Hong-Ming Zhu and Prof. Xuelei Chen from Cosmological Dark Energy and Dark Matter Research Group at NAOC, Prof. Ue-Li Pen from Canadian Institute of Theoretical Astrophysics, and Dr. Yu Yu from Shanghai Astronomical Observatories proposed a new method to measure neutrino masses based on observations of cosmological large scale structure. This work has been published in the Physical Review Letters (Phys. Rev. Lett. 113,131301 (2014)).

In the Standard Model, there are three neutrino species. In 1998, neutrinos were found to oscillate, which indicated that their masses are non-zero. However, neutrino mass is difficult to measure, and their values remain undetermined, only the sum of the neutrino masses are constrained by the observations of the galaxy large scale distribution. This traditional method is based on the fact that cosmic relic neutrinos will smooth the small scale density fluctuation, which suppresses the small scale power spectrum. However, the galaxy formation is a complicated process, the bias factor, i.e. the ratio of galaxy number fluctuation to density fluctuation may not be a constant factor but varies over different scales, this may induce systematic error in the measurement and reduce its accuracy.

Zhu et al. find that neutrinos will develop a coherent relative bulk velocity field with respect to cold dark matter (CDM) under the gravitational interaction. This relative bulk velocity will induce a cross correlation dipole along the relative bulk velocity direction between galaxies with different biases. By measuring this cross correlation dipole in large volume galaxy surveys, we could determine the neutrino mass. This method is robust to the scale dependent galaxy bias. Further more, this method could determine single neutrino mass instead of the sum of neutrino masses. Combining with information provided by other experiments, the absolute neutrino mass can be measured.