A new approach to generate a catalogue of potential historical novae

Ancient Chinese, Korean and Vietnamese observers left us records of celestial sightings, the so-called `guest stars' dated up to $\sim2500$ years ago. Their identification with modern observable targets could open interesting insights into the long-term behavior of astronomical objects, as shown by the successful identification of 8 galactic supernovae (SNe). Here we evaluate the possibility to identify ancient classical novae with presently known cataclysmic variables (CVs). For this purpose, we have developed a method which reconsiders in detail positions and sizes of ancient asterisms, in order to define areas on the sky that should be used for a search of modern counterparts. These areas range from a few to several hundred square degrees, depending on the details given in ancient texts; they should replace the single coordinate values given by previous authors. Any appropriate target (CVs, X-ray binaries etc.) within these areas can be considered as a valid candidate for identification with the corresponding ancient event. Based on the original descriptions of several hundred old events, we selected those without movement and without a tail (to exclude comets) and which did not only visible within a certain hour (to exclude meteors). This way, we present a shortlist of 24 most promising events which could refer to classical nova eruptions. Our method is checked by applying it to the known SN identifications, leading to a margin of error between 0 and 4.5 degrees, meaning that some SN remnants lay exactly inside the areas given by the historical reports while in some other cases they are laying at considerable distances.