Design of a High Spectral Resolution Lidar for Atmospheric Monitoring in EAS Detection Experiments

Download Design of a High Spectral Resolution Lidar for Atmospheric Monitoring in EAS Detection Experiments

Post on 29-Jun-2016




1 download

Embed Size (px)


<ul><li><p>Design of a High Spectral Resolution Lidar for Atmospheric Monitoringin EAS Detection Experiments</p><p>E. Fokitisa, P. Fetfatzisa, A. Georgakopouloua, S. Maltezosa, and A. Aravantinosb</p><p>aPhysics Department, National Technical University of Athens,9, Heroon Polytechniou, ZIP 15780, Athens, Greece</p><p>bPhysics Department, Technological Educational Institution of Athens,Agiou Spiridonos, ZIP 12210, Athens, Greece</p><p>A High Spectral Resolution Lidar (HSRL) is designed in order to achieve high accuracy measuring the aerosol tomolecular scattering coecients ratio. This type of Lidar consists of a Continuous Wave Single Longitudinal Modelaser beam at 532 nm, a receiver with a parabolic mirror and analyzes spectrally the scattered light in a Fabry-Perot Cavity. The great wavelength sensitivity of this system allows the separation of the aerosol and molecularcomponent due to their dierent levels of Doppler eects on the scattered laser light. We present a study of theeects of the various levels of CCD cooling on the sensitivity of this. Firstly we discuss preliminary experimentalresults based on a Fabry-Perot etalon with free spectral range (FSR) 0.1 cm1, the expected performance of aetalon with FSR of 0.05 cm1, under construction, and with nally an aerosol parameter analysis in SimulationCodes for EAS.</p><p>1. INTRODUCTION</p><p>In this paper we present the work of our teamin the atmospheric monitoring issue for EAS De-tection experiments. In the measurements of air-uorescence yield for the detection of Ultra HighEnergy Cosmic Rays, there is a contribution dueto the air-Cherenkov signal scattered mainly bythe aerosol particles. Therefore, there is a need tomonitor the degree of this type of scattering in or-der to correct the total signal and obtain the air-uorescence signal. The most common methodfor measuring the aerosol induced scattering isusing the Elastic or Raman Lidar Method. Inthis work, we present the progress in an alter-native method using the High Spectral Resolu-tion Lidar (HSRL). This is continuation of recentwork [3], where now we emphasize on the studyof the performance of a candidate laser source forthe HSRL. This apparatus is designed in order toachieve high accuracy measuring the aerosol tomolecular scattering ratio. It consists of a Con-tinuous Wave Single Longitudinal Mode (SLM)Laser beam at 532 nm, a receiver with a parabolic</p><p>mirror, and analyzes spectrally the scattered lightpassing through Fabry-Perot cavities.</p><p>We mention that [1], rst reported a highspectral resolution lidar (HSRL) using a narrow-band laser and a high resolution Fabry-Perotetalon to separate the aerosol (Mie) and molecu-lar (Rayleigh) scattering. More recently, (see [2]),demonstrated aerosol and temperature measure-ments using a HSRL based on an iodine vapor l-ter, and the temperature in the stratosphere wasobtained from the Mie-ltered signal.</p><p>The HSRL can give simultaneously, for eachatmospheric height layer, both the aerosol andthe molecular scattered intensity from a groundbased laser emitter. This capability is due tothe interferometric measurement method, that is,the Fabry-Perot interferometer in the LIDAR re-ceiver. Additionally, in the aerosol scatteringHSRL, the narrow linewidth of the laser trans-mitter makes it possible both to separate the con-tributions from aerosol particles and atmosphericmolecules, and to calculate the aerosol extinctionwithout assuming the lidar ratio (dened as theratio between the extinction and backscattering</p><p>Nuclear Physics B (Proc. Suppl.) 190 (2009) 261265</p><p>0920-5632/$ see front matter 2009 Elsevier B.V. All rights reserved.</p><p></p><p>doi:10.1016/j.nuclphysbps.2009.03.097</p></li><li><p>coecients) [5]. We take advantage of the re-cent technological developments in the solid statediode lasers which have lead to cost aordableHSRL emitters.</p><p>In section 2 we describe the laser source usedin this HSRL. The section 3 deals with the char-acterization of the molecular channel to be usedfor determining the scattering coecient due tothe atmospheric molecules (3.1), while the aerosolchannel characterization is presented in 3.2. Thelaboratory emulation scattering studies of theSLM laser beam is presented in Section 4. Thesection 5 discusses the limitations in the sensitiv-ity of the HSRL prototype. The Section 6 consid-ers the prospects of the present research eorts.</p><p>2. Laser sources</p><p>2.1. SLM laserWe need a very narrow line-width laser source</p><p>so that the line source uncertainty is much smallerthan the Doppler broadening of the scattered ra-diation from the air molecules. Only in such casewe may separate the Mie from molecular scat-tering using a high spectral resolution spectrom-eter as the one we will be describing. We areusing a Diode Pumped Solid State Laser contin-uous wave (CW) SLM laser at 532 nm (calledSLM at the rest of the paper) with a nominalpower of 100 mW and coherence length greaterthan 50 m or correspondingly, wavenumber uncer-tainty k =0.02 cm1. The tests conducted wereaiming at studying the stability of the interfer-ence fringe patterns of the 5 cm spacer etalon asthe current driving the SLM laser gradually is in-creased to the nominal value. The results will bepresented in the Section about the aerosol chan-nel. However, the question of the absolute laserwavelength cannot be answered by such measure-ments. The measurements so far conducted haveindicated that at the nominal operation condi-tions, there is a dominant operation mode and asecondary mode with intensity of the order of 2-4 % of the one of the dominant mode as will beshown in 2.3. For the absolute wavelength (or fre-quency) determination, one would need a facilitywhich would include a laser frequency standard.However, we believe that the knowledge of the</p><p>absolute frequency is not necessary for the aimsof the HSRL.</p><p>2.2. Procedure for laser stabilizationThe specic solid state laser we use includes</p><p>some proprietary design, but basically, it con-sists of an optical cavity and some lasing medium,while the emitted line structure depends on thetemperature of the lasing medium as the tem-perature is a thermodynamic quantity which isimportant in aecting the longitudinal and trans-verse modes. For this reason, the SLM laser hasa controllable excitation current and we had tostudy the laser line structure as a function of thecurrent as well as of its rate of change. The useof the available Fabry-Perot spectrometers cangreatly help in revealing the laser line structureand its changes as a function of excitation current.In comparison with He-Ne mode competition, wehave the dominance of a single mode as discussed.The minimum level of the intensity in this plot isconsistent with the expected curve of Airy func-tion. No noise substraction has been performedsince the latter is expected to be negligible as theexposure time was around 1/4000 s.</p><p>3. Molecular and Aerosol channels</p><p>3.1. Molecular channelThe radiation approaching the receiver, after</p><p>being scattered by the air molecules, is typicallyreected by a beam splitter and is directed to themolecular channel, while the rest is transmittedto the aerosol channel. In order to characterizethe molecular channel we perform the followingmeasurements: We have obtained interferogramsof a platinum-neon (Pt-Ne) Hollow Cathode spec-tra lamp operating at ambient as well as at highertemperatures. We have also used an interferencelter at 562 nm, with a passband 30 5 nm inorder to select a specic spectral line of Pt-Nespectrum. In this range there are two line ofNe, namely at 585 and 588 nm, possibly someplatinum lines. The interferogram presented istaken with the 2 cm etalon spacer and the aboveinterference lter. The two intensity histogramsindicate that the rst one, corresponding to coldlamp, is less inuenced by the Doppler eect than</p><p>E. Fokitis et al. / Nuclear Physics B (Proc. Suppl.) 190 (2009) 261265262</p></li><li><p>Figure 1. The two intensity histograms indicatethat the rst one, corresponding to cold lamp,shows smaller Doppler Width than the one at thesecond histogram.</p><p>the one at the second histogram, where the tem-perature is higher. The result is presented inFig. 1. This plot is a display of the superpo-sition of the interference patterns correspondingto the atomic transitions involved. As the FSRis given by FSR = 1/2d, where d is the etalonspacer length, selecting an etalon of d=5 mm weare able to study eectively the molecular scatter-ing. Thus, we plan to make corresponding mea-surements with available etalons having spacerlengths 2 mm and 5 mm, respectively, which arecapable to contain the whole line-breath of thelaser radiation scattered and therefore broadenedby the air molecules.</p><p>3.2. Aerosol channelWe have studied in a laboratory the perfor-</p><p>mance of the aerosol channel using two dierentlasers, one at 632.8 nm, based on an available He-Ne laser, and the other was the SLM laser, thelatter having appropriate power and linewidth tobe used in a HSRL. Characterization of the longerspacer etalon, such as the one with d=5 cm, usingHe-Ne lasers can be nicely studied by observing</p><p>Figure 2. Modes of He-Ne Laser recorded with a5 cm spacer etalon and 110 cm focal length lensprojecting the interferometer output. On top ofgure, we show the result of a preliminary simu-lation of the intensity distribution.</p><p>the separation of the He-Ne longitudinal modes,as seen in Fig. 2. This plot is obtained withfocal length 110 cm after the etalon. The laserresonators length is 38 cm, so the mode spacingc/2L = 400 MHz. This corresponds to k around0.08 cm1. We observe that some fringes haveangular distance as small as 1/5 of the free spec-tral range. As we see in this Figure, such severalmodes are seen separated but there may also besome overlapping within the narrow free spectralrange of the etalon used. In the same gure weshow result of a simulation of the total intensitydistribution of the modes involved. This veriesthe High Spectral Resolution of the etalon. Thisresult should not have been taken for grantedsince the eective nesse of our interferometer de-pends on several factors which must combine to</p><p>E. Fokitis et al. / Nuclear Physics B (Proc. Suppl.) 190 (2009) 261265 263</p></li><li><p>Figure 3. The interferogram obtained with SLMlaser and 5 cm spacer length etalon, used for de-termining the etalon nesse.</p><p>achieve this result. Such factors are etalon platesatness, their parallelism, the etalon reectivity,the etalons angular nesse, the laser linewidth,the eect of externally caused mechanical vibra-tions etc. The nesse of the Aerosol Channel pa-rameter is quantied for the 5 cm spacer lengthetalon by using the SLM laser. The interferogramobtained with the SLM laser is presented in Fig.3. We observe that the interferogram is domi-nated by one mode when operating the laser atthe nominal power output of 100 mW. These datacorrespond to an overall nesse of 17.5.</p><p>4. Laboratory emulation the SLM laserbeam scattering</p><p>This was done by using the SLM laser beamexpanded by a small telescope. The scattered ra-diation was collected by a Newtonian telescopeof 250 mm diameter, and 1.6 meter focal length.The focused radiation was directed to the 5 cmspacer etalon combined with 110 cm focal lengthlens. The resulting interferogram shows thatthe nesse is again around 17.5. This resultcan be considered as proof-of-principle that theexperimental setup has the capability to mea-</p><p>sure at a resolution of FSR/ Finesse = 0.1 /17.5cm1=0.006 1. By using an etalon, under de-velopment, with FSR=0.05 cm1, we expect toimprove somewhat on the above achieved in thelaboratory spectral sensitivity, and therefore al-low the separation of the aerosol and molecu-lar scattering contributions via the two describedchannels. We have, however, to caution that theachievable sensitivity in eld measurements of at-mospheric aerosols may be worse than the oneachieve under the described lab conditions.For in-stance, one expects that the shot-noise level isexpected to be higher at atmospheric heights giv-ing low aerosol scattering signal, an also the nightsky background level may aect the sensitivity ofthe proposed setup. Finally, we are soon commis-sioning a CCD based system operating at Liquidnitrogen temperature so that the dark count con-tribution in the overall error budget is minimized.</p><p>5. Limitations on the sensitivity of HSRL</p><p>The sensitivity of the HSRL is mainly aectedby the Night Sky Background, shot noise, darkcount, read noise of each CCD pixel. The op-eration of the HSRL is foreseen only during thenighttime, corresponding to the period when theEAS Fluorescence Detector Telescopes are oper-ating. During this period, in the absence of ar-ticial UV radiation, the main limitation in theHSRL is due to the night sky background in thenear UV range, i.e. in the range 320-400 nm,which is the window of sensitivity of the EAS Flu-orescence Telescopes [4].</p><p>To get a feeling for the Night Sky Background,we may use a criterion that the NSB numberof photons recorded in a time interval of 1-5minutes, by the LIDAR receiver, is considerablysmaller than the noise of the receiver CCD (com-bined dark count and electronic noise) at thesame interval. For a typical thermoelectricallycooled CCD , the expected dark count is on theaverage 0.06 e-/pixel/electron, while it does notexceed 0.1 e-/pixel/s at 0 oC. The next require-ment is that the total rate of night sky back-ground plus dark count plus electronic noise ismuch smaller than the signal corresponding tothe SLM radiation scattered at a specic height</p><p>E. Fokitis et al. / Nuclear Physics B (Proc. Suppl.) 190 (2009) 261265264</p></li><li><p>corresponding to a scattering angle.</p><p>6. CONCLUSIONS AND PROSPECTS</p><p>In this work we have explored the possibil-ity to measure the aerosol scattering coecientby comparing to molecular scattering coecientrecorded at the same time. We tried a HighSpectral Resolution LIDAR method. The resultsdone at this phase, obtained by microparticlesdispersed in water, seem to retain the linewidthof the Laser beam. They show a k =0.02 cm1</p><p>as measured by a system combining a Newtoniantype telescope and a Fabry Perot Interferometer.</p><p>We are at the stage to implement the opticaldesign in permanent optical bench for both chan-nels, aerosol and molecular, in order to have a sta-ble receiver which will be able to move in variousdirections for operating in bistatic Lidar mode.</p><p>Also we consider SLM Laser at the UV Rangeas well as Pulsed Laser operation mode eldtest of this LIDAR will be done by recording,in bistatic mode, the scattering of the SLMlaser beam by aerosol and molecules near thePBL heights. We are developing an etalon withFSR 0.05 cm1, compatible for operation in thenear UV range which is appropriate for opera-tion of Fluorescence Detector telescopes. Thebiggest challenge towards this goal is to acquire aUV laser source with the correspondingly narrowlinewidth.</p><p>7. ACKNOWLEDGMENTS</p><p>This work has been funded by the projectPENED 2003. The project is co-nanced 80 %of public expenditure through C - European So-cial Fund, 20 % of public expenditure throughMinistry of Development - Gene...</p></li></ul>