[LTER-im] publishing data collected outside of your LTER

[Gastil Gastil-Buhl]

Hi Tim,

That is a good discussion topic. I will respond with my experience at my
own site, but I will stop short of suggesting that how we do this at MCR
LTER is an official "best practice".

Most, perhaps all, new LTER sites begin after time series data has already
been underway for some time. In fact I believe this is part of what makes a
site attractive as an LTER site. These data are called "Legacy" data. At
MCR there were several legacy datasets that had been collected by
researchers who wrote the proposal for our site. Not all of those are still
ongoing; a few are "completed". My site did put them in the MCR LTER
catalog before I arrived and I maintain the ongoing ones, adding data each
year. I suggest it is useful to put these data into the site catalog even
if they are not ongoing because they form a context for future research at
your site.

We do have some data for MCR LTER that was created by others but we
cataloged. We call these "Exogenous" data. (Limnologists will get the
meaning.) Examples from my site include some Landsat image data. Sometimes
we call these "Reference" data, but for us that has a slightly different
meaning. Our fish taxonomy dataset is a compilation of data from external
sources, used by our local datasets; so we call that "Reference data" and
catalog it at our site. Exogenous datasets are almost the only ones in our
catalog where we put the owners and creators in the metadata as non-LTER
organizations, ie USGS.

The above examples my site has put in the LTER Network Data Catalog. A
counter example I can provide, as contrast, is a dataset where the data's
only association with MCR LTER was that it originated from one of our
investigators while working on a non-LTER project. That dataset he had
curated by EDI. It ends up in the same catalog, but thru the EDI portal,
not the LTER portal. Another counter example are some cruise data collected
for MCR LTER that went directly from the research vessel to the BCO-DMO
repository ("Rolling Deck to Repository").

One compromise you face is convenience to your site investigators of
designing the dataset best suited to them and trivial ease of location,
versus using external resources to curate (saving you time) and having data
users find the data thru more general catalog searches. As a new site it
may be tempting to fill your local data catalog so it does not look so
empty at the start. (We did that.) Keep in mind this may oblige you to
maintain any time series that looked like LTER at the start but may not
actually be LTER in the big picture.

Good question!
Gastil

On Sat, Jul 28, 2018 at 10:26 AM, Whiteaker, Timothy L <whiteaker at utexas.edu
> wrote:

> Dear IMs,
>
>
> What is appropriate with respect to archiving data (e.g., submitting data
> to EDI) that wasn't officially collected by your LTER site?  Here are some
> examples.
>
>
> Our new LTER will build from some long-standing work in our study area.
> There is a large database of observations data that we're inheriting from
> that prior work.  We will continue those observations.  Should I package up
> and archive the old observations, or just the ones produced since our LTER
> site started?
>
>
> Suppose we download GIS datasets published by others and crop them to our
> study area.  This becomes part of our project GIS that our whole team
> uses.  Is this something we should submit to an archive, even though the
> data may already be available (in uncropped form) elsewhere online?
>
>
> What about USGS stream gauge data that we use to calibrate our hydrologic
> models?
>
>
> Thanks!
>
>
> Tim Whiteaker
>
> BLE Information Manager
>
> _______________________________________________
> Long Term Ecological Research Network
> im mailing list
> im at lternet.edu
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lternet.edu/pipermail/im/attachments/20180728/490077ee/attachment.html>