c is the speed of light. The effective pulse width of the return signal, τw, is a combination of the transmitted laser pulse width, τL, detection system bandwidth, BW, and the effective target altitude within the laser footprint, Δh, according to [32]In either cases the number of background photoelectrons, NBG, associated within the measurement depends on the receiver geometry and the background irradiance, SBG [in W/(m2⋅nm⋅sr)], according to [32]where FW is the optical filter bandwidth (in nm) and FOV is the receiver telescope field-of-view. Table 2 lists the ocean and vegetation day background irradiances used in this study [32]. The detected signal converted from photoelectron to preamplifier output voltage, VS, is given bywhere q is the electron charge, M is the multiplication gain assuming an avalanche photodiode (APD) and Rf is the preamplifier trans-impedance gain.The total noise associated with the detected signal can be divided into fixed circuit noise and signal-dependent shot noise. The total circuit noise current spectral density (in A/Hz1/2), In, is given by [32]
where Id and F are the detector dark-current and excess-noise-factor, respectively, InA and VnA are the preamplifier integrated input current and voltage noise spectral densities, Rf is the preamplifier feedback resistance, and Cd is the detector and preamplifier equivalent input capacitance. The circuit noise is usually dominated by either the detector dark-current shot noise or the preamplifier noise. In this analysis, all circuit noises are referred to the detector input and the equivalent circuit noise-generated photoelectrons, Nn,C, is calculated fromSimilarly, the equivalent shot noise-generated photoelectrons, Nn,S, are calculated fromThis equation is applicable to either signal or background shot noises. Treating these noises as equivalent photoelectrons generated within the detector, (i.e., before the multiplication process) is done for comparison to the actual detected photoelectrons, while normalizing to the detection system gains. The resultant signal-to-noise ratio (SNR) is given byEquation (13) is applicable to either the on-line or off-line wavelengths, where Nn,BG is the background shot noise, calculated from Eq. (12).3.3 Sensitivity analysis and error budget
The CH4 column weighted average dry-air volume mixing ratio, Xmn, is obtained by measuring the differential optical depth and estimating the weighting function according to
where PG and PA are the ground and airborne pressures, respectively. Errors in Xmn retrievals results from both random and systematic sources. The CH4 weighting function is obtained at each ground elevation level and depends upon meteorological conditions which do not vary on shot-to-shot basis within the data averaging interval. Therefore, the retrieval of Xmn is dominated by the accuracy of the optical depth measurements. Thus, the total error, ε, is given bywhere εR is the random error associated with the lidar signal including the detection system, εA and εT are the systematic errors associated with the knowledge of the atmospheric environment and laser transmitter, respectively, and s is the number of shot average. Although, random error can be reduced by shot-averaging, measurements could be affected by variability of the surface conditions, speckle noise and aircraft velocity. Assuming that the ratio of the on- and off-line signals are appropriately normalized with respect to the ratio of the transmitted energies (0.1% laser fluctuation uncertainty), and speckle noise is reduced by shot averaging and receiver aperture, the random error is given by [4, 32]Atmospheric environmental sensitivity results from a combination of sensitivities to metrological data and the interference from molecules other than CH4 and consequently,
where εA,t, εA,p, εA,wv and εA,m are the systematic errors due to temperature, pressure, relative humidity and other interfering molecules, respectively. These are given by where dODwv and dODcd are the water vapor and carbon dioxide differential optical depths, respectively. The systematic errors from the transmitter result from the spectral quality and control of the laser source. The laser transmitter sensitivity is dependent on the beam spectral quality in terms of wavelength jitter and line-width and are given bywhere εj,on and εj,off, are the on-line and off-line laser line jitter errors and εp,on and εp,off, are the on-line and off-line laser line-width errors, respectively. The on and off-line laser line jitter errors are given bywhere Δλon/off is the on or off line wavelength position variance. The on-line and off-line laser line-width errors are given bywhere is the CH4 differential optical depth calculated using the effective absorption cross section, , defined bywhere G is the laser spectral intensity distribution profile, assumed to be Gaussian and is altitude-independent, σmn is the Voigt absorption cross section profile and L is the wavenumber ( in cm−1).4. Performance Evaluation of Methane DIAL
Based on the lidar parameters discussed in the previous section, the prediction of the CH4 DIAL system performance was evaluated. The DIAL system is being designed to operate from ground, medium altitude aircraft or a UAV. In the ground-based mode, RR-DIAL measurements can be achieved using atmospheric backscattering from boundary layer aerosols. From the aircraft or UAV, IPDA CH4 column average dry mixing ratio measurements can be achieved from surface reflected signals. Simultaneously, the same column measurements within the boundary layer can be achieved using signals from the residual layer (high scattering from aerosols or clouds on top of the convective boundary layer) and surface return.
4.1 Ground-based measurements
The ground-based operation of the DIAL system can provide three-dimensional CH4 profiles within the atmospheric boundary layer using a scanner. However, the performance was evaluated assuming horizontal path in this study. The on and off-line backscattered returns, and the background signals, were calculated from Eqs. (5) and 8, respectively. The associated noises from these signals as well as the circuit noises were calculated from Eqs. (10) through 12. Figure 5 compares the different detected signals to the corresponding noise as well as the total noise, per laser shot. On and off-line signal shot noise dominates in the near-field while circuit noise dominates in the far-field. Figure 6 shows the different errors associated with the optical depth measurements by applying Eqs. (15) to 25. Systematic errors associated with the uncertainties in the knowledge of temperature, pressure, relative humidity, and CO2 and H2O interference were estimated. The total systematic error was calculated in rms for all these uncorrelated errors [19]. Temperature sensitivity was calculated using ± 1°C uncertainty. Pressure sensitivity was calculated for ± 2 mbar deviation. Relative humidity sensitivity was calculated for ± 20% deviation. For the laser transmitter, ± 10 MHz laser line position uncertainty and 20 MHz half-line width were used for calculating the jitter and profiling sensitivity, respectively for on and off-line. For a single shot, receiver error dominates, as shown in Fig. 6. The atmospheric and transmitter errors are limited to about 0.1% of the optical depth measurements. One advantage of the ground measurement is that, under stable atmospheric conditions, longer time averaging could be used to increase the measurement sensitivity. This is indicated in the same figure, where shot averaging is applied and only total errors are illustrated. To maintain a given sensitivity, range can be traded for temporal resolution. Using the on and off-line signals from Fig. 5 the CH4 number density profile is retrieved applying the standard RR-DIAL, Eq [19].
Errors associated with the range resolved number density measurements are shown in Fig. 7. A fixed bias error of 0.13% results from systematic errors. The RR-DIAL random error was obtained from [19]The random error reduction by shot averaging is also illustrated in the figure. High precision measurements could be achieved with appropriate time averaging. For example, with a 5 minute average and 500 m range cell 1% precision is obtained up to 4.5 km range from the instrument. Simulations for ground-based RR-DIAL system performance indicate a total optical depth error of 0.5% for ranges up to 1.5 km for 1 sec averaging, which can be extended to 7 km with 5 min averaging.Fig. 5 Ground-based horizontal measurements on-line and off-line return signals and background signals as a function of range calculations. The associated noises are compared indicating that the total noise is shot-noise limited in near-field and limited by the detection circuit noise in far-field.
Fig. 6 Ground-based horizontal measurements optical depth errors. Total error is dominated by the receiver error due to the noise sources. Averaging reduces the error thereby extending the measurement range.
Fig. 7 Ground-based horizontal measurements RR-DIAL error. A fixed error of 0.13% indicates a measurement bias attributed to systematic sources. By averaging in time, the random error for a fixed range could be reduced or for a fixed error (0.5% horizontal dashed line) the measurement range could be extended. A 500-m range cell was assumed for this calculation.
4.2 Airborne nadir evaluation
For the airborne case, the system will be operated from either a mid-altitude aircraft or high altitude UAV in the nadir direction. Figure 8 shows the backscatter lidar return signals for both platforms, calculated using Eq. (5). The background signals are shown in the same figure assuming 90° solar elevation angle. For the UAV case, background could dominate the backscatter return, when high scattering target (vegetation) exists. Figure 9 shows the ground return pulse strength per shot versus ground elevation calculated from Eq. (6). Reflectivity from both ocean and vegetation scenes are used to estimate the range of signals from the ground return. Comparing both figures, the backscatter signal is much weaker than the ground return signals. To capture the full dynamic range of the ground and atmospheric returns, the return signal is optically split into high gain and low gain paths. The detected signal of each path can be further split after the preamplifier into high and low gain channels, then applied to two digitizer inputs. In order to maintain high signal-to-noise ratio, and for practical consideration, the detection system bandwidth was limited to 3.5-MHz. This bandwidth will reduce the signals dynamic range by spreading out sharp features such as the ground return.
Fig. 8 The backscatter lidar return signal simulation for the aircraft and UAV platforms. The background signals are shown for ocean and vegetation surfaces assuming overhead sun.
Fig. 9 Ground return pulse strength per shot versus ground elevation for ocean and vegetation surfaces for the aircraft and UAV platforms.
Figure 10 shows the error analysis associated with the optical depth measurement for the aircraft and UAV, applying Eqs. (15) through (25). The background signals are 241 and 709 photoelectrons for the ocean and vegetation returns, respectively. Figure 10 shows the column optical depth measurement errors resulting from systematic effects, random error and their combination. This is done for a range of ground elevations achievable by a medium altitude aircraft such as the NASA LaRC B-200. This illustrates the feasibility of obtaining high accuracy (< 1%) measurements of column CH4 amounts to the surface from an aircraft with high temporal (< 0.1 sec) and spacial (< 15 m) resolution. Similarly, Fig. 10 shows the systematic and random errors and the total error estimates to the surface from a high altitude UAV. The capability of trading precision against resolution, through averaging, points-out another advantage of the DIAL technique. The calculation results indicate the feasibility of obtaining high accuracy and high resolution CH4 measurements weighted near the surface from an airborne platform. These measurements can be extended to even lower reflectivity targets such as some variety of snow surfaces.
Fig. 10 Error analysis for the UAV (top) and the aircraft (bottom) platforms. (Left) Atmospheric sensitivities to ± 1°C temperature, ± 2 mBar pressure, ± 20% relative humidity and interfering molecules. The total atmospheric sensitivity is dominated by temperature. (Middle) Laser transmitter error due to ± 10MHz jitter and 20 MHz line half-width for the on-line and off-line. Off-line width error is less than 10−4% and not shown. Total transmitter error is dominated by the on-line. (Right) Receiver error for ocean and vegetation targets with 0.1 second average, marking the random error source. Combining the atmospheric sensitivity and transmitter error define the systematic error. Adding the receiver error produce the total measurement error that is less than 0.3%.
4.3 Boundary layer measurements
Most CH4 sources are located near the surface. A benefit of the pulsed DIAL system is the capability of combining signals from the atmospheric scattering from layers above the surface with ground return signals. This provides the capability to measure CH4 amounts between the atmospheric scattering layer and the ground directly. This unique capability can be achieved by the specified DIAL system as shown in Fig. 8. Generally, higher aerosol scattering layer exists on top of the atmospheric boundary layer. This scattering layer comprises of either a residual layer at night or near the top of convective boundary layer during day. Figure 11 illustrates the capability of making accurate measurements of average CH4 column amounts within the boundary layer from aircraft and UAV platforms. The errors in optical depth are dominated by random errors from signal shot noise from the atmospheric scattering layer and daytime background signals. By averaging signals over 10 sec optical depths with 0.5% and 1.2% errors can be obtained from the aircraft and UAV, respectively. This high precision provides unique remote sensing capability for identifying and mapping localized sources of both natural and anthropogenic CH4.
Fig. 11 Optical depth error versus averaging time calculated for the top of the boundary layer measurements from aircraft and UAV platforms. The total errors are dominated by the random error which can be reduced by averaging.
5. Conclusions
Development of a new remote sensing capability for understanding the distributions and variability of atmospheric CH4 is required for better understanding of the carbon cycle. The feasibility of CH4 remote sensing using the DIAL technique was analyzed. The study assumed RR-DIAL from ground and IPDA from aircraft and UAV platforms. The study is based on the planned development of the 1.645-μm pulsed laser technology at Fibertek, Inc., and using an integrated receiver and detection system being developed at NASA LaRC. Simulation for ground-based RR-DIAL system performance indicate a total optical depth error of 0.5% for ranges up to 1.5 km for 1 sec averaging, which can be extended to 7 km with 5 min averaging. Total CH4 number density error of 1% could be achieved with 5 min averaging and 500 m range cell with range up to 4.5 km. For airborne platform the feasibility of obtaining high accuracy measurements of column CH4 amount to the surface was examined. Airborne IPDA are capable of measuring the integrated average column dry mixing ratio with about 0.3% error when averaging the ground returns over 0.1 sec. Combining the signals from atmospheric scattering above the surface with ground return provides an enhanced capability to estimate CH4 amounts within the boundary layer. This unique capability can be achieved by the specified DIAL system. For example using a scattering layer above the boundary layer would provide capabilities to measure column integrated dry-air CH4 mixing ratios between the surface and that scattering layer with 0.5% and 1.2% total error from aircraft and UAV, respectively. Development of this capability would provide unique measurements to improve the understanding of climate, carbon cycle, atmospheric chemistry and environment and enables validation of satellite measurements. This system will provide key measurements not currently covered by satellites that have either lower sensitivity near the surface (AIRS) or are limited to daytime overland measurements (GOSAT) and are influenced by the presence of aerosols and clouds. This system would advance DIAL technologies and will complement MERLIN satellite IPDA measurements.
References and links
1. National Research Council, Earth Science and Applications from Space: National Imperatives for the Next Decade and Beyond (The National Academies Press, 2007).
2. NASA OCO-2 Mission, http://science.nasa.gov/missions/oco-2/.
3. A. Michalak, R. Jackson, G. Marland, and C. Sabine, A U.S. Carbon Cycle Science Plan (University Corporation for Atmospheric Research, Boulder, Colorado, 2011).
4. C. Kiemle, M. Quatrevalet, G. Ehret, A. Amediek, A. Fix, and M. Wirth, “Sensitivity studies for a space-based methane lidar mission,” Atmos. Meas. Tech. 4(10), 2195–2211 (2011). [CrossRef]
5. S. Solomon, D. Qin, M. Manning, M. Marquis, K. Averyt, M. Tignor, H. Miller, and Z. Chen, Climate Change 2007: The Physical Science Basis, Working Group I Contribution to the Fourth Assessment Report of the IPCC (Cambridge University Press, New York, New York, 2007).
6. ASENDS Science Definition and Planning Workshop Report (University of Michigan, 2008).
7. National Research Council, Frontiers in Understanding Climate Change and Polar Ecosystems, Workshop Report, (The National Academies Press, 2011).
8. W. Dillon, Gas (Methane) Hydrates - A New Frontier, Energy and Marine Geology, USGS (U.S. Department of Energy Government Printing Office, 1992).
9. M. Khalil, Atmospheric Methane: Sources, Sinks, and Role in Global Change, NATO ASI Series I: Global Environmental Change, (Springer-Verlag, 1993).
10. B. Metz, O. Davidson, P. Bosch, R. Dave, and L. Meyer, Climate Change 2007: Mitigation of Climate Change, Working Group III Contribution to the Fourth Assessment Report of the IPCC (Cambridge University Press, 2007).
11. X. Xiong, C. Barnet, E. Maddy, C. Sweeney, X. Liu, L. Zhou, and M. Goldberg, “Characterization and validation of methane products from the Atmospheric Infrared Sounder (AIRS),” J. Geophys. Res. 113, G00A01 (2007).
12. M. Buchwitz, R. De Beek, S. Noel, J. Burrows, H. Bovensmann, O. Schneising, I. Khlystova, M. Bruns, H. Bremer, P. Bergamaschi, S. Korner, and M. Heimann, “Atmospheric carbon gases retrieved from SCIAMACHY by WFM-DOAS: Version 0.5 CO and CH4 and impact of calibration improvements on CO2 retrieval,” Atmos. Chem. Phys. 6(9), 2727–2751 (2006). [CrossRef]
13. T. Yokota, Y. Yoshida, N. Eguchi, Y. Ota, T. Tanaka, H. Watanabe, and S. Maksyutov, “Global concentrations of CO2 and CH4 retrieved from GOSAT: First preliminary results,” SOLA 5, 160–163 (2009). [CrossRef]
14. E. S. A. The, CarbonSat mission, http://www.astrium.eads.net/en/press_centre/carbonsat-on-the-trail-of-greenhouse-gases.html.
15. Y. Yoshida, Y. Ota, N. Eguchi, N. Kikuchi, K. Nobuta, H. Tran, I. Morino, and T. Yokota, “Retrieval algorithm for CO2 and CH4 column abundances from short-wavelength infrared spectral observations by the Greenhouse gases observing satellite,” Atmos. Meas. Tech. 4(4), 717–734 (2011). [CrossRef]
16. S. Ismail, J. Crawford, I. Leifer, F. Hovis, R. Burnham, J. Hair, T. Refaat, A. Notari, J. Collins, S. Kooi, M. Hardesty, M. Devi, C. Benner, L. Brown, K. Sung, G. Diskin, A. Fix, and N. Abedin, “Development of a 1.65 μm pulsed laser DIAL System to map atmospheric CH4 distributions,” AGU Fall Meeting, A23B–0247, San Francisco, CA, (2010).
17. T. Refaat and S. Ismail, “A feasibility study of atmospheric methane DIAL measurements at 1.65 micron spectral region,” EGU General Assembly 13, 9675 (2011).
18. R. Measures, Laser Remote Sensing, (John Wiley & Sons, 1984).
19. S. Ismail and E. V. Browell, “Airborne and Spaceborne lidar measurements of water vapor profiles: a sensitivity analysis,” Appl. Opt. 28(17), 3603–3615 (1989). [CrossRef] [PubMed]
20. V. Wulfmeyer and J. Bösenberg, “Ground-based differential absorption lidar for water-vapor profiling: assessment of accuracy, resolution, and meteorological applications,” Appl. Opt. 37(18), 3825–3844 (1998). [CrossRef] [PubMed]
21. E. Browell, M. Fenn, C. Butler, W. Grant, V. Brackett, J. Hair, M. Avery, R. Newell, Y. Hu, H. Fuelberg, D. Jacob, B. Anderson, E. Atlas, D. Blake, W. Brune, J. Dibb, A. Fried, B. Heikes, G. Sachse, S. Sandholm, H. Singh, R. Talbot, S. Vay, R. Weber, and K. Bartlett, “Large-scale ozone and aerosol distributions, air mass characteristics, and ozone fluxes over the western Pacific Ocean in late-winter/early-spring,” J. Geophys. Res. 108(D20), 8805 (2003). [CrossRef]
22. R. Alvarez II, C. Senff, A. Langford, A. Weickmann, D. Law, J. Machol, D. Merritt, R. Marchbanks, S. Sandberg, W. Brewer, R. Hardesty, and R. Banta, “Development and application of a compact, tunable, solid-state airborne ozone lidar system for boundary layer profiling,” J. Atmos. Ocean. Technol. 28(10), 1258–1272 (2011). [CrossRef]
23. A. Fix, M. Wirth, A. Meister, G. Ehret, M. Pesch, and D. Weidauer, “Tunable ultraviolet optical parametric oscillator for differential absorption lidar measurements of tropospheric ozone,” Appl. Phys. B 75(2-3), 153–163 (2002). [CrossRef]
24. T. Refaat, S. Ismail, G. Koch, M. Rubio, T. Mack, A. Notari, J. Collins, J. Lewis, R. De Young, Y. Choi, N. Abedin, and U. Singh, “Backscatter 2-μm lidar validation for atmospheric CO2 differential absorption lidar applications,” IEEE Trans. Geosci. Rem. Sens. 49(1), 572–580 (2011). [CrossRef]
25. T. Shuman, R. Burnham, A. Nehrir, S. Ismail, J. Hair, and T. Refaat, “Efficient 1.6 micron laser source for methane DIAL,” Proc. SPIE 8872, 88720A (2013). [CrossRef]
26. J. W. Hair, C. A. Hostetler, A. L. Cook, D. B. Harper, R. A. Ferrare, T. L. Mack, W. Welch, L. R. Isquierdo, and F. E. Hovis, “Airborne high spectral resolution lidar for profiling aerosol optical properties,” Appl. Opt. 47(36), 6734–6752 (2008). [CrossRef] [PubMed]
27. S. Burton, R. Ferrare, C. Hostetler, J. Hair, R. Rogers, M. Obland, C. Butler, A. Cook, D. Harper, and K. Froyd, “Aerosol classification using airborne High Spectral Resolution Lidar measurements - methodology and examples,” Atmos. Meas. Tech. 5(1), 73–98 (2012). [CrossRef]
28. A. Scarino, M. Obland, J. Fast, S. Burton, R. Ferrare, C. Hostetler, L. Berg, B. Lefer, C. Haman, J. Hair, R. Rogers, C. Butler, A. Cook, and D. Harper, “Comparison of mixed layer heights from airborne high spectral resolution lidar, ground-based measurements, and the WRF-Chem model during CalNex and CARES,” Atmos. Chem. Phys. Discuss. 13(5), 13721–13772 (2013). [CrossRef]
29. E. Saiki, C. Weimer, and M. Stephens, “An investigation of high spectral resolution lidar measurements over the ocean,” Proc. SPIE 8159, 81590F (2011). [CrossRef]
30. R. Drever, J. Hall, F. Kowalski, J. Hough, G. Ford, A. Munley, and H. Ward, “Laser phase and frequency stabilization using an optical resonator,” Appl. Phys. B 31(2), 97–105 (1983). [CrossRef]
31. K. Numata, J. R. Chen, S. T. Wu, J. B. Abshire, and M. A. Krainak, “Frequency stabilization of distributed-feedback laser diodes at 1572 nm for lidar measurements of atmospheric carbon dioxide,” Appl. Opt. 50(7), 1047–1056 (2011). [CrossRef] [PubMed]
32. G. Ehret, C. Kiemle, M. Wirth, A. Amediek, A. Fix, and S. Houweling, “Space-borne remote sensing of CO2, CH4 and N2O by integrated path differential absorption lidar: a sensitivity analysis,” Appl. Phys. B 90(3-4), 593–608 (2008). [CrossRef]
33. L. Rothman, I. Gordon, A. Barbe, D. Benner, P. Bernath, M. Birk, V. Boudon, L. Brown, A. Campargue, J. Champion, K. Chance, L. Coudert, V. Dana, V. Devi, S. Fally, J. Flaud, R. Gamache, A. Goldman, D. Jacquemart, I. Kleiner, N. Lacome, W. Lafferty, J. Mandin, S. Massie, S. Mikhailenko, C. Miller, N. Moazzen-Ahmadi, O. Naumenko, A. Nikitin, J. Orphal, V. Perevalov, A. Perrin, A. Predoi-Cross, C. Rinsland, M. Rotger, M. Simeckova, M. Smith, K. Sung, S. Tashkun, J. Tennyson, R. Toth, A. Vandaele, and J. Auwer, “The HITRAN 2008 molecular spectroscopic database,” J. Quant. Spectrosc. Radiat. Transf. 110(9-10), 533–572 (2009). [CrossRef]
34. S. Drayson, “Rapid computation of the Voigt profile,” J. Quant. Spectrosc. Radiat. Transf. 16(7), 611–614 (1976). [CrossRef]
35. G. Anderson, S. Clough, F. Kneizys, J. Chetwynd, and E. Shettle, “AFGL atmospheric constituent profiles (0-120km),” Air Force Geophysics Laboratory, AFGL-TR-86–0110, 954 (1986).
36. L. Elterman, “UV, visible and IR attenuation for altitude to 50 km,” Air Force Cambridge Research Laboratories, AFCRL-68–0153, 285 (1968).
37. S. Ismail, R. Ferrare, E. Browell, S. Kooi, J. Dunion, G. Heymsfield, A. Notari, C. Butler, S. Burton, M. Fenn, T. Krishnamurti, M. Biswas, G. Chen, and B. Anderson, “LASE measurements of water vapor, aerosol, and cloud distributions in Saharan air layers and tropical disturbances,” J. Atmos. Sci. 67(4), 1026–1047 (2010). [CrossRef]
38. D. Bruneau, F. Gibert, P. H. Flamant, and J. Pelon, “Complementary study of differential absorption lidar optimization in direct and heterodyne detections,” Appl. Opt. 45(20), 4898–4908 (2006). [CrossRef] [PubMed]
