Nine statistically significant decay channels are observed in LHC data around a mass of 650 GeV. We interpret three of them as coming from a narrow resonance observed in $e^+e^-$ , 2 photons and ZZ which could be interpreted as a J=2 Kaluza Klein graviton candidate called T690 (T for tensor with J=2). This hypothesis is reinforced by noting that this signal disappears in ZZ when treated as a scalar. Given that the six other excesses have poor mass resolution, one cannot exclude the presence of an additional wide scalar resonance called H650. Assuming a Randall Sundrum model, we conclude that LHC observes the predicted sequence T376, T690 and T1000, as indicated by the data. At variance with the RS model, T690 weakly couples to gluon pairs, suggesting a composite model interpretation. Perturbative unitarity requirements predict $T^{ ++} \to W^+W^+$ and $T^+\to ZW$ resonances, again indicated by LHC data. This scenario offers excellent prospects for abundantly (109 events) producing a sequence of narrow resonances at future $e^+e^-$ colliders. For heavy scalars, the situation is less clear. Following ATLAS and CMS, we expect that the top loop contribution to the gluon-gluon fusion mechanism ggF could produce a deficit rather than an excess in the mass distribution of top pairs, which prevents a standard estimate of the statistical significance for heavy resonances. It seems that the pseudo-scalar and scalar resonances A490 and H650, indicated by other channels, create observable deviations in the t t analyses presented by ATLAS and CMS. The present note summarises these arguments and collects available indications in view of electing a future collider.
Categories: hep-ph hep-ex