Документ взят из кэша поисковой машины. Адрес оригинального документа : http://cosrad.sinp.msu.ru/experiments/tus/doc/HO2COSPAR2006.pdf Дата изменения: Wed Mar 19 12:50:15 2008 Дата индексирования: Mon Oct 1 19:13:37 2012 Кодировка:

HO2.1 Updated TUS space fluorescence detector for study of UHECR. V. Abrashkina, V. Alexandrovb, Y. Arakcheeva, J. Cotzomic, A. Diazc , M. Fingerd, G. Garipove, V. Grebenyukd, N. Kalmykove, B. Khrenove, S.H. Kimf , P. Klimove, V.Kovala, O. Martinezc, S.W. Namg, D. Naumovd, A. Olshevskyd, M. Panasyuke, I.H. Parkg, J.H. Parkg, E. Poncec, A. Puchkovi ,C. Robledoc, A. Rosadoc, I. Rubinsteine, S. Sharakine, A. Silaeve, L. Tkatchevd, V.Tulupove, B. Sabirovd, H. Salazarc, O. Saprykini, L. Villasenorj, I. Yashine, N. Zaikind
a- Special Construction Bureau " Progress", Samara, Russia. b- Department of Applied Research of Moscow State University, Moscow, Russia c- University of Puebla, Puebla, Mexico. d- Joint Institute for Nuclear Research, Dubna, Moscow region, Russia. e- D.V. Skobeltsyn Institute of Nuclear Physics of Moscow State University, Moscow, Russia. f- Yonsei University, Seoul, Korea. g- EWHA Womans University, Seoul, Korea. i- Rocket Space Corporation "Energia", Consortium "Space Regatta". j- University of Michoacan, Morelia, Michoacan, Mexico. Abstract. The TUS (Tracking Ultra-violet Set up) space fluorescence detector has to be launched in 20092010 as a separated platform in Foton-4 mission prepared by the Samara enterprise. This detector was designed for another satellite and the updated variant of the TUS detector for a new platform is presented. The data on UV glow of the atmosphere obtained in operation of one pixel of the TUS detector on board the Moscow State University "Universitetsky-Tatiana" satellite was taken into account in design of the updated TUS detector. The data on UV transient flashes registered in "Universitetsky-Tatiana" mission are of special interest. Electronics of the TUS detector able to select and register different types of UV events in the atmosphere is presented. 1. Introduction. Since the previous COSPAR2004 meeting the launching plan of the Samara Special Construction Bureau " Progress" satellites has changed and the TUS (Tracking Ultra-violet Set up) mission (Abrashkin et al, 2006) is now planned for operation at the Small Space Apparatus (SSA) separated from the main Foton-4 satellite, due to be launched in 2009-2010. SSA is a new platform being designed for operation with space instruments having mass 50-100 Kg, power consumption of 60-100 Wt at the orbits of 500-400 km heights. The platform will be oriented in space due to a scientific task. In transportation mode SSA is placed above the Foton-4 body so that the TUS mirror could be accommodated in full size of 1.8 m diameter, Fig.1a. The TUS photo receiver is placed at the side of the SSA body (Fig. 1a) and in operation mode it is put to the focal plane of the mirror by an arm. Fig. 1b presents the TUS detector in the operation mode. In the new design the mirror- concentrator consists of 6 Fresnel and a central parabolic mirror segments. The mirror segments will be arranged on one plane of honey comb plastic base, Fig. 2a. In a new design the full mirror area is 2 m2, the focal distance is 1.5 m. As a second option another variant of mirror construction has been considered: the mirror consists of 19 hexagonal segments of 20 cm side, Fig. 2b. Segments are regular spherical mirrors 1


with 4 slightly different curvature radii. All of them are focusing light to one focal point. In Fig. 3 the efficiency of collecting light to one pixel is presented as a function of the off axis angle in both options. The photo receiver operates in the range =0-5o. One can see that in both options the efficiency decreases with angle and at =5o it is of about 50% but at small angles the first option has better resolution. The focusing characteristics of the constructed mirrors will give the decisive answer for an option choice. The photo receiver design is as in the previous case (Abrashkin et al, 2006). A special attention is paid for an accommodation of the small LIDAR transmitter at the back side of the receiver. LIDAR is designed as a combination of the TUS receiver and a low energy consumption transmitter at wavelength of 405 nm with energy in the laser pulse of ~5 mJ. The laser angular width is less than the pixel angular size so the beam might be directed only to area of the atmosphere covered by one pixel. The LIDAR task is to measure a distance between TUS and a ground or cloud spot which backscattered the EAS Cherenkov radiation. For operation of the LIDAR see Khrenov et al, 2005. The main feature of the design is use of MEMS technology scanning mirror controlled by the TUS computer, analyzing the recorded EAS data and directing the laser to the atmosphere spot, where back scattered Cherenkov light came from. After COSPAR2004 the UV detector comprising one pixel of the TUS receiver was launched on board of the "Universitetsky-Tatiana" satellite. The data of this detector presented below, section 2, on background UV light are taken into account in simulation of the UHECR registration. In simulation the light losses during light collection to the pixel were revised and new results on the UHECR energy threshold and accuracy in measurement of primary energy and direction are presented below, section 3. The data on UHECR from the "Pierre Auger Observatory" were taken when expected rate of UHECR events in the TUS detector was calculated. The "Universitetsky-Tatiana" data on UV transient flashes presented in section 2, opened a new field of interest for the TUS experiment- a search for mechanism of the transient luminous events (TLE) by study the time-space image of TLE at the initial faint stage of UV radiation. The TUS electronics was updated, section 4, for study of different types of UV events: EAS, TLE, meteors and sub-relativistic dust grains. 2. Results of the "Universitetsky- Tatiana" mission. The "Universitetsky- Tatiana" (below a short name "Tatiana" is used) UV detector is shown schematically in Fig. 4. In the TUS receiver the photomultiplier tube (PMT) of Hamamatsu type R1463 was selected as an UV sensor. At the entrance window the collimator and UV filter are mounted. Collimator puts limit to the field of view of the detector (15o). It also restricts the PMT's cathode area open for the light to S=0.4 cm2. The aperture of the detector is S=0.02 cm2 sr. The filter cuts the light with wavelength >400 nm. Quantum efficiency of the cathode is 20% in the range of wavelength 300-400 nm and goes down at smaller wavelengths. The effective wavelength range of the detector is 300-400 nm, limited by the filter at higher wavelengths and by the cathode quantum efficiency and absorption of light in the atmosphere- at lower wavelengths. The selected PMT has a high energy resolution and a single photo electron (p.e.) signal is well resolved. Before the detector operation the 1 p.e. signal was measured as a function of the tube voltage and this characteristic is used for measuring signal in p.e. number. A second PMT, identical to the first one but shielded from ambient light, has been also installed to evaluate the background light produced in the optical elements of the detector (filter, PMT glass) by charged cosmic ray particles. Signals from both PM tubes are coming to the multiplexer and then to 10 bits ADC (Fig. 4). The block- diagram of signal analysis is shown in Fig. 5. The main feature of the electronics is the use of FPGA (of XILINX type) for digital analysis of the signals after ADC (digital oscilloscope method). Several digital oscilloscopes are used 2


with different time samples for several tasks. Digital integration is used for selection and registration of various event types with a value of integration time determined by the experimental task: Task 1. Monitoring of UV intensity on-route. Time between measurements is 4 s, integration time in every point is 64 ms. Task 2. Monitoring of the noise signals from the charged cosmic ray particles. Time interval is 32 s and integration time in every point is 1 s. Task 3. Registering short UV flashes by 2 oscilloscopes: 1) with the trace 4 ms, time sample 16µ s, integration time for event selection 256 µ s and 2) with the trace 64 ms, time sample 256 µ s , integration time for event selection 4 ms. The number of time samples in both oscilloscope traces is 256. The photomultiplier tube, registering UV light, is working with a constant anode current (the ADC code is equal to 128). Automatic control circuit controls the voltage (gain) of PMT on-line with the illumination of the tube (relaxation time 1 s). The PMT is operating all time: either at night or day side of the Earth. At day side the gain is less than 100 but at the darkest locations at night side the gain is at its maximum equal to 3 106. Two codes are recorded and sent as a result of measurement to the mission center: M- the PMT's voltage code and N- the ADC code, for details see Garipov et al, 2006. In Fig. 6 the example of the signal registration in task 1 (monitoring on-route UV light from the atmosphere) is presented. One can see that ADC code N is kept approximately constant (~128) and the UV intensity is determined by PMT voltage code M (see the scale on the right Yaxis). When the PMT gain is at maximum, UV intensity is determined by the ADC code N (an example of this regime is indicated by in Fig. 6). This method gives us a possibility to operate the UV detector in very wide range of UV intensity- of about 106. In Fig. 7 UV intensity at moonless night side of the Earth is presented as a function of time on one of the satellite circulation. In Fig. 7 the measurement of the UV light from the cities are also presented. In the first circulation the satellite crosses Mexico City () and Houston (). In the second circulation it crosses Los-Angeles (). The corresponding UV intensity for them (minus the natural intensity observed before, above the ocean) is: - 8 107 photons/cm2 s sr, - 108 photons/cm2 s sr and 1.5 108 photons/cm2 s sr. In Fig. 8 the UV intensity measured at a full moon night is presented. At a moonlit night a large variation of the UV intensity was registered which is correlated with the cloud cover variation. In moonlit nights the UV intensity depends on the moon phase, its local zenith angle. Average UV intensity varies from 108 to 3 109 photons/cm2 s sr. It is over 109 photons/cm2 s sr only in 20% of the night time of the lunar month. The moon scattered light could be used as standard light source for checking the stability of the photo sensors. In 2 years of the Tatiana operation the efficiency of selected PM tubes are stable in error of 10%. Twice a year the satellite comes to high latitudes (>60o) at night side of the Earth. At latitude ~70o an aurora lights was measured in the UV range. In Fig. 9 one of such measurement is presented. Typically the aurora UV intensity is of the order of the moon scattered light but the intensity of aurora light is changing with the latitude and solar activity. The data from the "blind" PMT has shown that measurable signals were observed only when the satellite went through the "South-Atlantic (Brazilian) Anomaly" where the intensity of cosmic ray particles (electrons) was about 104 particle/cm2 s sr. Even in this region the noise from particles is less than 10% of the UV light background signal from the atmosphere at moonless night. For the TUS pixel which will have higher intensity of a background light from the atmosphere due to large area of the mirror the "particle noise" will be negligible. The use of the digital oscilloscope allowed us to select and to record the temporal profiles of transient UV flashes. The flash-event finding algorithm is based on selection of an integrated in time ti signal with the largest amplitude (charge). Integration is done in digital form: the signal value is summed in number "n" of ADC time samples ts (n=ti / ts ). From this value the previous 3


value (also summed in "n" samples) is subtracted at every ADC step. The obtained difference value is recorded in RAM. If the new difference is larger than the previous one the data of the oscilloscope trace (256 time samples) is recorded in the RAM data section. At the end of a given period of time (in "Tatiana" case at the end of night side satellite circulation) the RAM data is retranslated to the main computer memory and then are sending to the mission center. As was mentioned before, two oscilloscopes are operating simultaneously. In Fig. 10 and 11 examples of UV pulses (flashes) selected by the first and the second oscilloscopes are presented. The number of photons in temporal profiles of Fig. 10 and 11 are of the order of 104 and 105 respectively. The number of UV photons radiated in the atmosphere flashes at distance 1000 km from the detector was estimated as 1022 -1023 photons, i.e. the energy radiated in near UV range (300-400 nm) is of the order of 0.01 -0.1 MJ. In some flashes with the saturated signals this energy is much larger- up to 1MJ. Geographical coordinates were associated to the measured UV flashes and their distribution over the globe is shown in Fig. 12. Registered UV flashes are concentrated in the equatorial region (50 flashes were registered at latitudes ±10o among 83 events at latitudes ±60o). Their average rate is of about 1 flash per 2 circulations. Analysis of the UV flash temporal profiles, their energies and global distribution has shown (Garipov et al, 2005b, Garipov et al, 2005c) that most of registered UV flashes are not the lightning but transient luminous events (TLE), presumably Elves and Sprites. The role of UV flashes of lower than registered energies, as a noise in search for UHECR events, were not investigated and has to be searched in the TUS mission. 3. Expected TUS performance in registration of different phenomena in the atmosphere. Simulation of the UHECR event registration was improved to compare with the previous results (see Abrashkin et al, 2006). New results include the increased area of the mirrorconcentrator, its focusing parameters, light collection efficiency by light guides, losses of light in reflection at the mirror surface and light guide surfaces. Efficiency of event selection by triggered several pixels for zenith angles in the range =50o -90o and primary EAS energies 40100 EeV was tried in simulation of electronics operation. Event selection of "vertical" EAS (<50o) having a signal from scattered back Cherenkov radiation was also tried. Attention was paid to a possibility of distinguishing the horizontal and up- going EAS from neutrino and horizontal EAS initiated by cosmic rays protons and nuclei. Preliminary estimates of the duty cycle in measuring cosmic ray particles show that EAS, initiated by protons and nuclei with zenith angles >50o , will be registered with the duty cycle ~20% . Vertical EAS will be registered with less duty cycle ~10%. EAS initiated by the electron neutrino will be registered with duty cycle ~7-10% (cloud cover closes the most effective atmosphere target for neutrino interaction). EAS from tau- neutrino will be registered with duty cycle 10-20% (EAS from tau- neutrino are mostly come in trajectories, skimming the Earth and EAS maxima are expected to be above the cloud). An interesting advantage of the neutrino induced EAS observation from space is a possibility to compare the rate of event above ocean and ground which could give an estimate of neutrino interaction cross-section, see PalomaresRuiz ,et al, 2006. Particle flux intensities in 30-50 EeV energy range measured in the "Pierre Auger Observatory" experiment (Sommers, 2005) and in the HiRes experiment (Sokolsky, 2006) have been used to evaluate the UHECR rate in the TUS detector. Extrapolation of the energy spectra above 50 EeV has been carried both under hypothesis of a constant power law index =1.84 (i.e. no cut off) and with a steep index =3 for E>50 EeV (as favored by HiRes data). Results are presented in Table 1.

4


At the orbit height 500 km, in the range of zenith angles =60o-90o for duty cycle 20% the TUS geometrical factor is 1000 km2 sr per year. In the range of zenith angles <60o for duty cycle 10% the geometrical factor is 1500 km2 sr per year. Expected full rate of particles with energy threshold E=50 EeV is 57 particles per year for the case of no cut-off and 35 particles per year for the case of cut-off. To justify the existence of cut-off the TUS detector has to operate during 3 years when the full statistics of particles over threshold energy 200 EeV will be 14 (no cut-off) and 2 (cut-off). A long operation is also important for search of the neutrino induced EAS. Operation in 3 years is possible if the starting orbit of SSA is not less than 500 km height. TLE events of energy 103 -105 erg in UV radiation shall be certainly the most frequent events registered by TUS (in every satellite circulation several TLE events of such energy is expected). At the early stage of TLE the initiating EAS may be revealed as it was predicted by Gurevich et al, 1999. Observation of micro meteors and, possibly, of sub-relativistic dust grains are anticipated by the TUS detector, Khrenov and Stulov, 2006. They also might be observed in the mountain experiment at the Mexico site, see section 5. 4. Electronics of the updated TUS detector. As mentioned above the TUS detector is aimed for study of different in nature signals with time scale variable from microseconds (back scattered Cherenkov light, fluorescence light from EAS) to hundreds of millisecond (TLE UV radiation, micro meteor UV emission). The adequate electronics was designed for registration of event images in various temporal scales. The general approach is that the analog front-end electronics is measuring all signals with a needed for EAS short integration time ~1µ s. Those short signals are converted to digital form in ADC with time sample equal to the above integration time. Then slower signals are selected and analyzed in the digital part of the electronics. In every pixel different type of expected phenomenon are observed by sampling signals in different time intervals: 0.8 µ s for EAS, 16 µ s for TLE and subrelativistic dust grains, 256 µ s for longer TLE, 4 ms for meteors. The data is finally recorded in a standard number of samples: 256. In correspondence the trace duration is 205 µ s, 4 ms, 65 ms and 1 s. In Fig. 13 the principal schematic of the electronics for 256 pixels, divided to 16 PMTs clusters, is shown. From every pixel a signal goes to the multiplexer controlled by the generator with frequency 1/0.8x16=20 MHZ which is sampling the operation of ADC. One ADC is used for 16 pixels. Digital information goes to FPGA where every signal is considered by the demultiplexer restoring the pixel address. Then signals go to 4 oscilloscopes with different time samples. Signal in the oscilloscope time sample is a sum of signals in the original time samples of 0.8 µ s. Every time sample the signal summed in 16 oscilloscope samples (in this time the disc of a horizontal EAS crosses the TUS pixel) is compared with the noise level. If the summed signal is times "a" greater than pixel noise "" the signal is identified as the 1-st stage trigger and is sent to the second selection stage where signals from all TUS pixels are analyzed. Here the 1st triggered pixels are considered in the pixel map. If "n" neighbor pixel signals are in coincidence during coincidence time "t", a command for recording the data from 256 time samples (128 samples before the trigger time and 128 samples- after) in all pixels is produced. Four oscilloscopes are triggered separately- by a trigger condition selected for every oscilloscope. The oscilloscope parameters (oscilloscope time sample, integration time) and the trigger conditions (values "a", "t" and "n") are given in FPGA but it could be changed by commands from the mission center during the flight. For selection of vertical EAS (1-2 pixel triggered) the condition of high signal in one time sample of 0.8 µ s (back scattered Cherenkov signal) is also applied. Oscilloscopes are working permanently but with various event rate as the PMT's voltage (gain) is controlled by the UV light level at the tubes cathodes. The average ADC signal is kept at a given level (this level is controlled from the mission center, as a preliminary value the ADC 5


average code is put equal to 128). At the day side gain of the tubes is very low and triggering by the selected conditions is not expected. At the night side the gain come up to a regular "operation" level and the TUS detector start to select useful events. In this design the TUS detector will operate as at moonless nights so at full moon nights. In the last case the energy threshold will be higher as the pixel noise is higher. In the designed electronics there is a chance to register simultaneous triggers of several oscilloscopes when the event of long bright light pulse triggers not only oscilloscope with a long trace but the oscilloscopes of shorter traces as well. Correlation between different phenomena could be revealed if the time consequence of triggers is analyzed. For example, the EAS trigger may be followed by the TLE trigger- in this case a delay between TLE and EAS is expected >100 µ s. In false triggering an "EAS trigger" or even several "EAS triggers" will arrive close in time to the brightest stage of TLE event. 5. Mountain detectors of the TUS type in Mexico.

At Cerra La Negra at the altitude of 4.5 km the fluorescence detector of the TUS type is being prepared for operation. The mirror- concentrator is a sum of segments presented in Fig.2b but number of segment is 37 and the mirror area is 3.8 m2. At focal distance ~1.5 m from the mirror plane the photo receiver of pixels with size 1.5 cm will cover FOV of the detector. Today the pixel number is 64 but it will be enlarged later to 256 pixels. Design of this detector is very close to the TUS detector, one of the construction differences is that mirror segments are made as glass mirror (not the carbon plastic as in TUS). Comparatively large mirror area allows us to look in near horizontal direction for EAS tracks at distances up to 60 km. At altitude 4.5 km due to high atmosphere transparency in horizontal direction 30% of EAS beam light would reach the detector. EAS of energies in the range 1-10 EeV will be registered with their position of cascade maximum inside of FOV at distances 20-60 km. The area of the atmosphere where the EAS could be registered in the range of those distances is ~200-250 km2. For estimated average EAS energy threshold of 1 EeV, the rate of EAS is expected to be ~1 events per hour for EAS zenith angles >60o. For those zenith angles position of EAS maximum is above the horizon. The main problem in this experiment is an achievement of satisfactory accuracy in the directional measurements in axial plane of the detector. Measurements of EAS angular velocity in several pixels will be used for determination of the angle in axial plane. Energy spectrum in the range of 1-10 EeV is interesting as in this range the change of spectrum exponent is observed in several experiments which may indicate a change in cosmic rays origin from Galactic to extragalactic. The same detector looking upward (elevation angle ~45o) could register micro meteors and sub-relativistic dust grains producing signals at altitudes in the atmosphere of 90-120 km. In this direction the transparency of the atmosphere for detector at the altitude 4.5 km is excellent (80% of beam light coming to the detector) and the efficient area of the atmosphere for those observations is ~1000 km2. The expected range of kinetic energies of micro meteors and subrelativistic grains in the mountain experiment is discussed in Khrenov and Stulov, 2006. This kind of experiment in the mountains will help for planning observation of the same objects by the TUS space detector at much larger area in the atmosphere.

6. Conclusion. A multi- purpose detector TUS is being prepared for launching in 2009-2010. It will give space-time images of the objects radiating UV light in the wavelength range 300-400 nm. The mirror- concentrator of area 2 m2 will be used for measuring very faint signals. The detector will measure images with a lateral resolution in the atmosphere 4-5 km (from orbits of height 400500 km) and time resolution depending on the origin of radiation source: for UHECR event 6


resolution is ~1 µ s, for TLE or sub-relativistic dust grains- resolution is of 16-256 µ s and for micro meteors- 4 ms. Acknowledgement. The work done on the TUS project is supported by the Federal Space Agency of RF. References. 1. Abrashkin V, Alexandrov V., Arakcheev Y., et al, The TUS space fluorescence detector for study of UHECR and other phenomena of variable fluorescence light in the atmosphere. Advances in Space Research, 37 (2006) 1876-1883. 2. Garipov G.K., Khrenov B.A., Panasyuk M.I., Tulupov V.I., Salazar H., Shirokov A.V., Yashin I.V., UV flashes in the equatorial region of the Earth. JETP Letters, 82 (2005a) 185-187. 3. Garipov G.K., Khrenov B.A., Panasyuk M.I., Tulupov V.I., Salazar H., Shirokov A.V., Yashin I.V., UV radiation from the atmosphere: Results of the MSU "Tatiana" satellite measurements. Astroparticle Physics 24 (2005b) 400-408. 4. Garipov G.K., Khrenov B.A., Panasyuk M.I., Rubinshtein I.A., Tulupov V.I., Salazar H., Shirokov A.V., Yashin I.V., UV radiation detector of the MSU research educational micro satellite "Universitetsky-Tatiana". Instruments and Experimental Techniques, 49 (2006) 126-131. 5. Gurevich A.V., Zybin K.P. and Roussel-Dupre R.A., Lightning initiation by simultaneous effects of runaway breakdown and cosmic ray showers. Phys. Lett. A 254 (1999) 79-87. 6. Khrenov B.A., Park I.H. and Salazar H., Detection of scattered Cherenkov radiation in cosmic ray observation from space. NIM A, 553 (2005) 304-307. 7. Khrenov B.A. and Stulov V.P., Detection of meteors and sub-relativistic dust grains by the fluorescence detectors of ultra high energy cosmic rays. Advances in Space Research, 37 (2006) 1868-1875. 8. Sokolsky P., Recent results from the HiRes Fly's Eye experiment, talk at the "Cosmic Vision" workshop, WWW.roma2.infn/uhe_workshop06 9. Sommers P., First estimate of primary cosmic ray energy spectrum above 3 EeV from the Pierre Auger observatory, 29-th ICRC (Pune) usa-sommers-p-abs1-he14-oral (2005). 10. Palomares-Ruiz S., Irimia A. and Weiler T.J. Acceptances for space-based and ground ­ based fluorescence detectors, and inference of the neutrino-nucleon cross-section above 1019 eV. Phys. Rev. D 73 (2006) 083003-1-083003-32.

Table Caption. Table 1. Integral UHECR intensity used in estimates of the TUS detector performance.

7


Figure Captions.

Fig.1 TUS detector at SSA: a. in transportation mode, b. in full size. Fig.2. The TUS mirror- concentrator: a. segmented Fresnel mirror, b. segmented spherical mirror. Fig.3. Efficiency of the mirror light collection to one pixel as function of the off- axis angle. Triangles- mirror option 1, squares- mirror option 2. Fig.4. Schematic of the UV light registration on board the "Tatiana" satellite. Fig.5. Electronics of the "Tatiana" UV detector. Fig.6. Raw data from the UV detector at moonless night side of the Earth. M is the code of PMT voltage and N is the code of ADC. Two circulations are shown. Region above the ocean where M is at maximum and N indicates the darkest place in the circulation is indicated by . is the UV light peak above Mexico city, ­same for Houston, - same for Los-Angeles. In xaxis "n" is the number of the 4 s time intervals where UV intensity is measured. Fig.7. Data on the UV intensity obtained from the data presented in Fig.6. Fig.8. UV intensity at the night side orbit with a full moon. Fig.9. Aurora UV light intensity at the Southern latitude of about 70o. Fig.10. Examples of transient UV flashes registered by oscilloscope with the trace of 4 ms. Fig.11. Examples of transient UV flashes registered by oscilloscope with the trace of 64 ms. Fig.12. World map of registered UV flashes. Circles - data of the oscilloscope with 4 ms trace. Triangles - data of the oscilloscope with 64 ms trace. Fig.13. Block diagram of the TUS electronics.

8


Figure 1.

Figure 2.

9


Figure 3.

Figure 4.

10


Figure 5.

Figure 6.

11


Figure 7.

Figure 8.

12


Figure 9.

Figure 10

13


Figure 11.

Figure 12.

14


Figure 13.

15