Nasa Open Datasets | Hacker News Reader

Nasa Open Datasets

open.nasa.gov

208 pointssquiggy2212 comments

If you are purely looking for some beautiful datasets, checkout NASA's Datacasting website[1]. They have a large RSS feed directory[2] with a lot of data on earth science. They even built their own RSS feed reader[3][4] to download and filter the data. And I think everything is already geo-tagged.

[1] http://datacasting.jpl.nasa.gov/

[2] http://datacasting.jpl.nasa.gov/feed_directory/

[3] http://datacasting.jpl.nasa.gov/download/

[4] http://sourceforge.net/projects/datacasting/

Also check out https://data.nasa.gov/data?category=&search=&type=datasets (linked to from open.nasa.gov) which all have consistent APIs -- so you can write code once and re-use it as opposed to having to write a new API wrapper for each disparate data set

Disclosure: I work for the company that helps run data.nasa.gov

Know of any interesting examples/fun projects where this data is used?

Must be fun!

It will be interesting to see how much uptake such a top-level index gets.

The search capability (e.g., https://news.ycombinator.com/item?id=10264811) returns so many results for typical naive queries ("temperature" returns 8300), that it's impossible to sift through them. They seem to be un-curated, so a niche product with marginal calibration gets equal play to flagship products with thousands of users.

People like me who use lots of this data every day have sat in panels and workshops and mused whether a searchable top-level index would be useful, because this is not how scientists work. They ask their friends which sources are best for their use, e.g., which have the right coverage and sampling in space, time, and wavelength. And they settle into something, and maybe make a big shift to a new product when it comes online.

But: something like this index could be helpful, to citizens, decision-makers (e.g., tree canopy depth maps for fighting forest fires), and to out-of-discipline scientists ("what if I brought in NASA sea surface temperature to my analysis of beach restaurant revenue fluctuations?").

For CO2, here's an example (https://co2.jpl.nasa.gov) which gives 5 simple use cases and sketches why you'd choose one over the other.

First off, awesome, I'm always looking for new datasets. So these complaints are fairly small, but currently stopping me from doing some things:

I can't see more than two items on this page: https://data.nasa.gov/data?&search=&type=datasets I'm just after a big list of datasets so I can skim through and find ones I'm intrigued by. Here's a screenshot: http://i.imgur.com/TtP630G.jpg (latest Chrome, mac)

The featured dataset links to a 404 page: https://open.nasa.gov/nasadatanauts/

Again, great to see the data out.

Edit - scratch the listing problem, it seems to be one of my extensions, either readability or pushbullet, breaking the site.

It is great to see them releasing data, but I'm surprised they went with something based on Socrata given there are other OSS products out there. I can't imagine how much they paid - I guess that won't be in the data catalog anywhere :(

Could you name some of the alternatives to Socrata, I really like what it does and its interface but would prefer something free.

There's http://ckan.org (disclaimer, I work with this) and http://dataverse.org if you're more academically minded. There are others, but hard to judge any claims of open-sourceness without access to source code.

In addition to being a beautiful design I like how NASA is making use of personas. As I've discovered in working with open data it isn't just developers who are interested.

In Michigan mappers, citizen activists and journalists are drawn to it in equal numbers with developers. Giving different views that allow them to get what they're looking for is very powerful.

http://www.skywatch.co/

SkyWatch is doing something similar. If you're into this sort of thing, check them out.

Wow, this is hugely improved. Used to be lots of scattered files everywhere with poor access and no consistency.

Reply on news.ycombinator.com