In addition, the pattern of sea surface temperatures at low latitudes is extremely important for regional climate variations (shown, for example, by the increased likelihood of heavy winter rainfall in California when the eastern tropical Pacific warms in El Niño events). (giss.nasa.gov)
Hence, studies of observed changes in regions influenced by an oscillation may be able to attribute these changes to regional climate variations, but decades of data may be needed in order to separate the response to climate oscillations from that due to longer - term climate change. (ipcc.ch)
BTW, my view that it is the regional climate variation that is not associated with the global change is very consistent with Vecchi and Soden's work showing that MPI is approximated by subtracting tropical mean SST from MDR SST: (climateaudit.org)