Averaging by month (which applies the current mean for December), we have a 0.5F cushion over 2016, 2012, 1991 & 1949. That means the final monthly mean could fall off a full 6F [6/12 = 0.5] and we'd still tie those years using monthly averaging. I'm kind of doubting December's mean falls off that much over the next three weeks given the current modeling. Note using daily averaging, 2012, 1991 & 1949 bump up one place with a mean of 54.3F.