Title: Characterization of exoplanetary atmospheres through a model-unbiased spectral survey methodology

Abstract: Context. Collecting a large variety of exoplanetary atmosphere measurements is crucial to improve our understanding of exoplanets. In this context, it is likely that the field would benefit from broad species surveys, particularly using transit spectroscopy, which is the most successful technique of exoplanetary atmosphere characterization so far.
Aims: Our goal is to develop a model-unbiased technique using transit spectroscopy to analyze every qualified atomic spectral line in exoplanetary transit data, and search for relative absorption, that is, a decrease in the flux of the line when the planet is transiting.
Methods: We analyzed archive data from HDS at Subaru, HIRES at Keck, UVES at VLT, and HARPS at LaSilla to test our spectral survey methodology. It first filtered individual lines by relative noise levels. It also corrected for spectral offsets and telluric contamination. Our methodology performed an analysis along time and wavelength. The latter employed a bootstrap corroboration.
Results: We highlight the possible detections of Mn I and V II in HD 209459b data taken by HDS at Subaru ($5.9\sigma$ at 5916.4 $\angstrom$, $5.1\sigma$ at 6021.8 $\angstrom$). The previous detection of Ca I in the same planet is classified as inconclusive by our algorithm, but we support the previous detection of Sc II ($3.5\sigma$ at 6604.6 $\angstrom$). We also highlight the possible detection of Ca I, Sc II, and Ti II in HD 189733 data taken by UVES at VLT ($4.4\sigma$ at $6572.8 \angstrom$, $6.8\sigma$ at $6604.6 \angstrom$, and $3.5\sigma$ at $5910.1 \angstrom$), in addition to the possible detection of Al I in WASP-74b data taken by UVES at VLT ($5.6 \sigma$ at $6696.0 \angstrom$).