Proceedings of the IMC, Giron, 2014 1 RETRAM: recognition and trajectories of meteors Jean-Jacques Maintoux, Sylvain Azarian, Jérémy Maintoux, Frédéric Rible ARRL / RETRAM, Velizy-Villacoublay, France contact@retram.org Meteor detection and tracking is the main activity of the RETRAM group, part of the French ARRL organization. Our project uses passive radar techniques and real-time processing to detect and recognize falling objects and tries to estimate their trajectory to help in fireball recovery. The experiment started in the vicinity of Paris, France. This paper shows our observations and analyses, then it describes our technical approach, our first passive radar station and first meteor 3D localization. Finally, we describe the evolution of the system and subsequently its extension in the form of a network of stations grouping radio and optical detectors. 1 Introduction RETRAM (as REcognition and TRAjectories of Meteors) is a group of amateurs working together to make experiments around meteors detection and their trajectories, using radio signals through passive radar techniques. Goals for the RETRAM project – our wishes – are:  to experiment with new methods for detecting and recognizing meteors using radio signals;  to design and to test new algorithms and processing to obtain automatic detections;  to design and to test new techniques to reconstruct a meteor path and to help in fireball recovery;  to use optical detection if possible to enhance results;  to cooperate with the scientific community to share the outcome of the project. RETRAM is built around the use of radio signals because they allow a constant survey in all weather conditions (therefore not limited by the visibility conditions) and at any time (not limited by lighting conditions). 2 Observations and Principle of RETRAM To avoid electromagnetic hazard and electromagnetic compatibility risks, RETRAM started with classic passive radar techniques using transmitters of opportunity. Starting in 2012, a survey was dedicated to the observation of meteors using the well-known Graves transmitter and some aeronautical VHF VOR beacons. (Some measurement reports are available on our WEB site (RETRAM - Recognition & trajectory, 2014) 1 . These transmitters allowed to confirm:  the minimum “radar” budget necessary to observe the meteors; 1 http://www.retram.org/wpcontent/uploads/2014/06/3D_FM_Res ults.pdf  the various signals we have to measure (signals level, speed, spectrums…). We were able to confirm previous observations done by amateurs and other works like Close et al. (2002), Close et al. (2011) and to define the criteria we have to measure to develop our project. Optical and radio observations Optical observations performed simultaneously by radio measurements and Doppler analysis both revealed the optical trail is correlated (in time) with the radio head echo of the meteor. This head echo is characterized by a huge Doppler slope (or fast Doppler shifting). The head echo was observed using different transmitters and the Figure 1 shows such a correlation. The phenomenon of head echo and its radio detection were also detected using VOR beacon as shown in Figure 2. Figure 3 shows a head echo and non-specular train signals collected at VHF frequencies (Close et al., 2011). The head echo, characterized by a fast Doppler shift (or high penetrating speed in atmosphere) is red circled. The head echo is followed by a long train (depending on the object and environment) presenting a very low Doppler shift (or very low speed) depending also of the environment in high atmosphere. Principle of RETRAM Previous observations and measurements were performed with signals (waveforms in radar wording) having no or poor temporal information (named narrow band signals). These waveforms allow integration of the signal over a relatively long time and made Doppler slope measurement possible with good accuracy. But it was impossible to find the location of meteors at any time. To enhance the meteor path reconstruction, the RETRAM project is based on a larger waveform bandwidth, with a good Range/Doppler ambiguity, permitting to measure temporal information with an enhanced accuracy (better than 1km). The Doppler accuracy is still limited by the integration time. As shown by numerous papers and measurements, the VHF band offers the best results for the