Dipion invariant mass distribution of the anomalous Upsilon(1S) π⁺ π⁻ and Upsilon(2S) π⁺ π⁻ production near the peak of Upsilon(10860)

Considering the defects of the previous work for estimating the anomalous production rates of $e^+e^-\to \Upsilon(1S)\pi^+\pi^-,\,\Upsilon(2S)\pi^+\pi^-$ near the peak of the $\Upsilon(5S)$ resonance at $\sqrt s=10.87$ GeV [K.F. Chen {\it et al}. (Belle Collaboration), Phys. Rev. Lett. {\bf 100}, 112001 (2008)], we suggest a new scenario where the contributions from the direct dipion transition and the final state interactions interfere to result in not only the anomalously large production rates, but also the lineshapes of the differential widths consistent with the experimental measurement when assuming the reactions are due to the dipion emission of $\Upsilon(5S)$. At the end, we raise a new puzzle that the predicted differential width $d\Gamma(\Upsilon(5S)\to\Upsilon(2S)\pi^+\pi^-)/d\cos\theta$ has a discrepant trend from the data while other predictions are well in accord with the data. It should be further clarified by more accurate measurements carried by future experiments.