An Nd:YAG laser-based sodium temperature/wind lidar was developed for the measurement of the northern polar mesosphere and lower thermosphere at Tromsø (69.6N, 19.2E), Norway. Coherent light at 589 nm is produced by sum frequency generation of 1064 nm and 1319 nm from two diode laser end-pumped pulsed Nd:YAG lasers. The output power is as high as 4W, with 4 mJ/pulse at 1000 Hz repetition rate. Five tilting Cassegrain telescopes enable us to make five-direction (zenith, north, south, east, west) observation for temperature and wind simultaneously. This highly stable laser system is first of its kind to operate virtually maintenance-free during the observation season (from late September to March) since 2010.
© 2017 Optical Society of America
The polar mesosphere and lower thermosphere (MLT) region is a complex and important region, here the neutral atmosphere and ionospheric plasma interact, and the region is coupled with the lower atmosphere as well as the ionosphere. In order to advance our knowledge on the coupling between the neutral atmosphere and the ionosphere, it is essential to measure relevant neutral and ionospheric physical parameters simultaneously.
Sodium lidar has been playing an important role to probe the MLT region in the studies of the dynamics of atmospheric tides or gravity waves, etc [1–4]. With a great advancement of the laser technology and measurement techniques, current sodium lidars are capable of measuring not only sodium density variation but also background neutral temperature and wind speed simultaneously. Other kinds of resonance scattering lidars such as potassium lidars [5,6] or iron lidars  were developed especially for similar observations of the neutral atmosphere in the MLT region. Incoherent scatter (IS) radar is a powerful tool for ionospheric plasma observations, since the IS radar is able to measure electron density, electron and ion temperatures, and ion velocity with high time and altitude resolution. Togather, they present powerful, complementary measurements. However, to the best of our knowledge, there have been few common volume observations at high latitudes conducted by the IS radar and the lidar, except in the case of Sondrestrom facility, Greenland (66.9°N, 50.9°W) .
To make these kinds of simultaneous observations in the polar region, we installed a new sodium resonance scattering lidar at the European incoherent scatter (EISCAT) radar site, Tromsø, Norway (69.6°N, 19.2°E). Sodium resonant light (589 nm) is generated by summing 1064 nm and 1319 nm laser pulses from injection-seeded Nd:YAG lasers. The lidar is a substantial upgrade in many aspects from its progenitor, which operated at Syowa station, Antarctica between 2000 and 2002 [9,10]. Of importance for the new sodium lidar is a capability for temperature and wind velocity observations using the narrowband sodium lidar technique . Thus, using the new sodium lidar and the EISCAT radar, we are able to carry out simultaneous and common volume observations of fundamental neutral and ionospheric parameters, i.e. temperatures and radial velocities of neutrals with lidar and ions with radar, in the high-latitude MLT region. At a remote site, especially in the polar region, limited manpower is one of the biggest issues to undertake successful observations. So, the lidar should be reliable and based on proven technology.
In this paper we present an injection seeded, laser diode-pumped Nd:YAG laser-based sodium lidar in detail, including several design improvements to achieve robust, maintenance-free observation for temperature and wind measurements in the MLT.
2. Design concept and the layout of the newly-developed lidar
The lidar is based on sum frequency generation (SFG) of injection-seeded Nd:YAG lasers, developed for the sodium lidar previously installed at Syowa station, Antarctica, in 1998 , with the additional capability to measure line-of-sight winds . The idea of upgrading the Antarctic lidar was highlighted with (i) all solid-state laser source based on laser diode (LD) end-pumped Nd:YAG lasers , (ii) high laser power and high stability, (iii) capability of measuring mesospheric temperature and wind using a narrowband laser technique, (iv) production of continuous wave (cw) 589 nm light, from cw 1064 nm and cw 1319 nm Nd:YAG lasers , for Doppler-free spectroscopy for the absolute frequency monitor, and laser frequency locking and shifting, (v) simultaneous five-direction observation mode with five sets of receivers, and (vi) almost maintenance-free system.
In order to simultaneously measure both temperature and wind, we follow a narrowband sodium lidar technique established by the Colorado State University group as a proven technique . The technique incorporated a precise cw laser frequency control at 589 nm with an absolute frequency monitoring system, i.e., Doppler-free spectroscopy . To do that, we sum-frequency-generate single mode cw 589 nm light by combining portions of available cw 1064 nm and cw 1319 nm laser power though a periodically poled lithium niobate (PPLN) crystal with a “quasi-phase matching” period of 9.48 μm. To lock the cw 589 nm light at the peak of the Na D2a transition, νa, Doppler-free saturation spectroscopy with a heated sodium vapor cell is applied. The other portion of the seeder laser power is used for injection seeding of pulsed Nd:YAG systems. In addition, acousto-optic (AO) frequency modulators are inserted in the light path of the cw 1064 nm seeder laser to carry out fast and exact frequency shifts between ν1064, ν1064 + 630 MHz and ν1064 - 630 MHz (see Section 2.2 and Fig. 2.), giving rise to pulsed SFG outputs, respectively, at νa, ν+ = νa + 630 MHz and ν- = νa - 630 MHz. These cw sub-systems are collectively termed as the “Injection Seeder Unit” described in Section 2.2 below.
The pulsed Nd:YAG laser systems, injection seeded by appropriate single frequencies at 1319 nm and frequency-modulated 1064 nm, are end-pumped by laser diodes (LD) instead of flash lamps, as used in the previous system . Direct and efficient pumping of the Nd:YAG rods at 808 nm causes less thermal lens effect and it does not require much cooling water, unlike lamp-pumped Nd:YAG lasers. Efficient lasing enables high repetition rate, thus increase the system’s thermal stability. The new lidar employed the laser with repetition rate of 1000 Hz, i.e. pulse interval 1 msec. In that duration, the laser pulses complete a round trip from the ground to 150 km, passing through the sodium layer (80-115 km altitude). As shown in Section 2.1, “Pulsed 589 nm light source “, the pulsed 589 nm light is generated by SFG in an LBO crystal. Compared with the flash-lamp pumped Nd:YAG lasers with 40 mJ/pulse and 10 Hz at 589 nm used at Syowa station, Antarctica, the pulse energy of the new lidar is lower (4 mJ/pulse), however, the total power has become larger (4W) due to the 1000 Hz repetition rate. High power laser enables us to conduct five-direction observations with a set of receivers. The long lifetime LD pump laser is also a great advantage to minimize downtime.
The system parameters of the pulsed laser unit, along with injection seeder unit and receiver are summarized in Table 1. We first discuss in more detail the generation of 589 nm light source for lidar transmitter in terms of Pulsed laser unit (Section 2.1) and Injection seeder unit (Section 2.2). All the optical components are on the same optical bench (1.2 m (width) x 3.6 m (length) x 0.8 m (height)) including transmitting mirrors to the sky. We finish Section 2 with a discussion of the receiver.
2.1 Pulsed 589 nm light source
Figure 1 illustrates the pulsed laser unit, which was manufactured by MegaOpt Co., Ltd. The light sources employed are injection-seeded Nd:YAG lasers. A compact end-pump method is applied comprising Nd:YAG rods for 1064 and 1319 nm. Each oscillator for 1064 and 1319 nm employs two Nd:YAG rods. The pump sources are fiber-delivered quasi-cw LD lasers which directly pump the Nd:YAG rods at 808 nm. We adopt end-pumping instead of side-pumping for higher efficiency. Since the LD light sources are placed outside the oscillators, the oscillators are very compact. The LD light sources at a repetition rate of 1000 Hz deliver 200 micro sec duration pulses at 808 nm with a peak power up to 70 W. The Nd:YAG lasers for 1064 and 1319 nm produce linearly polarized output beams due to intracavity thin film polarizers (TFP). The TFPs are also used to inject the seed beams to the cavities. Pockels cells are used for Q-switched operation. By injection seeding from 1064 and 1319 nm narrowband lasers (Mephisto, by InnoLight, now provided by Coherent), the wavelength of the laser can be controlled. An LBO of length 25 mm is used as a nonlinear optical crystal for sum-frequency generation. Both surfaces of the crystal are AR-coated for 589, 1064 and 1319 nm. The LBO is housed in an oven and temperature-controlled at approximately 39.5°C. The Q-switched delay is individually adjusted for the 1064 and 1319 oscillators, causing temporal overlapping to yield the maximum output power at the sodium D2 line. With injection locking, we obtained 4.15W at the sodium D2 line (589 nm) with the 1064 nm power of 4.22 W and the 1319 nm power of 4.27 W. The conversion efficiency was 59.4%. The pumping LDs have not been replaced even though we ran the laser more than 10,000 hours. The Q-switch 1064 and 1319 nm lasers have pulse widths of 35 and 70 nsec, respectively, while the 589 nm has a pulse width of approximately 35 nsec. The measured beam quality factor (M2) of the pulsed 589 nm light is 1.5. After a collimation telescope, the 589 nm light is sent to the sky with a divergence of 0.2 mrad.
2.2 Injection seeder unit
Figure 2 illustrates the injection seeder unit. The seeder unit is comprised of three parts; sum-frequency generation (SFG) of the cw 589 nm, Doppler-free saturation spectroscopy, and the frequency shifter using AO devices in the path of the 1064 nm seed laser. Absolute frequency control of the laser is crucial for the temperature/wind observations with the resonance lidar.
To control the frequency of pulsed 589 nm light accurately, we monitor and control the frequency of the cw 589 nm light generated from 1064 and 1319 nm seeders . Narrowband, single-mode lasers from the seeders are collimated through a lens (f = 100 mm). After the polarization conversion from elliptical to linear with quarter-wave plates, the lasers go through Faraday isolators. A part of the cw 1319 nm laser (120 mW) goes to the pulsed 1319 nm Nd:YAG cavity and a part of the cw 1064 nm (200 mW) goes to an AO frequency shifter before injection seeding the 1064 nm pulsed Nd:YAG laser. The remaining cw 1064 and 1319 nm lasers focus on the AR coated PPLN at a temperature of 74 °C. The largest SFG output of 21 mW at 589 nm was derived from the 665 mW of 1064 nm and 320 mW of 1319 nm. The conversion efficiency was 2.1%. After we remove the fundamentals with a prism, part of the 589 nm light is monitored by a wavemeter (WSU-30, HighFinesse) that is always being calibrated with a frequency stabilized He-Ne laser. The other part of the 589 nm light, attenuated to ~0.3 mW goes through a sodium vapor cell heated at 60°C for the Doppler-free spectroscopic diagnostics. Example of scanned Doppler-free spectrum around D2a peak with ~1.0 MHz frequency step is shown in Fig. 3 in the GHz scale relative to the 32P3/2 -> 32S1/2 transition. Large dips at the D2a peak, enhancements at the crossover and dips at the D2b peak are clearly seen. During observations, the fine structure around the D2a peak (see figure insert) in a range of 150 MHz is scanned with ~1 MHz frequency resolution. We lock the frequency of the SFG output to the absolute sodium D2a peak frequency, at νa. This Doppler-free peak is re-scanned and the SFG output re-locked to it every hour throughout the observations. Between the repeated scans, the wavemeter (with 5 MHz accuracy) is used to monitor the 589 nm frequency to alert to laser mode hops. Consequently, by injection seeding, the pulsed 589 nm frequency can be locked to the D2a peak (νa) to within 1-2 MHz (neglecting frequency chirp).
To shift the frequency between νa, νa + 630 and νa-630 MHz, we built an AO frequency shifter using two AO frequency modulators (Brimrose) in the 1064 seed laser line and shifted the frequency of ν1064 by + 315 MHz and −315 MHz one way, leading to a round trip shifts of + 630 MHz and −630 MHz, as shown in Fig. 2. The one-way AO diffraction efficiencies at 1064 nm are about 60%. The direct output powers from the AO frequency shifter at ν1064, ν1064 + 630 MHz, ν 1064-630 MHz are 91.4 mW, 39.4 mW, and 30.9 mW, respectively. Since the pulse build-up time from the Q-switch trigger is a function of injection power, we rotate the λ/4 plates in the ν1064, ν1064 + 630 MHz and ν1064-630 MHz laser lines by different amount, thus tune the power of all laser lines to the same value of ~30 mW. This leads to equal pulse build-up times, resulting in the same seeding powers at all 3 frequencies. To obtain optimum temporal pulse overlap, the Q-switch for the 1319 oscillator is set to trigger before that for the 1064 oscillator. The electronics of the InnoLight seeders have a function of matching and stabilizing the cavity length by monitoring pulse build-up time.
All the receivers are placed in the receiver container which is 8 m from the laser transmitter container. Strong backscatter signal below 15 km altitude is not detected in this configuration. Backscattered photons from the five output beams are collected with five receivers. Each receiver consists of a telescope (Celestron C14; diameter = 355 mm, f = 3910 mm), collimating lenses, an interference filter (FWHM = 3 nm) and a photomultiplier tube (PMT) (Hamamatsu, H7421-40) as shown in Fig. 4. The receivers are assembled with custom-made tilting Aluminum frames, so that the pointing direction can be flexibly changed between the zenith and 30 degree from the zenith. When all the telescopes face the vertical direction, a power aperture product is comparable to 1.92 Wm2. Five-direction observations (vertical, north, south, west and east) have been performed since October 2012. Easy alignment of the laser/telescope axes can be done by comparing a real-time star chart with CCD camera images from both a finder telescope and the main telescope. By inserting a mirror between the main telescope and the PMT, the image through the main telescope can be monitored with another CCD camera. The 589 nm beam is first captured in a finder telescope with a wider field-of-view (FOV) and then it is easily guided to the narrow FOV (1 mrad) of the main telescope. Photocount data from the PMTs are recorded with a multichannel scalar (Sigma Space Co., AMCS-USB + ) which can be operated with LabView-based software. The signals are received with 0.64 μsec bins (96 m range resolution). The integrated photocount data is saved every 5 seconds, although the laser frequency is shifted by the AO frequency shifter every 1 minutes. The photocount linearity of the PMT was measured in the laboratory for the data calibration. A custom-made iris mask that cuts out received light trajectory to a FOV ~1 mrad, largely reduces background photon noise.
Due to the extremely cold weather in the winter polar region, temperature-controlled housing is quite important. The laser unit, receiving telescopes, control PCs are placed in three containers, respectively. The container size is 2.4 m (width) x 6.1 m (length) x 2.6 m (height). All the electronics, air conditioner, windows on the roof etc., were assembled with great care. The telescope apertures are exposed to the sky through flat glass windows. After doing environmental testing in the field, the containers were shipped to Norway. The inner temperature of the containers is kept between 17 °C and 23 °C through the year. Pictures of the lidar and the containers can be found in the web site .
3. Performance of the sodium lidar
After making several upgrades, our lidar system matured to become a robust instrument for simultaneous temperature and wind measurements. Based on the LD end-pumped Nd:YAG laser technology, the laser is highly stable and does not require much care for its maintenance. Currently observations are limited to polar winter, i.e. between late September and March. We optimize the system every summer, primarily the laser power, to prepare for the observation season. It is worth noting that during the 6-month observation period, we do not touch any optical or electrical laser component other than turning on and off the power supply. To our knowledge, thanks to the LD end-pumped Nd:YAG laser technology, this is the first maintenance-free resonance lidar system in operation. Thus the lidar observation does not require laser experts or highly skilled operators. No system trouble nor replacement of optical components has taken place since observation started in 2010. Daily decrease of the 589 nm power is vanishingly small. The pulse build-up time difference between the 1064 and 1319 nm laser pulses slightly increases during the observation period. Thus, a 10% power decrease occurs during the observation season (e.g. after about 2000 hours in the 2015 season).
Lidar observations in Tromsø began from October, 2010. In the first and the second seasons, we observed temperature and sodium density, and after the third season, wind observation was included. Five-direction observations started in 2012. The laser light was divided into five branches and directed to the sky through AR-coated and optical-polished container windows. Figure 5 shows photocount data received by one telescope when all the power was transmitted in the vertical direction. Strong signal from the sodium layer was detected although the measurement time was only 5 seconds. Despite the high repetition rate, the background level is quite low presumably because the iris mask works well. Temporal variations of sodium density, temperature, and zonal/meridional winds are shown in Fig. 6. The uncertainties of typical measurements induced by photon noise and laser locking fluctuation for the temperature and wind are estimated. With 0.5 km vertical and 10 min temporal resolutions the error bars are estimated to be 1.0 K and 1.5 m/s, respectively, at the sodium layer peak (e.g., 90 km), and 5 K and 10 m/s, respectively at both sodium layer edges (typically around 80 km and 105 km). Wind bias of about 5 m/s probably due to frequency chirp  is observed in hourly-mean vertical line-of-sight wind, which is generally accepted to be 0 m/s in the MLT region. However, it is not a problem for us because we always conduct vertical observations, from which the wind bias in all the directions can be subtracted. The details for retrieving temperature and wind have recently been thoroughly documented . Initial results for atmospheric science studies have already been reported [20–25]. Quick looks of all the lidar data can be found at the web site .
We have demonstrated a mature and robust temperature/wind lidar deployed at the EISCAT radar site in Tromsø (69.6N, 19.2E), Norway. The high power output of 4W at 589 nm is to the best of our knowledge the highest power among existing Na Doppler-resonance lidars, allowing efficient simultaneous observations in five Cardinal directions with high temporal resolution. The demonstrated system stability has allowed maintenance-free operation through the late September to March observation season since 2010.
Japan Society for the Promotion of Science (JSPS) KAKENHI (JP22403010, JP24310010, JP25287126, JP20360405, JP19651004); MEXT Special Funds for Education and Research (Energy Transport Processes in Geospace); MEXT Grant-in-Aid for Nagoya University Global COE Program
The authors gratefully acknowledge Dr. C.-Y. She for helpful advice and discussions, and Dr. David Krueger for sharing the data analysis program. The authors also thank Dr. Toru Takahashi for taking the lidar data and for making plots. The authors thank the three reviewers for their detailed reviews that improved the manuscript.
References and links
1. C.-Y. She, D. A. Krueger, R. Akmaev, H. Schmidt, E. Talaat, and S. Yee, “Long-term variability in mesopause region temperatures over Fort Collins, Colorado (41°N, 105°W) based on lidar observations from 1990 through 2007,” J. Atmos. Terr. Phys. 71(14–15), 1558–1564 (2009). [CrossRef]
2. C. S. Gardner and A. Z. Liu, “Seasonal variations of the vertical fluxes of heat and horizontal momentum in the mesopause region at Starfire Optical Range, New Mexico,” J. Geophys. Res. 112(D9), D09113 (2007), doi:. [CrossRef]
4. B. Clemesha, D. Simonich, and P. Batista, “Sodium lidar measurements of mesopause region temperatures at 23° S,” Adv. Space Res. 47(7), 1165–1171 (2011). [CrossRef]
5. J. S. Friedman, C. A. Tepley, S. Raizada, Q. H. Zhou, J. Hedin, and R. Delgado, “Potassium Doppler-resonance lidar for the study of the mesosphere and lower thermosphere at the Arecibo Observatory,” J. Atmos. Terr. Phys. 65(16–18), 1411–1424 (2003). [CrossRef]
6. J. Höffner and F.-J. Lübken, “Potassium lidar temperatures and densities in the mesopause region at Spitsbergen (78°N),” J. Geophys. Res. 112(D20), D20114 (2007).
7. X. Chu, Z. Yu, C. S. Gardner, C. Chen, and W. Fong, “Lidar observations of neutral Fe layers and fast gravity waves in the thermosphere (110-155 km) at McMurdo (77.8°S, 166.7°E), Antarctica,” Geophys. Res. Lett. 38(23), L23807 (2011).
8. C. J. Heinselman, J. P. Thayer, and B. J. Watkins, “A high-latitude observation of sporadic sodium and sporadic E-layer formation,” Geophys. Res. Lett. 25(16), 3059–3062 (1998). [CrossRef]
9. T. D. Kawahara, T. Kitahara, F. Kobayashi, Y. Saito, A. Nomura, C.-Y. She, D. A. Krueger, and M. Tsutsumi, “Wintertime mesopause temperatures observed by lidar measurements over Syowa station (69°S, 39°E), Antarctica,” Geophys. Res. Lett. 29(15), 1709 (2002).
10. T. D. Kawahara, C. S. Gardner, and A. Nomura, “Observed temperature structure of the atmosphere above Syowa Station, Antarctica (69°S, 39°E),” J. Geophys. Res. 109(D12), D12103 (2004).
11. C.-Y. She and J. R. Yu, “Simultaneous three-freqency Na lidar measurements of radial wind and temperature in the mesopause region,” Geophys. Res. Lett. 21(17), 1771–1774 (1994). [CrossRef]
12. T. D. Kawahara, T. Kitahara, F. Kobayashi, Y. Saito, and A. Nomura, “Sodium temperature lidar based on injection seeded Nd:YAG pulse lasers using a sum-frequency generation technique,” Opt. Express 19(4), 3553–3561 (2011). [CrossRef] [PubMed]
13. C.-Y. She, J. D. Vance, T. D. Kawahara, B. P. Williams, and Q. Wu, “A proposed all-solid-state transportable narrowband sodium lidar for mesopause region temperature and horizontal wind measurements,” Can. J. Phys. 85(2), 111–118 (2007). [CrossRef]
14. N. Saito, K. Akagawa, M. Ito, A. Takazawa, Y. Hayano, Y. Saito, M. Ito, H. Takami, M. Iye, and S. Wada, “Sodium D2 resonance radiation in single-pass sum-frequency generation with actively mode-locked Nd:YAG lasers,” Opt. Lett. 32(14), 1965–1967 (2007). [CrossRef] [PubMed]
15. J. Yue, C.-Y. She, B. P. Williams, J. D. Vance, P. E. Acott, and T. D. Kawahara, “Continuous-wave sodium D2 resonance radiation generated in single-pass sum-frequency generation with periodically poled lithium niobate,” Opt. Lett. 34(7), 1093–1095 (2009). [CrossRef] [PubMed]
17. S. Nozawa, “Sodium lidar pictures at Tromso,” http://www.isee.nagoya-u.ac.jp/~nozawa/LIDAR/index_lidaroperation.html
18. T. Yuan, J. Yue, C.-Y. She, J. P. Sherman, M. A. White, S. D. Harrell, P. E. Acott, and D. A. Krueger, “Wind-bias correction method for narrowband sodium Doppler lidars using iodine absorption spectroscopy,” Appl. Opt. 48(20), 3988–3993 (2009). [CrossRef] [PubMed]
19. D. A. Krueger, C.-Y. She, and T. Yuan, “Retrieving mesopause temperature and line-of-sight wind from full-diurnal-cycle Na lidar observations,” Appl. Opt. 54(32), 9469–9489 (2015). [CrossRef] [PubMed]
20. S. Nozawa, T. D. Kawahara, N. Saito, C. M. Hall, T. T. Tsuda, T. Kawabata, S. Wada, A. Brekke, T. Takahashi, H. Fujiwara, Y. Ogawa, and R. Fujii, “Variations of the neutral temperature and sodium density between 80 and 107 km above Tromsø during the winter of 2010-2011 by a new solid-state sodium lidar,” J. Geophys. Res. 119(1), 441–451 (2014). [CrossRef]
21. T. T. Tsuda, S. Nozawa, T. D. Kawahara, T. Kawabata, N. Saito, S. Wada, Y. Ogawa, S. Oyama, C. M. Hall, M. Tsutsumi, M. K. Ejiri, S. Suzuki, T. Takahashi, and T. Nakamura, “Decrease in sodium density observed during auroral particle precipitation over Tromsø, Norway,” Geophys. Res. Lett. 40(17), 4486–4490 (2013). [CrossRef]
22. T. T. Tsuda, S. Nozawa, T. D. Kawahara, T. Kawabata, N. Saito, S. Wada, C. M. Hall, S. Oyama, Y. Ogawa, S. Suzuki, T. Ogawa, T. Takahashi, H. Fujiwara, R. Fujii, N. Matuura, and A. Brekke, “Fine structure of sporadic sodium layer observed with a sodium lidar at Tromsø, Norway,” Geophys. Res. Lett. 38(8), L18102 (2011).
23. T. Takahashi, S. Nozawa, M. Tsutsumi, C. Hall, S. Suzuki, T. T. Tsuda, T. D. Kawahara, N. Saito, S. Oyama, S. Wada, T. Kawabata, H. Fujiwara, A. Brekke, A. Manson, C. Meek, and R. Fujii, “A case study of gravity wave dissipation in the polar MLT region using sodium LIDAR and radar data,” Ann. Geophys. 32(10), 1195–1205 (2014). [CrossRef]
24. T. T. Tsuda, S. Nozawa, T. D. Kawahara, T. Kawabata, N. Saito, S. Wada, C. M. Hall, M. Tsutsumi, Y. Ogawa, S. Oyama, T. Takahashi, M. K. Ejiri, T. Nishiyama, T. Nakamura, and A. Brekke, “A sporadic sodium layer event detected with five-directional lidar and simultaneous wind, electron density, and electric field observation at Tromsø, Norway,” Geophys. Res. Lett. 42(21), 9190–9196 (2015). [CrossRef]
25. T. Takahashi, S. Nozawa, T. T. Tsuda, Y. Ogawa, N. Saito, T. Hidemori, T. D. Kawahara, C. M. Hall, H. Fujiwara, N. Matuura, A. Brekke, M. Tsutsumi, S. Wada, T. Kawabata, S. Oyama, and R. Fujii, “A case study on generation mechanisms of a sporadic sodium layer above Tromsø (69.6° N) during a night of high auroral activity,” Ann. Geophys. 33(8), 941–953 (2015). [CrossRef]
26. S. Nozawa, “Sodium lidar data,” http://www.isee.nagoya-u.ac.jp/~nozawa/indexlidardata.html.