Sprite observations over France in relation to their parent thunderstorm system Lars Knutsson OBERON : Through the house give gathering light, By the dead and drowsy fire : Every elf and fairy sprite Hop as light as bird from brier ; And this ditty, after me, Sing, and dance it trippingly. (W illiam Shakespeare, A midsummer night‘s dream) Abstract As a part of the European research program CAL, sprite observations were carried out from the OMP observatory in the French Pyrenees during the summer 2003. Images of the sprites were taken by two remotely controlled CCD cameras. The 23 July was considered particularly interesting because we then had access to data concerning both cloud-to-ground and intracloud lightning activity. This day was therefore chosen as the object of the present study. A large thunderstorm with two convective cores, one to the north and the other to the south, developed over the South of France during the late afternoon, and about two hours after sunset, the first sprite was detected. During a little more than three hours, 13 sprites were observed, 7 over the northern system and 6 over the southern system. The images enabled us to determine the azimuth angle of each sprite from the OMP observatory. 12 of the 13 sprites could be associated to positive cloud-to-ground flashes, and by putting together the sprite directions and the locations of the associated flashes on the radar images, we managed to get a rough idea of the position of the sprites in the storm system, and also to estimate their vertical and horizontal extent. Satellite images were included at this point of the study, and it appeared clear that sprites tend to occur over the stratiform region of the storm system in the area with the coldest (highest) cloud tops. The associated positive flashes were also within or close to this portion of the storm. The sprite occurrences were studied in relation to the cloud-to-ground and to the intracloud activity. W e found that sprites seem to occur in a late stage of each storm system, when the rate of negative cloud-to-ground flashes has considerably decreased, and when the ratio of positive cloud-to-ground flashes is much higher then during the most active phase of the storm. Globally, the intracloud activity is also low during the sprite-producing periods, but sudden —bursts“ of intracloud lightning could frequently be observed at the moment of the sprite. The peak current of the positive flashes was found to be rather weakly correlated to their sprite-generating capacity. The available Schumann resonance measurements seem to indicate that the charge moment is a much more adequate parameter in this respect. The areal coverage of the radar echo was calculated. The result supports the idea that sprite events tend to appear almost exclusively over large thunderstorm systems. Contents 1 Introduction… … … … … … … … … < < < < < < < < < < < < < < < < , 1 ) Background and theory< < < < < < < < < < < < < < < < < < < < ,, 2 ),3/ history of sprites… … … … … … … … … … … … … … … … … … … … ... 2 2.2 A worldwide phenomenon… … … … … … … … … … … … … … … … … … .. 3 2.3 The TLE family… … … … … … … … … … … … … … … … … … … … … … .. 5 2.3.1 Sprites… … … … … … … … … … … … … … … … … … … … … … … 5 2.3.2 Blue jets… … … … … … … … … … … … … … … … … … … … … … . 7 2.3.3 Blue starters… … … … … … … … … … … … … … … … … … … … … 8 2.3.4 Elves… … … … … … … … … … … … … … … … … … … … … … … ... 9 2.3.5 Sprite haloes… … … … … … … … … … … … … … … … … … … … ... 11 2.3.6 Trolls… … … … … … … … … … … … … … … … … … … … … .......... 11 2.3.7 Gnomes… … … … … … … … … … … … … … … … … … … … … … .. 11 2.3.8 Pixies… … … … … … … … … … … … … … … … … … … … … … … .. 11 2.3.9 And what else ?… … … … … … … … … … … … … … … … … … … .. 11 2.4 Theory… … … … … … … … … … … … … … … … … … … … … … … … … … . 12 2.4.1 Lightning… … … … … … … … … … … … … … … … … … … … … … 12 2.4.2 The global atmospheric electric circuit… … … … … … … … … … .. 13 2.4.3 W hat causes TLEs ?… … … … … … … … … … … … … … … … … ... 15 3 The CAL project… … … … … … … … … … … … … … … … … … … … … … … .. 21 3.1 W P5… … … … … … … … … … … … … … … … … … … … … … … … … … … . 21 4 Instrumentation and available data… … … … … … … … … … … … … … … ... 22 4.1 Images from the Pic du Midi… … … … … … … … … … … … … … … … … ... 22 4.2 ARAMIS… … … … … … … … … … … … … … … … … … … … … … … … … . 23 4.3 Météorage… … … … … … … … … … … … … … … … … … … … … … … … … 24 4.4 SAFIR… … … … … … … … … … … … … … … … … … … … … … … … … … .. 25 4.5 Meteosat… … … … … … … … … … … … … … … … … … … … … … … … … .. 27 4.6 Schumann resonance measurements… … … … … … … … … … … … … … ... 27 5 Case study… … … … … … … … … … … … … … … … … … … … … … … … … … 29 5.1 The meteorological situation… … … … … … … … … … … … … … … … … .. 29 5.2 Observed sprites… … … … … … … … … … … … … … … … … … … … … … . 31 5.3 The sprites in relation to the cloud-to-ground lightning… … … … … … … 34 5.3.1 The northern storm system… … … … … … … … … … … … … … … . 38 5.3.2 The southern storm system… … … … … … … … … … … … … … … . 42 5.4 Intracloud activity… … … … … … … … … … … … … … … … … … … … … ... 46 6 Conclusion… … … … … … … … … … … … … … … … … … … … … … … … … … 53 Acknowledgements… … … … … … … … … … … … … … … … … … … … … … … … . 55 References… … … … … … … … … … … … … … … … … … … … … … … … … … … ... 56 Appendix 1, Estimated vertical and horizontal extent of the sprites… … … … … .. 61 Appendix 2, Charge moments for some of the associated +CGs… … … … … … … 62 Appendix 3, Soundings from the airport of Lyon… … … … … … … … … … … … ... 63 Appendix 4, Intracloud activity 1 second before and 1 second after each sprite… 64 1 Introduction Sprites are transient luminous glows at altitudes of approximately 40-90 km above large thunderstorms. They are the most dramatic visible evidence of electrodynamic coupling between thunderstorm systems and the overlying mesosphere and lower ionosphere. Although they are much more rare than —ordinary“ lightning, they may have long-term effects on the atmosphere such as chemical changes, persistent heating of ionospheric electrons, and increased production of mesospheric and stratospheric nitrogen oxides (Reising, 1998). All these effects can influence global climate. The understanding of the physical processes behind sprites, the meteorological conditions that generate them, and their global occurrence rate, are therefore of interest not only for a small group of scientists, but for humanity as a whole. An important issue is to find out whether sprites significantly influence the global electrical circuit. In the conventional picture, the main components of Earth‘s electrical circuit include thunderstorms, the conducting ionosphere, the downward fair-weather currents and the conducting Earth. The thunderstorms are the generator that drives currents upward from cloud tops and maintains an electric potential of hundreds of kilovolts between the ionosphere and the ground. It is possible that sprites play an important role in this circuit. This is a topic of future research. For easily understandable reasons, sprites are very difficult to observe from the ground. Occurring at high altitude over the massive cloud cover of large thunderstorms, they can be seen only under certain conditions from high mountains or from aircrafts. This is of course why they have remained unknown to science for so long. Scientific study of sprites began only a little more than ten years ago. Since then, there has been an explosion of research in optical and radio measurements and theoretical modeling, not only of sprites, but of all the parent mesospheric phenomena that have been discovered recently : blue jets, elves, trolls… If sprite exploration is a very young research area, sprite exploration in Europe is even younger. The first sprites over the European continent were observed in year 2000 (Neubert et al., 2001), and the following summer campaigns have shown that these light emissions are a regularly occurring phenomenon in this part of the world. The summer 2003 produced several large thunderstorms over France and was therefore a —good“ summer for the European sprite research. A great number of observations were made and the results are waiting to be analyzed. In this study, we concentrate on the 23 July 2003. This was one of the few days for which we had access to data concerning the intracloud activity. 1 2 Background and theory 2.1 The history of sprites The 6 July 1989 is known to be the day when the phenomenon now called sprites was discovered. A group of scientists from the University of Minnesota, lead by Professor John R. W inckler, were then testing a low-light-level TV camera intended for a sounding rocket flight. W hile directing their camera to the northern horizon, they recorded a twin flash that seemed to propagate from some distant cloud tops up in the stratosphere. The flash lasted for less than 30 milliseconds and was associated with a very active storm center located on the northwestern side of Lake Superior. Its vertical extent was about 20 km and thus much greater than that of cloud-to-ground flashes. The publishing of these results (Franz et al., 1990) gave birth to an intense research in the electrodynamics of the middle atmosphere. But the phenomenon did not suddenly appear on that particular day. For some time already, rumors had been telling about —upward flashes“ and airplane pilots claimed to have seen —strange things“. In fact, for more than a century there had been accounts from credible witnesses, but the science community had generally