gavin971 / climate-1

Data and analysis of global warming

Climate

Summary

The goal of this project is to provide climate data and analysis in computable form (in terms of IPython Notebooks). Part of our goal is to reproduce graphs and results from literature. Ultimately we would like to start from the raw data (e.g. from temperature stations, satellites, ice cores and other measurements) and provide the analysis in a notebook, so that anyone can easily rerun it and see exactly how the final results and graphs were obtained from original data.

Preview the IPython Notebooks

If you just want to see the notebooks, the best way is to preview them with the links below. When you're ready to develop or run the code yourself, pull the source for this GitHub repository and execute the notebooks on your own machine.

• Satellite Temperature (RSS)
• Satellite Temperature (RSS) maps
• Berkeley Fit
• Temperature Data
• Sea Ice Index
• Sea Ice Index Arctic
• Temperature Fits Reproduction
• Sea Level Trend New York City

Motivation

The motivation for this work is the following:

• Reproduce graphs and results from literature.
• Provide notebooks that anyone can easily rerun to reproduce all our results exactly (including all the graphs).
• Provide a "pipeline", that takes raw data and calculates final results, so that anyone can easily see what exact analysis was done.
• Let the data speak for itself, we try hard not to "jump to conclusions" (or have "leaps of faith"), that does not strictly (scientifically) follow from the data.
• Following Feynman's "never trust the experts", we want anybody to form his or her opinion about the issue, and this project might help by doing the hard work (of getting the data, understanging it, analysing it and calculating results/graphs), that one has to do anyway in order to understand the issues. By the nature of the process, the analysis can be wrong, but by making all the analysis public in a computable form, it can be discussed, improved or dismissed.
• In particular (as implied by the previous points), we try to strictly only include analysis that is scientifically agreed upon and undisputed, or if that is not possible, to provide all possible weaknesses of the given approach.

Please send us ideas for improvements (for example you can open a GitHub issue or send us an email) or send a GitHub pull request (even better :).

License

All files (notebooks, Python code) are licensed under MIT license, see the LICENSE for more information. The data directory contains files that were downloaded from various sources and they are covered by their own respective licenses.