The relatively large measured value of θ 13 has opened various possibilities to determine the neutrino mass ordering, among them using PINGU, the low-energy extension of the IceCube neutrino telescope, to observe matter effects in atmospheric neutrinos, or a high statistics measurement of the neutrino energy spectrum at a reactor neutrino experiment with a baseline of around 60 km, such as the Daya Bay II project. In this work we point out a synergy between these two approaches based on the fact that when data are analysed with the wrong neutrino mass ordering the best fit occurs at different values of varDelta m 312 for PINGU and Daya Bay II. Hence, the wrong mass ordering can be excluded by a mismatch of the values inferred for varDelta m 312, thanks to the excellent accuracy for varDelta m 312 of both experiments. We perform numerical studies of PINGU and Daya Bay II sensitivities and show that the synergy effect may lead to a high significance determination of the mass ordering even in situations where the 3individual experiments obtain only poor sensitivity.