Friday, January 20, 2017

Weekend Rebel Science Excursion - 58

I became tired of the "standard deviation" (std. dev.) graph technique, so I tried to create a new technique to replace it.

I rebelled, and mind you, this took place while I was proving that the Earth is not flat (Once Upon A Time In The West - 2).

Fig. 2

If you will be so kind as to check out Fig. 1 and Fig. 2 (cursorily or perfunctorily, your choice) you can see that summer and winter no longer happen at the same time on the poles (once called edges) of the Earth.

The peak of the quantity of sea ice in the Arctic takes place at the time the lowest quantity of sea ice takes place in Antarctica.

Fig. 3
When the Earth was still flat, summer and winter levels of ice would be at their maximum and minimum at the same time.

Oh, for the simple life again, or for the return of the great Flatulence.

Fig. 4
Anyway, let's get back to the cool and informative new graph technique I want to share with you.

The "std. dev." technique is used and shown in Fig. 7 below, I added two comments "not very telling" to that graph with pointers to the edges of the typical standard deviation.

Fig. 5
The new technique is a most-and-least boundary on the high side and the low side of the main stream of data being graphed (in this case the million km2 area of sea ice).

The new Dredd Blog technique is to construct the high side line out of the high year value (highest million km2 for that day of the year, 1-366, out of all years in the database table).


The same for the other side of the data flow, except it is the lowest million km2 for that day of the year.

As you can see, since the values merge when the current year is the lowest or highest year of sea ice extent, trends and essential values are at once obvious when the lines approach, go away from, or merge with one another (all graphs - Fig. 1 - Fig. 6).

Fig. 7
Doing that is not as easy as it looks, or as hard as it looks either.

The thing about it is that as you go through the central values in the database (the ice extent for a particular year, month, and day) you can preserve the lowest and highest values for that day of the year quite easily.

I use variables named "lowExtent" and "highExtent" to record the low and high for that day of the year, which requires nothing more than a comparison as to which value is higher or lower as the case may be.

Both numerically and visually, just those numbers alone, in the CSV file, and on the graph, tell a story to the alert observer (see the red numbers in the 2017 data below).

Then, when you are writing the graph data to your storage medium, you can easily add the highest and lowest values, as well as the year, along with all the other data:
The red-numbered years are the high extent beginning and ending years for that day of the year (1-15).

They tell a story about sea ice extent highest levels being set decades ago, but they also tell a story about the global warming caused low sea ice extent records being set before our very eyes.

That is the "art" and science of rebelling and telling a big story with the usage of the fewest numbers in a dataset.

The previous post in this series is here.

No comments:

Post a Comment