Open Access
Add to BibSonomyAbstract
We developed an airborne compact rotational Raman lidar (CRL) for use on the University of Wyoming King Air (UWKA) aircraft to obtain two-dimensional (2D) temperature disman tributions. It obtained fine-scale 2D temperature distributions within 3 km below the aircraft for the first time during the PECAN (Plains Elevated Convection At Night) campaign in 2015. The CRL provided nighttime temperature measurements with a random error of <0.5 K within 800 m below aircraft at 45 m vertical and 1000 m horizontal resolution. The temperatures obtained by the CRL and a radiosonde agreed. Along with water vapor and aerosol measurements, the CRL provides critical parameters on the state of the lower atmosphere for a wide range of atmospheric research.
© 2016 Optical Society of America
1. Introduction
Temperature is a key parameter in the atmosphere and plays a major role in many important atmospheric processes. The vertical temperature structure controls the stability of the atmosphere and affects the development of clouds and the atmospheric boundary layer [1–3]. Because of the importance of vertical temperature distribution, many in situ and remote sensing technologies have been developed to provide spatially and temporally resolved temperature profiles. Recent reports and reviews from thermodynamic profiling technology workshops [4–6] have summarized the capabilities of current measurement systems and their indispensable roles in supporting atmospheric research and forecasting weather and air quality, but they have also revealed significant gaps in fulfilling existing needs.
Remote sensing technologies, including microwave radiometry (MWR), infrared (IR) radiometry, and lidar, offer better temporal coverage than in situ sampling. MWR and IR radiometry are widely used passive remote sensing systems that can operate day and night [7,8].
MWR observations are usually collected in the zenith direction at about five to ten frequency channels from 50 to 60 GHz to retrieve temperature profiles, with a root-mean-squared accuracy of ~0.6 K close to the surface and degrading to ~1.5–2 K in the middle troposphere [9,10]. Precision better than 0.5 K for the 1500 m closest to the surface can be achieved by combining multiple-angle observations [11]. The vertical resolution of an MWR measurement decreases rapidly from about 300 m at a height of 500 m to 500 m at a height of 1 km [10]. With such coarse vertical resolution, MWR can capture only a surface inversion and has difficulty capturing elevated or multiple inversions [11]. Absolute calibration of MWR is still challenging in that radiation absorbed by atmospheric gas must be corrected for in temperature retrieval [5,12,13]. In addition, because of the increased use of wireless communication, radio frequency interference has become problematic for MWR [5].
The temperature profile obtained by a typical IR radiometer is retrieved from observations of the 15- and 4.3-μm absorption bands of CO2. Use of the integrated calibration approach makes the accuracy of IR radiometry measurements better than 1% of the ambient radiance, so the IR radiometer is particularly well suited for long-term observations [14,15]. However, when the temperature profile is measured from the ground up, the vertical resolution of the IR radiometer is lower because all weighting functions peak at the surface [5,16]. Generally, high spectral IR measurements is able to resolve vertical temperature structures better than the MWR [16,17]. In addition, IR radiation from clouds above the IR radiometer will affect temperature measurements [5], a new algorithm was developed to retrieval profiles of temperature and water vapor, and cloud liquid path and effective radius for a single liquid cloud layer [18]. The complexity of temperature retrieval from MWR and IR radiometer measurements makes it difficult to quantify measurement errors and assimilate these measurements into a numerical model [5,6], 1-d variational retrieval approaches for MWR and IR radiometer were developed in recent years and thus provide full covariance matrices for retrievals [16,18]. Thus, data from these particular retrieval algorithms are more easily assimilated into numerical models.
Raman lidar, an active remote sensing system, is widely used to profile water vapor and temperature [19,20]. It uses the different temperature dependencies of two pure rotational Raman (RR) signals from atmospheric molecules to obtain atmospheric temperature profiles [21–23]. The high vertical and temporal resolutions of Raman lidar measurements make fine-scale temperature structure measurements possible. Raman lidar is a nearly ideal remote sensing system for data assimilation [24] because its straightforward retrieval method makes it easy to quantify the measurement error for each profile [25,26]. Long-term ground-based measurements have shown that Raman lidar can provide reliable daytime and nighttime temperature measurements [20,26]. However, measurements from ground-based vertical-pointing lidar are not able to provide temperature variability from the meso-beta to meso-gamma scale, which is needed to address recent advances in Earth system modeling and new requirements in weather and climate process studies [6]. Although ground-based scanning Raman lidar has the potential to provide spatial measurements up to 10 km [20], an airborne system is more suitable to provide such measurements.
Unlike elastic lidar [27–29], the number of airborne Raman lidar systems is limited because the Raman signal is very weak. An upward-looking airborne Raman lidar for measuring temperature in the lower stratosphere and upper troposphere using the Rayleigh backscattering signal and the vibrational Raman backscattering signal has been developed [30,31]. The airborne vibrational–rotational Raman (VRR) lidar has provided high-quality two-dimensional (2D) distributions of water vapor and aerosols [25,32]. However, no airborne Raman lidar system has yet provided highly resolved temperature measurements in the lower troposphere, where important energy and mass exchanges occur. Therefore, we still rely on multileg aircraft transects, which are inherently nonsimultaneous and limited by flight altitude, to provide fine-scale temperature variations [33]. This motivated us to add the capability of measuring temperature to our airborne Compact Raman Lidar (CRL) to provide simultaneous water vapor, aerosol, and temperature measurements.
In early 2015, the CRL was modified by the addition of two RR detection channels for temperature measurements. The updated CRL was deployed on the University of Wyoming King Air (UWKA) aircraft during the PECAN (Plains Elevated Convection At Night) campaign, which was a large, collaborative effort to explore the dynamics and microphysics of nocturnal MCSs (Mesoscale Convective Systems), convection initiation, and undular bores as they interact with the stable boundary layer and the low-level jet [34]. The PECAN project, which took place from 1 June to 16 July 2015 in and around Kansas, collected 120 h of airborne CRL data. The modified CRL is described in Section 2. Data processing is presented in Section 3. Observation examples from PECAN are discussed in Section 4, followed by a brief summary.
2. Lidar system
The four-channel CRL was initially developed to obtain 2D distributions of water vapor, aerosols, and clouds while deployed on the UWKA in 2010 [25]. In early 2015, additional channels to measure temperature were added to the receiver box in the CRL. Figure 1 is the schematic diagram of the updated receiver box. The five channels in the 12-in. × 18-in. × 3-in. receiver box include N2 and H2O VRR channels for water vapor measurements, low-J and high-J pure RR channels for temperature measurements (J is the rotational quantum number), and an elastic channel for aerosol and cloud measurements. The main system parameters of the CRL are listed in Table 1. In addition, the optical characteristics of the interference filters in these five channels are listed in Table 2, and spectral transmission curves of the filters in the low-J and high-J channels are presented in Fig. 2.
Table 1. Main System Parameters of the CRL
Table 2. Optical Characteristics of the Interference Filters in the CRL
Fig. 2 Transmission curves of the filters in the low-J and high-J channels. The rotational Raman backscattering spectra of N2 and O2 at 300 and 250 K and the elastic backscattering spectra are shown for comparison.
As illustrated in Fig. 1, the light collected by the telescope is divided into two beams by a beam splitter (BS1). The reflected beam, which includes VRR signals from nitrogen and water vapor molecules, is further split and guided into the N2 and H2O Raman channels by another beam splitter (BS2) and a reflector mirror (M1), respectively. The ability of the CRL to measure water vapor and aerosols in the daytime was demonstrated in early studies using narrow-band interference filters [25,35]. Because the PECAN measurements were performed during the nighttime, a 2.8-nm-bandwidth filter with a higher transmittance than narrow-band filters was utilized in the H2O Raman channel to increase its signal-to-noise ratio (SNR).
A polarizing beam-splitter cube (PBS) splits the beam that passes through BS1, which includes pure RR signals and elastic signals. Because the RR signal has a high linear depolarization ratio of 0.75 [20,23] and the depolarization ratio of the elastic signal from molecule and aerosol backscattering in the troposphere is usually <0.3 [36,37], and thus a greater fraction of the total RR signal and a smaller fraction of the total elastic signal go into the cross-polarizing component (with respect to polarization vector of the laser emission) than into the co-polarization component. To suppress the elastic signals in the low-J channel more effectively, the polarization vector of the laser emission is rotated to make sure that the cross-polarization component in the echo can pass through the PBS and go into the low-J channel. Because of the limited receiver box size, only the low-J channel, the central wavelength (CWL) of which is blueshifted from the laser line by ~0.7 nm, is located after the PBS to detect the cross-polarization component. In addition to the PBS, a wide-band filter (WF) (LL01-355-50, Semrock, Inc., Rochester, NY, USA) and two narrow-band filters (NF1 and NF2) (Materion, Mayfield Heights, OH, USA) are employed in the low-J channel (as illustrated in Fig. 1). The co-polarization component reflected by the PBS is split by BS3 and guided into the high-J channel and the elastic channel. An NF and a WF are used in the high-J channel.
The transmission curves of the NFs measured by the manufacturer is limited to ~0.1% due to the background noise of measurements (Fig. 2). Theoretical transmissions at 354.7 nm for the NFs in the low-J and high-J channels are 10−3 and 10−5, respectively. The transmission curves of the WFs were derived from typical normal incidence transmission and the refractive index provided by the manufacturer for the angles of incidence (AOI) listed in Table 2. The AOI on the WFs was optimized experimentally. The combined filters can provide 107 and 108 suppression of the elastic signal in the low-J and high-J channels, respectively. These suppressions are sufficient for temperature measurements in a cloud-free atmosphere [22].
3. Data processing
Pure RR signals received by the high-J and low-J channels can be written as [23]
where PH(r) and PL(r) are RR signals in the high-J and low-J channels, respectively; r is the distance from the scattering volume to the lidar; CH and CL are lidar constants for the high-J and low-J channels, respectively; OH(r) and OL(r) are the overlap functions of the high-J and low-J channels, respectively; N(r) is the number density of air molecules; tH(Ji) and tL(Ji) are the optical transmissions of the high-J and low-J channels at the Ji RR line, respectively; (dσ/dΩ)π(Ji) is the RR backscattering cross section at the Ji RR line; Ji is the rotational quantum number; ηi is the relative volume abundance of N2 and O2; τ(λ,r) is the atmospheric optical depth from the lidar to the scattering volume at wavelength λ; and λ0, λH, and λL are the transmitted laser wavelength, the high-J RR wavelength, and the low-J RR wavelength, respectively. Because λ0, λH, and λL are almost the same, the atmospheric optical depth at these three wavelengths can be considered almost the same.Because the dependence of the high-J and low-J RR backscattering cross sections on temperature trends in opposite directions, the temperature can be derived from the ratio Q(r) of the high-J and low-J channel signals, which is expressed as
The temperature dependence of the RR backscattering cross-section ratio can be expressed as [22,23]Thus, the simple relationship between the RR signal ratio and the temperature is expressed aswhere C is the ratio of the lidar constants of the high-J and low-J channels, g(r) is the ratio of the overlap functions of the high-J and low-J channels, A and B are constants, and T(r) is the atmospheric temperature in the scattering volume. The atmospheric temperature can be derived from Q(r) by rearranging Eq. (5):where a and b are calibration coefficients.All altitude-dependent items in Eqs. (1)–(6) are given as functions of r to simplify the expressions. The temperature at an altitude of z can be derived from T(r) by using the UWKA flight altitude z0 and the pitch and roll angle of the aircraft.
Equation (6) indicates that Q(r), g(r), a, and b must be determined to retrieve the atmospheric temperature. Therefore, the essential steps in our Raman lidar temperature-processing algorithm include preprocessing, determination of g(r), calibration, and uncertainty analysis, which are detailed in the following subsections.
3.1 Preprocessing data averaging
CRL raw data are normally saved as a 10-shot average via onboard averaging with a 0.6-m vertical resolution. Raw signals need to undergo further averaging to reach the required resolutions or SNR for temperature retrieval.
For the results reported in this study, we first obtained a 1-s average signal profile for each channel. The standard deviation of signal profiles measured in 1 s was considered the random uncertainty of the 1-s average signal while assuming the atmosphere was horizontally uniform during the 1-s measurement period. As the mean speed of the UWKA aircraft was ~90 m/s [27], the 1-s average signal profile corresponded to a horizontal resolution of ~90 m. Next, the 1-s average signal underwent a few basic preprocessing procedures, including ground surface detection, aircraft pointing correction, background subtraction, and further signal averaging to reach the required resolutions. For the given vertical and horizontal resolutions, N vertical sampling bins and M 1-s profiles were averaged. Different horizontal and vertical resolutions can be selected for the data analysis of different measurement targets. Random uncertainty in the averaged signal with the required resolutions is given as
where δP(r) is the random uncertainty in the averaged signal at the selected resolution and δP1 s(r) is the random uncertainty in the 1-s average signal.3.2 Determination of overlap difference between low-J and high-J channels
Although we can assume that the overlap function ratio g(r) = 1 for measurements within the full overlap range, it could be significantly different than 1 if the low-J and high-J channels have different short-range overlap functions. Equation (6) shows that g(r) needs to be determined to correct this effect. For ground-based Raman lidar, g(r) can be determined, in principle, by comparing the lidar-measured Q(r) profile with a “true” Q(r) profile derived from a temperature profile measured by a radiosonde collocated with the lidar [26]. However, the radiosonde will drift because of horizontal winds, which makes the comparison with the lidar challenging. With an airborne Raman lidar, the ability to change its operating altitude allows for the design of a multilevel flight along the wind and thus a better way to determine g(r), provided the temperature profile remains constant during this time period.
An example of determining g(r) from a designed flight is shown in Fig. 3. The flight occurred on a clear and calm night. The aircraft with CRL climbed to an altitude of ~3.1 km, flew a flat leg of ~90 km along the wind according to flight level measurements, then descended to ~2.5 km and flew back along the same ground-track at a level altitude. The RR signal ratio profiles Q(z) measured by the CRL in the two flat legs of the flight are shown in Fig. 3(a). The agreement of the two measurements between 1.8 and 2.2 km indicates that the air mass sampled during the two flat legs was the same. The difference between the two Q(z) profiles for altitudes between 2.2 and 2.5 km was used to determine g(r) for the Q(z) profile measured in the lower leg [red line in Fig. 3(a)]. The determined g(r), shown in Fig. 3(b), indicates that the difference between the overlap functions of the low-J and high-J channels for the ratio of RR signals measured within ~300 m needs to be corrected.
Fig. 3 (a) Rotational Raman signal ratio profiles Q(z) measured by the CRL during two flat legs at altitudes of 2.5 and 3.1 km. (b) g(r) determined from measurements shown in (a).
3.3 Calibration
The CRL temperature measurement was calibrated using an in situ temperature measurement taken separately on the UWKA. The calibration temperature should be collected during a spiral ascent or descent flight and in a clear sky to eliminate the effect on pure RR signals of elastic signal leakage caused by a heavy aerosol layer and clouds. In our study, the in situ temperature measurements were collocated with the CRL measurements of the RR signal ratio at a range of 450 m from the UWKA during the spiral flight, and these data were used to obtain the temperature profile. As shown in Fig. 3(b), at 450 m, g(r) = 1 and any potential impact of overlap correction is eliminated. Equation (6) can be written as
where T and Q are functions of altitude z. From Eq. (8), a and b can be determined from a linear regression between ln Q(z) and1/T(z), as illustrated in Fig. 4.In Eq. (8), the calibration coefficient a is an indicator of the sensitivity of the CRL temperature measurement, which is determined by the RR spectra, NFs, and laser wavelength. The wavelength of the third-harmonic YAG laser varies with the temperature of the YAG crystal at a rate of about 2 pm/K [38], as does the center wavelength of the NFs. Both temperature-dependent wavelengths shift in the same direction when there is a temperature change. The temperature of the YAG crystal and the filters during measurements is affected mainly by the temperature of the aircraft cabin, which is controlled to within a few degrees by an air-conditioning system. Therefore, the spectra of the received RR signal and a should not have been exposed to temperature variations. Thus, the slopes of the measurements taken on three days are practically the same, as seen in Fig. 4.
The calibration coefficient b is not only controlled by the RR spectra, but is also affected by the ratio of the lidar constants of the high-J and low-J channels [as seen in Eq. (6)], which is determined by the optical and electrical efficiencies in these two channels. Because the CWL of the low-J channel is located on the edge of the WF transmission curve (seen in Fig. 2), the optical efficiency of the low-J channel changes with the shift of the CWL of the WF. Simulation has shown that a 10-pm shift of the CWL of the WF in the low-J channel induces an ~0.9% variation of b, which will lead to a temperature error of ~3.5 K under a temperature of 300 K. Furthermore, the CWL of the WF varies with temperature by about <5 pm/K. Therefore, b slowly varies with the cabin temperature and needs to be taken into account to improve the accuracy of temperature retrieval. The differences between the intercepts of the three measurements in Fig. 4 were corrected by shifting each of the three measurements to result in the linear fit.
The a and b derived from Fig. 4 were used to estimate the temperature, while errors associated with temperature-dependent b were corrected by comparing temperatures measured by the CRL near the aircraft with those measured on the UWKA aircraft. To reduce the overlap effect, we used the CRL measurement at ~150 m from the UWKA, where g(r) was more stable. A lapse rate of 6.5 K/km was assumed to apply to the CRL measurements from 150 m below the aircraft to flight level. The difference between the CRL-estimated flight-level temperatures and the in situ measurements was used to correct the error introduced by variations in b. Because the temperature in the aircraft cabin changed slowly, the variation in b was also very slow. Thus, sharp variations in the temperature difference, which could be caused by random noise or a complex vertical temperature distribution, were eliminated by using a low-pass filter.
3.4 Uncertainty analysis
The uncertainty in the CRL-measured temperature is determined using the standard error propagation method with Eq. (6), which yields
where δX is the uncertainty of variable X. Uncertainty in the real-time correction of b was not included in Eq. (9). In addition, we smoothed the distribution of the retrieved 2D temperature using a mean filter with a 9 × 9-square slip window; thus, the random error of the smoothed temperature measurement was reduced to one-ninth of its original value.4. Case studies
Two CRL measurements from the PECAN campaign are presented and discussed here to illustrate the performance of the CRL.
4.1 Case 1: random errors and comparison with radiosonde measurements
A flight dedicated to collecting calibration data and comparison data was carried out in the cloud-free nighttime on June 30, 2015. The UWKA aircraft flew two level legs to collect data for the correction of the difference between the short-range overlap functions of the high-J and low-J channels, as illustrated in Fig. 3. Next, the UWKA flew level while approaching the fixed PECAN Integrated Sounding Array 3 site (PISA 3, 38.94 N, 99.56 W) and then performed a spiral descent and spiral ascent over the PISA 3. After the UWKA spiral descent and spiral ascent over the PISA 3, a radiosonde was launched from the PISA 3 to collect data for water vapor and temperature profiles at 0400 UTC.
The CRL temperature measurements collected with 1000-m horizontal and 45-m vertical resolution during the level leg while approaching the PISA 3 site are presented in Fig. 5(a). The spatial inhomogeneity of the temperature field is clearly seen in the figure. In the absence of solar heating, weak near-surface temperature inversions can be identified in the measurements, but they are highly localized. Measurement errors associated with the calibration and random noise were investigated for this leg. Figure 5(b) shows that the calibration uncertainty is <0.2 K. Uncertainties associated with the drifting of b were neglected because the cabin temperature was relatively stable during the high-altitude level flight. Figures 5(d)–5(f) show the random uncertainties obtained by using different vertical and horizontal averaging. Reducing the resolution can effectively reduce the random uncertainty in the temperature measurements, and random uncertainty increases quickly with increasing range because of the r–2-dependent decrease in the signal. The approximate ranges where the random uncertainty was <0.5 K for different resolutions in the nighttime are listed in Table 3. The random uncertainty of the temperature measured within about 800 m below the UWKA was <0.5 K for 1000-m horizontal and 45-m vertical resolution. Figure 5(c) shows the total uncertainty in the CRL temperature measurements, which within 3 km below the UWKA was <3 K using a 1000-m horizontal and 45-m vertical resolution and a 9 × 9 slip window mean filter. The sharp increase in the random uncertainty around 3:27 and 3:40 UTC in Figs. 5(d)–5(f) was caused by the aircraft making a turn. When the aircraft turns, the SNR at the same altitude will decrease because of the increased distance between the sampling point and the aircraft along the laser path.
Fig. 5 (a) Distribution of temperature measured by the CRL on June 30, 2015. (b) Calibration uncertainty, (c) total uncertainty, and (d)–(f) random uncertainties that arose from random noises in the lidar signals using different horizontal (ΔX) and vertical (ΔZ) resolutions.
Table 3. Approximate Range of the CRL Temperature Measurement with a Random Uncertainty of <0.5 K Under Different Horizontal (ΔX) and Vertical (ΔZ) Resolutions
The temperatures measured by the CRL (black line) and the Rosemount 102 probe on the UWKA (red line) during the ascent of the aircraft and the temperatures measured by radiosonde (blue line) are compared in Fig. 6(a). The CRL temperature measurements were derived from the 10-s average signals with a 45-m vertical resolution, which was ~3.8 km from the PISA 3; the error bars indicate the uncertainty of the measurement. Figure 6(a) shows that the vertical temperature in the three measurements had a consistent structure and that the CRL captured the near-surface inversion well. The difference between the CRL and the radiosonde measurements as a function of altitude is presented in Fig. 6(b), which shows that the differences at most of the sampling altitudes were <1 K. The altitude-dependent variation of the difference could result from the spatial variations of the temperature, as seen in Fig. 5(a). The comparison of the three temperature measurements demonstrates the reliability of the CRL in measuring temperatures within ~3 km below the aircraft.
Fig. 6 (a) Comparison of temperature profiles from the radiosonde (blue line), the Rosemount 102 probe on the UWKA (red line), and the 10-s average CRL measurement with a 45-m vertical resolution (black line). (b) Difference between the radiosonde and CRL temperature measurements (solid line), and 1σ error of CRL measured temperature (dash line).
4.2 Case 2: water vapor, aerosols, and temperature structure around a MCS
The MCS was an essential research target in the PECAN campaign. The UWKA track around the MCS is presented in Fig. 7(a) and CRL measurements of aerosols backscatter, water vapor, and temperature using 500-m horizontal and 30-m vertical resolution around the MCS are shown in Figs. 7(b)–7(d). The UWKA was sampling inflow as the MCS moved in a southwesterly direction. For safety reasons, the UWKA kept at least 5 miles from any detectable radar echo from the onboard precipitation radar. The different colored lines in Fig. 7(a) correspond to the different flight legs in approaching the MCS. Except for the black and red lines, the middle of each line indicates that the UWKA was approaching the MCS. The vertical distributions of water vapor and aerosols clearly indicate that inflow boundary-layer air was pushed upward close to the MCS because of the outflow generated by the MCS [39]. The air vertical velocity measurements taken on the UWKA aircraft and presented in Fig. 7(f) are in good agreement with the water vapor and aerosol structures. Further uplifting of the boundary-layer air usually leads to new generation of convection. A cold pool associated with cooling from evaporation is an essential part of the development and propagation of an MCS. Around 7:01, 7:25, and 7:40 UTC, the UWKA crossed cold pool boundaries, as indicated by the sharp temperature drops. However, because of the turbulent environment and operating at night, the UWKA aircraft had to make turns. Although the aircraft could not get deep into a cold pool, the CRL measurements offered the first fine-scale view of one near the leading edge of an MCS.
Fig. 7 (a) UWKA track in a mesoscale convective system (MCS). CRL measurements of (b) aerosol lidar scattering ratio (LSR), (c) water vapor mixing ratio, and (d) temperature distributions with 500-m horizontal and 30-m vertical resolution around an MCS observed on July 1, 2015. (e) Comparison of temperature measurements by a Rosemount 102 probe and the CRL on the UWKA. (f) Air vertical velocity measurement on the UWKA.
Figure 7(e) shows the similarity of the in situ and remote sensing temperature measurements taken by a Rosemount 102 probe on the UWKA and by the CRL at 30 m from the UWKA. Both measurements have the same temperature variation, and the correlation coefficient R2 = 0.97. However, there is a bias of about 1 K between the measurements that could be due to the vertical temperature gradient and the uncertainties in short-range g(r) correction. The in situ data showed that there is a temperature drop of up to 3 K across the cold pool edges. CRL measured temperature structures have even greater horizontal temperature gradients at low altitude and there are complex interactions across the cold pool boundary within the lower 500 m in the boundary layer.
5. Summary
In this paper, we presented an airborne compact rotational Raman lidar (CRL) deployed on the University of Wyoming King Air (UWKA) aircraft to measure temperature profiles below the aircraft. A combination of filters was used to suppress the elastic signal in pure rotational Raman channels. Measurements made during the PECAN campaign indicated the feasibility of the overall design of the CRL. We discussed the calibration of CRL temperature measurements and the correction of the near-range overlap functions between the low-J and high-J channels to highlight the flexibility of the airborne lidar system.
Measurements taken during the PECAN campaign served to illustrate the accuracy of the temperature measurements taken by the updated CRL and the potential for using the CRL for atmospheric boundary-layer observations. Our results showed that CRL temperature measurements taken within 800 m below the UWKA and at 45-m vertical and 1000-m horizontal resolution have random errors of <0.5 K. Comparisons with radiosonde and in situ measurements indicated that the CRL can provide 2D fine-scale temperature structure ~3 km below the UWKA at night. Measurements of temperature, water vapor, and aerosols around an MCS by the CRL demonstrated the potential of such airborne profiling in supporting a wide range of atmospheric boundary-layer research in the future.
There are some limitations to the CRL. Its lower laser power and smaller telescope limit the detection range and the accuracy of its measurements, especially during the daytime. A Multifunction Airborne Raman Lidar (MARLi) with a more powerful seed laser, larger telescope, and improved optical design that will improve our ability to perform airborne profiling of temperature, water vapor, aerosols, and clouds is under development.
Funding
National Science Foundation (NSF) (AGS-1337599 and AGS-1359645); National Natural Science Foundation of China (NSFC) (41405032).
Acknowledgments
Radiosonde data were provided by NCAR/EOL under the sponsorship of the National Science Foundation, http://data.eol.ucar.edu/. We thank Drs. Andreas Behrendt, Jens Reichardt, Paolo Di Girolamo, and David Whiteman for their support and advice during the development of our airborne temperature Raman lidar and the UWKA crew for their hard work during PECAN.
References and links
1. B. Stevens, D. H. Lenschow, G. Vali, H. Gerber, A. Bandy, B. Blomquist, J.-L. Brenguier, C. S. Bretherton, F. Burnet, T. Campos, S. Chai, I. Faloona, D. Friesen, S. Haimov, K. Laursen, D. K. Killy, S. M. Loehrer, S. P. Malinowski, B. Morley, M. D. Petters, D. C. Rogers, L. Russell, V. Savic-Jovcic, J. R. Snider, D. Straub, M. J. Szumowski, H. Takagi, D. C. Thornton, M. Tschudi, C. Twohy, C. Twohy, M. Wetzel, and M. C. van Zanten, “Dynamics and chemistry of marine stratocumulus-DYCOMS II,” Bull. Am. Meteorol. Soc. 84(5), 579–593 (2003). [CrossRef]
2. Y. P. Richardson, K. K. Droegemeier, and R. P. Davies-Jones, “The influence of horizontal environmental variability on numerically simulated convective storms. Part I: Variations in vertical shear,” Mon. Weather Rev. 135(10), 3429–3455 (2007). [CrossRef]
3. N. A. Crook, “Sensitivity of moist convection forced by boundary layer processes to low-level thermodynamic fields,” Mon. Weather Rev. 124(8), 1767–1785 (1996). [CrossRef]
4. Committee on Developing Mesoscale Meteorological Observational Capabilities to Meet Multiple Needs, National Research Council, Observing Weather and Climate from the Ground Up: A Nationwide Network of Networks (The National Academies, 2009).
5. R. M. Hoff, R. M. Hardesty, F. Carr, T. Weckwerth, S. Koch, A. Benedetti, S. Crewell, D. Cimini, D. Turner, W. Feltz, B. Demoz, V. Wulfmeyer, D. Sisterson, T. Ackerman, F. Fabry, and K. Knupp, “Thermodynamic Profiling Technologies Workshop report to the National Science Foundation and the National Weather Service,” http://nldr.library.ucar.edu/repository/collections/TECH-NOTE-000-000-000-853.
6. V. Wulfmeyer, R. M. Hardesty, D. D. Turner, A. Behrendt, M. P. Cadeddu, P. Di Girolamo, P. Schlüssel, J. Van Baelen, and F. Zus, “A review of the remote sensing of lower tropospheric thermodynamic profiles and its indispensable role for the understanding and the simulation of water and energy cycles,” Rev. Geophys. 53(3), 819–895 (2015). [CrossRef]
7. D. Cimini, T. J. Hewison, L. Martin, J. Güldner, C. Gaffard, and F. S. Marzano, “Temperature and humidity profile retrievals from ground-based microwave radiometers during TUC,” Meteorol. Z. 15(51), 45–56 (2006). [CrossRef]
8. W. F. Feltz, W. L. Smith, H. B. Howell, R. O. Lnuteson, H. Woolf, and H. E. Revercomb, “Near-continuous profiling of temperature, moisture, and atmospheric stability using the Atmospheric Emitted Radiance Interferometer (AERI),” J. Appl. Meteorol. 42(5), 584–597 (2003). [CrossRef]
9. S. Crewell, H. Czekala, U. Löhnert, C. Simmer, T. Rose, R. Zimmermann, and R. Zimmermann, “Microwave radiometer for cloud carthography: A 22-channel ground-based microwave radiometer for atmospheric research,” Radio Sci. 36(4), 621–638 (2001). [CrossRef]
10. J. C. Liljegren, S.-A. Boukabara, K. Cady-Pereira, and S. A. Clough, “The effect of the half-width of the 22-GHz water vapor line on retrievals of temperature and water vapor profiles with a 12-channel microwave radiometer,” IEEE Trans. Geosci. Remote Sens. 43(5), 1102–1108 (2005). [CrossRef]
11. S. Crewell and U. Löhnert, “Accuracy of boundary layer temperature profiles retrieved with multifrequency multiangle microwave radiometry,” IEEE Trans. Geosci. Remote Sens. 45(7), 2195–2201 (2007). [CrossRef]
12. U. Löhnert and O. Maier, “Operational profiling of temperature using ground-based microwave radiometry at Payerne: prospects and challenges,” Atmos. Meas. Tech. 5(5), 1121–1134 (2012). [CrossRef]
13. G. Maschwitz, U. Löhnert, S. Crewell, T. Rose, and D. Turner, “Investigation of ground-based microwave radiometer calibration techniques at 532 hPa,” Atmos. Meas. Tech. 6(10), 2641–2658 (2013). [CrossRef]
14. H. E. Revercomb, H. Buijs, H. B. Howell, D. D. Laporte, W. L. Smith, and L. A. Sromovsky, “Radiometric calibration of IR Fourier transform spectrometers: solution to a problem with the High-Resolution Interferometer Sounder,” Appl. Opt. 27(15), 3210–3218 (1988). [CrossRef] [PubMed]
15. R. O. Knuteson, H. E. Revercomb, F. A. Best, N. C. Ciganovich, R. G. Dedecker, T. P. Dirkx, S. C. Ellington, W. F. Feltz, R. K. Garcia, H. B. Howell, W. L. Smith, J. F. Short, and D. C. Tobin, “Atmospheric emitted radiance interferometer: Part II, Instrument performance,” J. Atmos. Ocean. Technol. 21(12), 1777–1789 (2004). [CrossRef]
16. W. G. Blumberg, D. D. Turner, U. Löhnert, and S. Castleberry, “Ground-based temperature and humidity profiling using spectral infrared and microwave observations. Part II: actual retrieval performance in clear-sky and cloudy conditions,” J. Appl. Meteorol. Climatol. 54(11), 2305–2319 (2015). [CrossRef]
17. U. Löhnert, D. D. Turner, and S. Crewell, “Ground-based temperature and humidity profiling using spectral infrared and microwave observations. Part I: simulated retrieval performance in clear-sky conditions,” J. Appl. Meteorol. Climatol. 48(5), 1017–1032 (2009). [CrossRef]
18. D. D. Turner and U. Löhnert, “Information content and uncertainties in the thermodynamic profiles and liquid cloud properties retrieved from the ground-based atmospheric emitted radiance interferometer (AERI),” J. Appl. Meteorol. Climatol. 53(3), 752–771 (2014). [CrossRef]
19. I. Mattis, A. Ansmann, D. Althausen, V. Jaenisch, U. Wandinger, D. Müller, Y. F. Arshinov, S. M. Bobrovnikov, and I. B. Serikov, “Relative-humidity profiling in the troposphere with a Raman lidar,” Appl. Opt. 41(30), 6451–6462 (2002). [CrossRef] [PubMed]
20. J. Reichardt, U. Wandinger, V. Klein, I. Mattis, B. Hilber, and R. Begbie, “RAMSES: German Meteorological Service autonomous Raman lidar for water vapor, temperature, aerosol, and cloud measurements,” Appl. Opt. 51(34), 8111–8131 (2012). [CrossRef] [PubMed]
21. J. Cooney, “Measurement of atmospheric temperature profiles by Raman backscatter,” J. Appl. Meteorol. 11(1), 108–112 (1972). [CrossRef]
22. Y. F. Arshinov, S. M. Bobrovnikov, V. E. Zuev, and V. M. Mitev, “Atmospheric temperature measurements using a pure rotational Raman lidar,” Appl. Opt. 22(19), 2984–2990 (1983). [CrossRef] [PubMed]
23. A. Behrendt, “Temperature measurements with lidar,” in Lidar: Range-Resolved Optical Remote Sensing of the Atmosphere, C. Weitkamp, ed. (Springer, 2005).
24. M. Grzeschik, H.-S. Bauer, V. Wulfmeyer, D. Engelbart, U. Wandinger, I. Mattis, D. Althausen, R. Engelmann, M. Tesche, and A. Riede, “Four-dimensional variational data analysis of water vapor Raman lidar data and their impact on mesoscale forecasts,” J. Atmos. Ocean. Technol. 25(8), 1437–1453 (2008). [CrossRef]
25. B. Liu, Z. Wang, Y. Cai, P. Wechsler, W. Kuestner, M. Burkhart, and W. Welch, “Compact airborne Raman lidar for profiling aerosol, water vapor and clouds,” Opt. Express 22(17), 20613–20621 (2014). [CrossRef] [PubMed]
26. R. K. Newsom, D. D. Turner, and J. E. M. Goldsmith, “Long-term evaluation of temperature profiles measured by an operational Raman lidar,” J. Atmos. Ocean. Technol. 30(8), 1616–1634 (2013). [CrossRef]
27. Z. Wang, P. Wechsler, W. Kuestner, J. French, A. Rodi, B. Glover, M. Burkhart, and D. Lukens, “Wyoming Cloud Lidar: instrument description and applications,” Opt. Express 17(16), 13576–13587 (2009). [CrossRef] [PubMed]
28. 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]
29. D. Bruneau, J. Pelon, F. Blouzon, J. Spatazza, P. Genau, G. Buchholtz, N. Amarouche, A. Abchiche, and O. Aouji, “355-nm high spectral resolution airborne lidar LNG: system description and first results,” Appl. Opt. 54(29), 8776–8785 (2015). [CrossRef] [PubMed]
30. W. S. Heaps and J. Burris, “Airborne Raman lidar,” Appl. Opt. 35(36), 7128–7135 (1996). [CrossRef] [PubMed]
31. J. Burris, W. Heaps, B. Gary, W. Hoegy, L. Lait, T. McGee, M. Gross, and U. Singh, “Lidar temperature measurements during the Tropical Ozone Transport Experiment (TOTE)/Vortex Ozone Transport Experiment (VOTE) Mission,” J. Geophys. Res. 103(D3), 3505–3510 (1998). [CrossRef]
32. D. N. Whiteman, K. Rush, S. Rabenhorst, W. Welch, M. Cadirola, G. McIntire, F. Russo, M. Adam, D. Venable, R. Connell, I. Veselovskii, R. Forno, B. Mielke, B. Stein, T. Leblanc, S. McDermid, and H. Vömel, “Airborne and ground-based measurements using a high-performance Raman lidar,” J. Atmos. Ocean. Technol. 27(11), 1781–1801 (2010). [CrossRef]
33. B. D. Sipprell and B. Geerts, “Finescale vertical structure and evolution of a preconvective dryline on 19 June 2002,” Mon. Weather Rev. 135(6), 2111–2134 (2007). [CrossRef]
34. B. Geerts, D. Parsons, C. L. Ziegler, T. M. Weckwerth, D. D. Turner, J. Wuman, K. Kosiba, R. M. Rauber, G. M. McFarquhar, M. D. Parker, R. S. Schumacher, M. C. Coniglio, K. Haghi, M. I. Biggerstaff, P. M. Klein, W. A. Gallus, B. B. Demoz, K. R. Knupp, R. A. Ferrare, A. R. Nehrir, R. D. Clark, X. Wang, J. M. Hanesiak, J. O. Pinto, and J. A. Moore, “The 2015 Plains Elevated Convection At Night (PECAN) field project,” Bull. Am. Meteorol. Soc.submitted.
35. P. T. Bergmaier, B. Geerts, Z. Wang, B. Liu, and P. C. Campbell, “A dryline in southeast Wyoming. Part II: airborne in situ and Raman lidar observations,” Mon. Weather Rev. 142(8), 2961–2977 (2014). [CrossRef]
36. D. Liu, Z. Wang, Z. Liu, D. Winker, and C. Trepte, “A height resolved global view of dust aerosols from the first year CALIPSO lidar measurements,” J. Geophys. Res. 113(D16), D16214 (2008). [CrossRef]
37. S. Groß, V. Freudenthaler, M. Wirth, and B. Weinzierl, “Towards an aerosol classification scheme for future EarthCARE lidar observations and implications for research needs: EarthCARE aerosol classification,” Atmos. Sci. Lett. 16(1), 77–82 (2015). [CrossRef]
38. S. Z. Xing and J. C. Bergquist, “Thermal shifts of the spectral lines in the 4F3/2 to 4Z11./2 manifold of an Nd:YAG laser,” IEEE J. Quantum Electron. 24(9), 1829–1832 (1988). [CrossRef]
39. S. F. Corfidi, “Cold pools and MCS propagation: Forecasting the motion of downwind-developing MCSs,” Weather Forecast. 18(6), 997–1017 (2003). [CrossRef]
