Well, the sum of the values in the second and fourth column equals the value in the third column, but with an irregular discrepancy of 0.00000020 which I would attribute to somehow taking the first column as values linear in time, which is an error since these values describe a DATE which increments by one day in each row, whereas the raw value increments by 0.0027 or 0.0028.
As to square roots, I don’t suppose it bothers you that you are turning up your nose at Gauss’s construction of the heptadecagon by finding a solution in radicals ( i.e. square roots, recursively applied ) to the 16th order cyclotomic polynomial.
“Many were the Shugs and Zuls who knew what it was to roast in the belly of the Slor THAT day, I can tell you.”
Well, the sum of the values in the second and fourth column equals the value in the third column, but with an irregular discrepancy of 0.00000020 which I would attribute to somehow taking the first column as values linear in time, which is an error since these values describe a DATE which increments by one day in each row, whereas the raw value increments by 0.0027 or 0.0028.
You’re good! That was the very subtle one.
But there are two otehrs.
First, the “oh by the way” one. This is the data for the Arctic sea ice for the early days in May, 2016. (The date record you properly saw. But 2016 is a leap year, and I see there are slight differences between the comparable “days” in 2013, 2014, and 2015 when the “decimal year” was 365 days, not 366. (That is because they are “rounding off the decimal date (2016.1234) to only 4 places per day (0.0027 difference between days. Yet, .0027 x 366 is still only .9882 of a leap year.
So, you see, the “math” is right, but the information is wrong. Not wrong by much, but still not “right”.
( Still need to compare the earlier years exactly to see if they are using 365.24 (which is also slightly off since we need a 400 year leap date as well) as the “days per year” for a non-leap year. )
Then that brings up the question of “how do you compare measurements from year-to-year across leap year differences: February 10 to February 10 is always the same “day-of-year” but their decimal dates are different. Yet, do you compare May 30 to June 1? Or to May 30 to May 30 to May 30 every time? Also. But data is often compared between using monthly values. If the information “jumps” back and forth by one “day” every four years, what does that do a month-to-month average?
The second error - and much more important is not in the math (well, the arithmetic), but in the data. (Which is the important thing I wanted to show these numbers for.)
The satellite sensor failed on May 5, and the “daily sea ice area” column 3 got “stuck” at 4.6884198.
Now, this is in millions of square kilometers - and the idea that any “information is correct after the 4.68 digits is rather foolish in itself, but we will pretend the satellite really is accurate to the 1/10 sq kilometer through the ice and clouds and atmospheric diffusion and difraction of the radar signals over distances involved.
The fourth column (historical average sea ice for that date) continues to change each day. The error - we hope! - will NOT get input into future historical averages.
The second column, the daily sea ice anomaly is the simple difference between the daily historical average and the reported daily measurement.
Column 2 is dead wrong, even if the mathematics is exactly right!
And don’t even get me into the mathematical “errrors” of black hole manufacturing!