File talk:Global Temperature Anomaly.svg

From Wikimedia Commons, the free media repository
Jump to navigation Jump to search

Production Process

[edit]

How do you generate the SVG? Is it a machine translation of the PDF or ps file from http://data.giss.nasa.gov/gistemp/graphs/ or is it re-created from the source data from http://data.giss.nasa.gov/gistemp/graphs_v3/Fig.A2.txt

I had re-plotted the graph of the data using R and ggplot() on https://en.wikipedia.org/wiki/Talk:Global_warming_hiatus#Update_the_graphs_for_2015.3F but it would be interesting to know how to edit the existing graphs into SVGs, if that's what you are doing. Drf5n (talk) 18:52, 5 June 2015 (UTC)[reply]

Answering my own question: the SVG file is not a machine translation of the graphics on the site. It is an artfully coded SVG graphics file. Take a look at the code within the svg file (with 'curl https://upload.wikimedia.org/wikipedia/commons/f/f8/Global_Temperature_Anomaly.svg' or something similar. Drf5n (talk) 05:10, 6 June 2015 (UTC)[reply]

The formulas for the calculations are given both as comments in the source code, and in the comment for the upload 17 October 2013. They are: x-axis: pixels = 6*year − 11200; y-axis: pixels = −500*temp + 400.

However, I have a whole spreadsheet that I use to create svg text out of csv data copied from the source website. I wish there was some easy way to upload the spreadsheet here, but as there isn't, I'll try to give the spreadsheet cell formulas here, in a similar way to what I did above. --Nigelj (talk) 21:58, 23 January 2015 (UTC)[reply]
A B C D E F G H I J
1 Year Annual_Mean 5-year_Mean Year to x px Annual_mean to y px line text for annual_mean line graph year px-4 data points for annual_mean line graph 5-year_mean to y px line text for 5-year_mean line graph
2 1880 -0.21 * 80 505 M 80 505 80 <path class='annual_points' d='M 80 505 h4'/>
3 1881 -0.13 * 86 465 L 86 465 82 <path class='annual_points' d='M 82 465 h8'/>
4 1882 -0.16 -0.19 92 480 L 92 480 88 <path class='annual_points' d='M 88 480 h8'/> 495 M 92 495
5 1883 -0.19 -0.2 98 495 L 98 495 94 <path class='annual_points' d='M 94 495 h8'/> 500 L 98 500
  • The formula in the first cell of column D is =6*A2-11200
  • The formula in the first cell of column E is =-500*B2+400
  • The formula in the first cell of column F is =CONCATENATE(" M ", D2," ",E2)
  • The formulae in the first two cells of column G are
    • =D2
    • =D3-4
  • The formula in the first cell of column H is =CONCATENATE(" <path class='annual_points' d='M ",G2," ",E2," h4'/>")
  • The formula in the third cell of column I is =-500*C4+400
  • The formula in the third cell of column J is =CONCATENATE(" M ",D4," ",I4)

I hope this is helpful. --Nigelj (talk) 22:17, 23 January 2015 (UTC)[reply]

@Nigelj: The production is a bit different now due to the new sizing, but similar with comments in file and the whole thing is easy to reverse engineer. I just updated the data for 2018 except for the blue bars that show the "95% confidence interval.". Question- where is the data for the blue bars coming from? Maybe for simplicity they should just be cut? — Preceding unsigned comment added by Efbrazil (talk • contribs) 00:55, 8 February 2019 (UTC)[reply]

Date on Image

[edit]

Do we need to change the data on the image? The date of the image is 2013 which corresponds to when it was created. But that misleadingly suggests that data is from 2013 containing only earlier years. You have to go digging to see when the data is actually updated. 108.31.157.52 20:43, 27 May 2016 (UTC)[reply]

I'm sorry,I don't see where you are seeing the 'date of the image' to be 2013. Could you explain where you see that? --Nigelj (talk) 23:05, 26 September 2016 (UTC)[reply]

A better graph?

[edit]

There are a few improvements in this published graph that we may look at incorporating next update, if we can find reliable tabulated data. http://csas.ei.columbia.edu/2016/09/26/a-better-graph/ --Nigelj (talk) 23:00, 26 September 2016 (UTC)[reply]

Error in y-axis labelling

[edit]

There is an obvious error in the labelling of the y-axis - 0.9 at the top of the graph should read 1.0. Mikenorton (talk) 13:02, 27 January 2017 (UTC)[reply]

Oops, sorry. You're absolutely right. I've fixed that. --Nigelj (talk) 16:38, 27 January 2017 (UTC)[reply]

Graph is now modernized in several ways

[edit]

I bumped fonts, changed the aspect ratio, added a background to the key and graph, and improved compatibility with SVG importing for Microsoft Office. Primary goal was to make it look better in most versions of Wikipedia, especially mobile. Data is the same and SVG source should look similar, but several lines of code have changed and the view box / offsets have changed, so if this graph is generated from scratch then that generation will need to be modified. If you generate the data in Excel then you simply need to change the multiplier / offsets for the new viewbox. Let me know if there are questions or issues (or complaints).--Efbrazil (talk) 23:48, 31 January 2019 (UTC)[reply]

--2600:6C48:7006:200:B056:6066:1296:EF0B 01:50, 5 February 2019 (UTC)== How is 100 year old data verifiable by any independent means? ==[reply]

Temperature (and time-of-day) measurement methods 100 years ago were different then than they are today. There remains no consistent global coverage, or validation of temperature measurement accuracy and precision and calibration of measurements existent today. I cannot even independently verify temperature in my own home even if I read three digital thermometers in the same room yesterday within one minute. Why should anyone trust this? It is not a matter of belief or consensus- it is a matter of Truth. Always question. --2600:6C48:7006:200:B056:6066:1296:EF0B 01:50, 5 February 2019 (UTC)[reply]

No source provided for updated description

[edit]

Noted that someone added these lines to the description:

 Beware that this graph gives a misleading and inaccurate representation because a quick check of temperature records in the USA in the 30's shows that temperatures were warmer than present temperatures. This is also misleading because it purposefully only includes recent climate data and checking back any period of years shows there are regular climate change cycles.

-- 2600:8805:3800:D1:2A0:D1FF:FEAE:B38B 20:22, 17 April 2019 (UTC)[reply]

Suggestions for improvement

[edit]

I don't have SVG capability, so I suggest the following improvements:

(1) keep the dots, but make the spikey jagged annual trace (line) even lighter or non-existent (see File:20191021 Temperature from 20,000 to 10,000 years ago - recovery from ice age.png as an example using only separate dots) so that the smoothed trace is much more dominant than the spikey annual data,
(2) remove, reduce in number, or make lighter, the grid of horizontal and vertical lines (lay readers don't need precise grid lines),
(3) change "Temperature vs baseline" to "Temperature change" (less technical, for lay readers), and
(4) consider eliminating the "Annual mean" and "Five year average" legends altogether as being unnecessary for those who understand smoothing, and confusing to those who don't.

RCraig09 (talk) 15:57, 13 February 2020 (UTC)[reply]

I make these suggestions to make the chart friendlier for our predominantly non-scientist audience. Feel free to comment on these proposed changes. —RCraig09 (talk) 02:25, 20 February 2020 (UTC)[reply]