Sorting a mess I: Addressing velocity-axis correlation of cross-correlation functions. A census of atomic and ionised species in KELT-9b's atmosphere

Ultra-hot Jupiters (UHJs) are the most extreme class of close-in gas giants. We combine 13 transit observations of KELT-9b from six high-resolution spectrographs spanning 316-960 nm to conduct a census of atomic and ionised species in its atmosphere using cross-correlation. We introduce an agnostic strategy to mitigate velocity-axis correlation bias in parameter uncertainties. A forward model of the cross-correlation function (CCF) and a Monte Carlo bootstrap of flux uncertainties measure the CCF-derived velocity profiles of individual species and assess their time dependence. The stellar reference frame is determined, yielding Vsys = -17.99 +/- 0.04 km s-1 and v sin i = 110.54 +/- 0.05 km s-1. All previously reported species are recovered except Tb II; the combined analysis detects 29 species, including C I, Si I, K I, V II, Co I, Zr II, La II, Pr II, and Nd II. All 29 recovered species exceed the conventional 5 sigma threshold in the bootstrap Mahalanobis significance test. The collective Doppler shifts show a blueshift of Delta Voffset = -8.02 +/- 0.17 km s-1 relative to the stellar rest frame, indicating a net day-to-night wind, with no explicit time dependence when all species are homogenised. These results provide a catalogue of detections and retrieved Doppler parameters for a companion study of KELT-9b.