Something forecasters use in the fall months to help predict the upcoming winter is indeed Siberia. In October, the anomaly of snow cover over the region has been proven to correlate with observed Arctic Oscillation anomalies in the following winter. For example, if the month of October held a very below normal snow cover situation for Siberia, the Arctic Oscillation would most likely be on the positive side (warmer weather) for the winter.
If we apply that rule to the observed October 2012 snow cover, here is what I find:
-Generally warm start to winter.
-Slight improvement through Nov., Dec.
-Cooler towards January.
-Brutal cold in January-February.
Whether this happens remains to be seen, but Siberia is known to stick to its guns with this theory, so the above conclusions could very well verify.