Six weeks of climate data ends early: Southern Polar Region #4, 90° to 120° East

This fourth slice of Antarctic Circle does a good job of exposing some flaws in my method of collecting and averaging data. The region only has 319 seasonal readings from 14 stations in 56 years and this is not even the worst covered slice of the Antarctic. (Notice that this slice is the only one whose coastline is almost entirely outside the Antarctic Circle.)

This coverage pattern does not show exactly where the 14 stations are, but instead superimposes a 10 × 10 grid over the region and measures which stations are close to the grid marks. Some grid places have nothing close and do not register.

The problem with this particular region makes me wonder if the pattern we saw in Region 3 of a massive drop in temperatures has anything to do with new stations coming online in colder regions. In this region, there were only 3 Summers out of 21 between 1955 and 1975 that had any readings and in the 1975-1988 time interval there were only two.

I could instead just look at the last two intervals, but this problem makes me wonder if other data we have seen has been affected by stations coming online and going offline in significantly different parts of these large slices.

I'm going to work on my program and start the whole thing over again in the Arctic on Friday. The new method will set a threshold for how many seasonal readings a station has in the requested time interval and only count the data from the most reliable stations. I will spend the next few days looking at different thresholds - right now thinking the magic number should be somewhere between 75% and 90% - and on Friday I'll show the map of the slices of the Arctic again. Wednesday and Thursday I'll share some results and thoughts about the new method.

Not unlike my method for collecting polling data, I fiddled with the system until I felt it was close to right. It would be a dream come true if this system gets results as good as that one. I'll be happy with data I can explain and with luck, convince others to use some of my methods when they talk about climate.