CO2 PPM (long-term time series)

[rufuspollock]

We already have 50 years of CO2 #56 but what about much longer time series e.g. using ice core data. What's the best data we have there and where's it from?

[kiliakis]

@rgrp

• vostok ice core CDIAC has data about co2 concentration from 400k years ago
• NOAA has also co2 concentration data from 800k years ago

[rufuspollock]

@kiliakis can we have both in one data package? In any case, just pick easiest and start with that one.

[kiliakis]

@rgrp Here is a first version containing dataset from vostok ice core: https://github.com/kiliakis/co2-ppm-ice-core

Links from data.okfn.org:

• validation results
• view link

[rufuspollock]

@kiliakis perfect though i note the data table does not display because of the column title. That's a bit of a bug in viz but also reflects that column name has "nasty" characters - could i ask that we rename the column to something simple like "Age" and then have the full info in the description (as you already have)?

@pdehaye do you want then to review and we merge in?

[kiliakis]

Okay, I fixed the problem with the column names. Here is the view link: http://data.okfn.org/tools/view?url=https%3A%2F%2Fgithub.com%2Fkiliakis%2Fco2-ppm-ice-core

[rufuspollock]

@kiliakis looking pretty good - added a few comments in the issues of that repo. Now its time to transfer across to datasets.

@pdehaye please also cast your eye over for improvements.

[rufuspollock]

@kiliakis also strictly to satisfy this issue we'd want a nice dataset merging this dataset with some other co2 measures to get a nice continuous dataset up to the present with dates in a nice format. To be clear i think we want to keep vostok dataset but have another dataset like co2-ppm-longterm or something.

[kiliakis]

@rgrp I can merge the ice core dataset together with co2-ppm dataset and create the co2-ppm-longterm. There will be a gap between 2400 years ago ( end of vostok ice core data) and 1959 AD (start of annual co2 ppm measures ).

[kiliakis]

@rgrp I think the merging of co2-ppm and co2-ppm-ice-core into co2-ppm-longterm is completed.

• Validation link: http://data.okfn.org/tools/validate?url=https%3A%2F%2Fraw.github.com%2Fkiliakis%2Fco2-ppm-longterm%2Fmaster%2Fdatapackage.json
• View link: http://data.okfn.org/tools/view?url=https%3A%2F%2Fgithub.com%2Fkiliakis%2Fco2-ppm-longterm#resource-co2-mm-mlo
• datapackage: https://github.com/kiliakis/co2-ppm-longterm

Please review. I am facing a problem when trying to add two graphs from two different resources to the data package viewer. Is this possible?

[pdehaye]

I have looked at it in details.

The "Trends" and "Historial record" here are a bit confusing as I didn't see these terms used elsewhere.

I don't know about two graphs from two different resources. @rgrp ?

Otherwise it looks good!

[kiliakis]

@pdehaye I see what you mean. These terms are firstly introduced in the first paragraph of the README file and then used in the data description, citations and sources. I can change it though. @rgrp , @pdehaye Do you have any suggestions?

[pdehaye]

Actually, it is fine to use Trends and Historical series, but the first paragraph needs a bit of reorganisation and stream lining. In the first paragraph, do a Markdown list, and make sure in each it is the terms "Trends" and "Historical series" that appear first.

[rufuspollock]

@kiliakis merging in the co2-ppm does not make much sense to me tbh:

• the scales are massively out
• we end up with duped data which is never great

At most i would add in like the 1950 datapoint and the 2010 datapoint or something (but frankly i wouldn't botther).

Also when i look at the datasets page i just see the co2-ppm graph.

Overall: I would suggest we stick with historical data here and not dupe in the co2-ppm stuff (i agree we may want to rename co2-ppm to something more meaningful but lets make that another issue)

[kiliakis]

@rgrp ok no problem :)