1

Where could I ask question like this one?

Systematic bias in NCDC GSOD climate dataset?

Open Data is good, but maybe some site dedicated to this area in particular would be good.

3
  • the bureau of meteorology :) – user310756 Aug 9 '13 at 10:45
  • 1
    Im not aware of a stack exchange site for that (physics might be the closest), all the SE sites are here stackexchange.com/sites#science If you want a stack exchange site on climate started propose it at area 51 – Richard Tingle Aug 9 '13 at 10:53
  • That said your question does seem to be purely about the data; what that data means doesnt directly matter – Richard Tingle Aug 9 '13 at 10:57
4

There is a Weather Forcasting proposal on area51.

You can support it and if it gets enough of a following it will become a Stack Exchange site.

At the moment, there is no suitable site for such a question.

3
  • "weather forecasting" seems to be very narrow field of climatology - I am asking on data from past 30 years. And honestly, I doubt site proposal formulated as "for hobbyists, storm chasers, weather forecasters" will attract any professionals who understand the topic. – Tomas Aug 9 '13 at 12:06
  • @Tomas - fair enough. By all means - suggest a climatology Stack Exchange - perhaps similar to MathOverflow, which is for research level maths. – Oded Aug 9 '13 at 12:13
  • I think the Geoscience proposal is much more viable. It's not about math but usualy the proposals made without focus on actual practical needs end up closed. Thank you though, inspired me to look at area51 again. – Tomas Aug 9 '13 at 12:51
4

In this particular case, you seem to be asking a question about how to do a particular piece of statistical analysis: I read it that you're looking for a statistical method to quantify, and put confidence on, any bias.

Cross Validated is the best place for that and other questions about statistics and data visualisation.

For questions about climate and sustainability, there's the Sustainability Stack Exchange.

The GeoScience proposal on Area 51 could also take such questions, should it launch. You can help get it launched, by following the proposal, by adding example questions, and by voting on the existing example questions.

And I've just recreated the Climate Change proposal on Area51, which could also take such questions

Geoscience Climate Change, Mitigation and Adaptation

5
  • "you seem to be asking a question about how to do a particular piece of statistical analysis" - wrong. You misread the question. BTW I don't think it is a good idea to create "Climate change" proposal. It has once been closed and the lack of followers suggests it should fall under Geoscience. – Tomas Aug 9 '13 at 14:09
  • Hi Tomas. I did read the question. And I think, whether you know it or not, in the first case your question is a statistics one: is this pattern significant? Their chat room might be able to help turn it into a good stats question. As for the Climate Change Q&A, it will stand or fall on its own merits. If you don't want to see it launched, don't join it. I think the description of it makes it clear why it's not a subset of Geoscience, but do open a question on discuss.area51 about it if you like. – 410 gone Aug 9 '13 at 14:12
  • I think it is clear that statistics is not a problem there; it is the data itself. I am not asking how to perform linear regression. – Tomas Aug 9 '13 at 14:20
  • I know you're not. I'm trying very gently to suggest that a naive linear regression maybe isn't the very best tool for the question you're trying to ask. I know from Cross-Validated that your statistical abilities are more sophisticated than that. – 410 gone Aug 9 '13 at 14:31
  • Well, I use only the tools actually needed for the task. Even the naive regression is enough in this case to show there is some principial bias, not random or rounding errors. You can comment on the statistics used and propose some better methods. I don't see how could this shift me in finding the real cause of the problem behind this pattern, but you can try it. – Tomas Aug 9 '13 at 14:39
0

I think it is clear that there is no better site now to ask than it is currently asked - Open Data.

The Geoscience proposal on Area51 seems to be the most viable site to ask in the future.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .