Plot of the extent of Arctic and Antarctic ice since the late seventies  I think the total is decreasing, while the Antarctic ice is increasing.

This uses data (I believe originally from NSIDC) kindly formatted and provided by +Peter Bromberg.

Because of the seasonal fluctuations, things can be a little tricky: when you begin or end your measures will alter the local average at the beginning and end of the plot.  To minimize this, I end plots a few points early on all but the top graph.  This truncation should be conservative in that it biases us against the general trends. In other words, the apparent changes we see on these plots would be larger if we included the last few points of the dataset.  Also, I think this measurement window effect is less of an issue when we plot the total ice (sum of Arctic and Antarctic) because their seasonal trends are opposite each other and partly cancel.  

While the top and bottom plots show lines that indicate the overall slope of the data, the middle plot uses a locally weighted regression which does not assume that the slope is constant. For reference on all plots, I include a black horizontal line that shows the average of all points on a plot.  

I would need to know more about this dataset than I do to say whether the slopes are significantly different from zero.  But certainly the apparent trend is for the total Arctic and Antarctic ice to be decreasing while the Antarctic ice may be slightly increasing. 

My reformatted data file is here
My R code for plotting data is here

(h/t +Richard L. Brandt )
Shared publiclyView activity