Comment on Japanese Detection of Air Fluorescence Light from a Cosmic Ray Shower in 1969

1 Comment on Japanese Detection of Air Fluorescence Light from a Cosmic Ray Shower in 1969 Bruce R. Dawson School of Chemistry & Physics, University of Adelaide, Adelaide 5005 Australia Abstract We examine the claim made by Hara et al.[1] in 1969 of the observation of a 1019eV cosmic ray extensive air shower using the air fluorescence technique. We find that it is likely that fluorescence light was observed, confirming this as the first such observation. The work of Hara et al. and their friendly competitors at Cornell University paved the way for modern experiments like the Pierre Auger Observatory and the Telescope Array. 1 Introduction Investigations into the feasibility of detecting air fluorescence light from extensive air showers were conducted in the 1960’s by groups led by Suga in Japan and Greisen in the United States. Results from the Japanese experiment, reported by Hara et al.[1] in 1969, are reviewed here. In that report, the authors say “One event is very likely due to the atmospheric scintillation [fluorescence] light from an air shower whose primary energy and distance are about 1019eV and 3 km, respectively”. This was the first reported observation of fluorescence light from an air shower. The purpose of the present short note is to the review this observation in the light of our modern understanding of fluorescence detection. The Japanese experiment ran at the Dodaira Observatory (altitude 876 m) for a period of 5 months from December 1968. The fluorescence telescope consisted of a 1.6 m diameter Fresnel lens focussing light onto a camera of 24 PMTs, each of which imaged a 4.5◦degree portion of the sky. For the observation described here, the telescope with its field of view of 23◦×20◦was centered at an elevation of 30◦. The design was similar to Greisen’s Cornell telescope [2]. However the Japanese design had the advantage of a larger Fresnel lens, and faster electronics. The rise and fall-times of pulses on the cathode-ray tube displays were 0.12µs and 0.2µs respectively. The potential fluorescence observation (event #12 in Fig 3 of [1]) triggered 8 PMTs with an angular track length of 18.4◦and a duration of 1.9µs. In the next section we review the event geometry before considering the shape of the light profile received at the telescope. 2 Event Geometry I have taken the PMT trigger times from Fig 3 of [1], and using an estimate of the PMT pointing directions, I have made fits to the standard timing equation, ti = t0 + Rp c tan χ0 −χi 2  to extract shower axis parameters t0, Rp and χ0 from the eight (χi, ti) data points. I have assumed a vertical shower-detector plane (SDP), and I have guessed at a timing uncertainty of 0.05µs for each point. The SDP and the axis parameters are illustrated in Figure 1. Results of various timing fits are shown in Figure 2. Because of the rather short angular track length of the event (18.4◦), the timing fit suffers a large degeneracy in the parameters Rp and χ0; there is no curvature evident in the ti versus χi plot, meaning that while we do have an estimate of the the angular speed ω of the light spot across the camera, we have no information about dω/dt. The best fit in Figure 2 returns χ0 = 38◦and Rp = 3.6 km, but we show that a wide range of values of χ0 give acceptable fits. Other values not shown (e.g. χ0 > 90◦) also give reasonable fits. arXiv:1112.5686v1 [physics.hist-ph] 24 Dec 2011 2 Figure 1: The shower axis and the telescope define the shower-detector plane (SDP). The timing fit returns the axis parameters χ0 and Rp, and the time t0 at which the shower passes the point of closest approach. (Image from D. Kuempel). The implication of this degeneracy is that the axis geometry of this event is very uncertain - the shower could be vertical (χ0 = 90◦assuming a vertical SDP) with an impact parameter of Rp = 2.7 km, or the shower could be approaching the detector with a zenith angle of 60◦(ie χ0 = 30◦) and Rp = 3.6 km. The latter geometry of an inclined, approaching shower would produce a light signal dominated by Cherenkov light. If this were the case, the claim of Hara et al. of the observation of fluorescence light would be in question. However, Hara et al. correctly point out that there is important information in the shape of the light profile recorded by the telescope. We test this in the next section. 3 Light Profile Figure 3 and Figure 4 of [1] give information on the flux of light received by the telescope as a function of time (or angle χi). The key point is that the flux profile is rather flat. I have performed some simulations of a shower with a range of axis geometries consistent with the timing fits from the previous section. The simulated shower had a fixed energy of 5 × 1018eV and a depth of maximum Xmax = 680 g/cm2. The aim of the exercise is not to fit the observed light profile, but to illustrate the change in the light profile shape as a function of the shower geometry. Figure 3 shows the results of these simulations. We find that flat light profiles are only see

This content is AI-processed based on open access ArXiv data.