To my semi-trained eye, it looks like if you stare at the data long enough, you can convince yourself there's an upward trend in temps, you stare at it a little longer, you can convince yourself that there's some longish timescale periodicity too, all within the error bars. Looks like a bunch of noise to me, and I look for weak signals in noise for a living, so I do know what I'm talking about to a certain extent. That said, I do remember people predicting about 10 years ago that melted ice caps can act as a negative feedback mechanism for high latitude temperatures, and thus, global warming->longer/colder winters at high latitudes.