Antarctic Sea Ice Trend by Pixel

Updated twice.

——-

The antarctic sea ice trend is important as another baseline for Steig et al as a method of checking whether conclusions about temperature are reasonable.

As before, each ice pixel represents the trend in ice area per year. Brown is land, black is shoreline, grey is ocean. White represents zero trend while the maximum red value is +10Km^2/year for each 25×25Km pixel. The trend is for the full length of the satellite data 1978 -2009, only pixels with 120 months of ice data were used for the trend plot. This was arbitrarily chosen because pixels with 5 years of data aren’t sufficient to show a low noise trend.

The data has more variation in it than I expected. From previous work I knew the South edge of the graph would show a strong red color to the plot. The tip of the peninsula on the left is deep blue as expected because this is one of the fastest warming areas on earth. As we expect though ice has grown slightly so there is slightly more red color than blue.

The next graph is temperature plot, be careful because higher temperature trend is in red which would in theory add more blue (less ice) to the first graph.

This next graph is from an area weighted temperature reconstruction I did

I can’t make anything out of this but I had to show it so that others could understand better what noisy data can do.

In exploring the reasons behind the splotchy look of the Arctic, it is important to look at individual pixels to understand what creates the trend. Below is a typical ice pixel as recorded by the NSIDC. For most of the year the pixel is water only. You can see why the noise level is high.

The next 3 plots are of pixels which are adjacent to each other.

Well there it is. I was quite surprised the trend wasn’t smoother from pixel to pixel. I notice that in the first plot, the ice is demonstrating a growing trend on the top and bottom edges of the plot quite strongly while at the peninsula only the very tip is showing melting. The same spot where seven manned surface stations are maintained.

The green smudge at the tip of the peninsula represents seven manned surface stations.

Update: I’ve added plots which represtent the locations of the four trend graphs above. The first trend graph above is located at the topmost dark green point in the gray ocean area in figure below.The pixel had less than the required number of months (120) for the trend plot in Fig 1. The second green dot represents the three adjacent pixels represented in the other 3 trend plots.

Update again:

All points have a minimum of 30 months of data

All points have a minimum of 20 months of data, forced a scale change +/- 15Km^2/Year

The shorter months of data give a better view of why the ice level is increasing in the Antarctic. You can see the maximum extent of the north south edges have grown strongly while the east west edges have reduced in maximum extent in the last 30 years. The net is positive ice growth.