Data Policy Implementation Guidelines

Implementation Guidelines

1. It is the responsibility of the Platform Principal Investigator(s) to work with all Scientific Investigators involved in a Research Project to determine the status of the collected data with regard to restrictions on access, distribution and use. The Research Project Investigators should reasonably determine the appropriate data restrictions to be applied to the collected data (including the scope of affected individuals). The Investigators are not required to make their desires known to EOL or OFAP prior to the award of a Research Platform.




2. Restricted Access: Platform PIs may request that the data collected from a particular Research Project be subject to restricted access, distribution and/or use for a period of one (1) year from the date the Production Data Set is ready for distribution, provided that:
   1. no third-party contractual obligations or understandings state otherwise;
   2. at least two months prior to the start of the observational (data acquisition) phase of the Research Project, a written request is submitted to the EOL Director stating the specific details of the desired data restrictions and the reason(s) for requesting them; and
   3. all Research Project Investigators holding substantial interest in the Project must concur with such data restrictions.

   The EOL Director, at his/her sole discretion and before the observational (data acquisition) phase of the Research Project, will decide whether to grant the request.

3. During the first year after a Production Data Set has been released for a Research Project, the appropriate EOL Facility will notify the PIs whenever EOL makes a distribution of those data to other than the Research Project Scientific Investigators. This notification may take the form of a public report on the EOL web server that can be accessed on-line at the discretion of the Platform PIs (it is then the PIs' responsibility to access such reports as frequently as they may deem necessary).
4. As a routine part of a Research Platform's deployment, EOL expects to distribute Low-Volume Production Data Sets free of charge to all Platform PIs. High-Volume Production Data Sets will be made available free of charge to all Platform PIs via network transfer. If arrangements are made at the time of Facility Request, up to two sets of Production Data may be provided on physical media (it will then be the responsibility of the Platform PIs to share data-on-media with other investigators, as necessary). Any investigator may obtain a copy of data-on-media upon reimbursement for Out-of-Pocket expenses.

   For the foreseeable future, and to the extent possible, all non-restricted Production Data Sets will also be made available for no-cost network transfer to Community members.

5. EOL will reduce support for older data sets. This is necessary due to budget realities as well as due to changes in technology. Full support will be provided to a data set for ten years after data collection. This support will include maintenance of the data set on the NCAR Mass-Storage System (MSS). Metadata and rudimentary code to access the data will also be supported for the same ten years. After ten years, data redundancy will no longer be maintained, and the access code and metadata may suffer from a lack of maintenance. Any data will be left on the NCAR MSS for as long as practical after this period, subject to characteristic limitations and lifetimes of Mass Store files. After ten years, EOL may not make provisions for migrating data to new storage frameworks.

   Data sets will be considered largely unsupported ten years after data collection. If a scientist desires specific assistance for recovery of (or access to) such unsupported data, that scientist may be required to reimburse EOL for all costs (including manhours) incurred during that recovery effort. Actual data recovery in a usable form would not be guaranteed.

6. EOL will not widely distribute Preliminary Data. EOL recognizes the value of Preliminary Data for field project decision support, and for review of data accuracy by the Platform PIs. Field Project Decision Support may require that Preliminary Data be made available in real time (or near-real time) from EOL platforms. There is a danger of scientists incorporating Preliminary Data into reports or publications derived from a Field Project (this incorporation may be entirely inadvertent). Therefore, Preliminary Data should not be incorporated into a Project Database for potential further distribution.

   When Preliminary Data are provided, these data will be provided to all Platform PIs equally, subject to limited practical considerations, and subject to any prior arrangements among those PIs or between the Platform PIs and other Research Project Investigators (see Items 1 and 2). Distribution of data may be either over a network, or via a limited number of data copies on convenient media (it may be required that those copies are equally shared among all PIs). In cases were a high-volume EOL platform is located remotely, it may not be practical to provide distant users with full-bandwidth data in near real-time; under those circumstances, equal access will be provided only to Platform PIs on a local network at the remote field site.

   It is the Scientific Investigators' responsibility to reasonably handle access and use of Preliminary Data in the field. Investigators may be required to access EOL data through their own computer systems, or to read media on their own drives. It is further recognized that EOL may not be able to directly assist Investigators in routine data retrieval/transfer, and Investigators should plan for this need when fielding their own staff. If special arrangements have been made with EOL by Platform PIs, and costs are appropriately covered in the NSF Deployment Pool budget allocation, EOL will accommodate Platform PIs by providing additional specified support (e.g., network access to remote sites, transformation of EOL data into other formats, or routine support in archiving data on their systems).

   Investigators provided with Preliminary Data agree that such data are understood to be unverified, and possibly subject to some error or inaccuracies. The Investigators also agree to not act as redistributors of Preliminary Data once the Field Phase of a project ends. Upon delivery of the Production Dataset, EOL requests that Investigators surrender or destroy all Preliminary Data, and copies thereof. EOL expects that all Investigators are aware of the potential problems with the use of Preliminary Data and, as experts in their field, will provide timely and needed guidance to EOL while EOL is evaluating the Preliminary Data prior to producing its Production Dataset.

7. A Project Data Center will be provided with a copy of an EOL Production Dataset provided that EOL recovers its Out-Of-Pocket Expense for such delivery. In the event a Platform PI desires to transfer their rights in a Production Dataset to the Project Data Center, such delivery, if appropriate, may be free of charge. In the event that Platform PIs have imposed restrictions on the Production Dataset, EOL may not release the Production Dataset to the Project Data Center without written authorization from the PIs. In the event a restricted Production Dataset is provided to a Project Data Center, the Project Data Center shall handle the Dataset in accordance with the expressed wishes of the Platform PIs.
8. As part of its mission, EOL develops new instruments and new measurement techniques. EOL often must use data from a Research Project to test and demonstrate new capabilities and/or report upon instrument function. Therefore, EOL requires unrestricted access to and use of such data for test and demonstration purposes. In cases where data restrictions have been granted, Platform PIs will be notified of intended EOL data use, and EOL will discuss such use with them in view of the PIs' imposed restrictions.
9. A Platform PI, and under certain circumstances, other Scientific Investigators, may be asked to sign a letter acknowledging that s/he will accept the EOL Data Policy and accompanying Guidelines.

Glossary

EOL:

The Earth Observing Laboratory of NCAR and its staff.

Community:

The scientific research community at large.

Facility:

A physical research facility, such as an aircraft or radar.

also:

One of the groups in EOL which includes the Research Aviation Facility, Research Technology Facility, Design and Fabrication Services and the Research Data Program.

High-Volume Dataset:

A large data set generated from Research Project data which requires considerable effort to copy and distribute.

Low-Volume Dataset:

A small data set (typically consisting of only several gigabytes) generated from Research Project data which requires minimal effort to copy and distribute.

NSF:

The U.S. National Science Foundation

OFAP:

NSF's Observational Facilities Advisory Panel

Out-of-Pocket Expense:

Costs, which may include but are not limited to: media, shipping, overhead, fees, administrative expenses and time incurred by EOL for processing and distributing data and/or datasets. Except where agreed to otherwise, it is expected that the data recipient shall bear the reasonable cost of data/dataset distribution subject to adjustments based on NSF funding obligations or the extent of Research Project contributions by EOL.

Platform:

(Often used interchangeably with Facility) A physical research facility, such as an aircraft or radar.

Platform Principal Investigator (or PI):

A Facility-supported Principal Investigator whose name appears on an EOL "Request for Facility Support", and to whom support for use of a research Platform has been granted for a Research Project. [Illogically enough, this term is also considered to cover instances where there may be many "PIs".]

Preliminary Data:

Any EOL data collected and/or processed during or after a Research Project that have not been completely reviewed and/or analyzed by EOL.

Production Data Set:

The presumed final, processed EOL data for a given Research Project that have been carefully reviewed and/or analyzed by EOL and determined to be the highest quality EOL can produce. Such data are not guaranteed to be free of errors, but EOL will, to the best of its ability, inform potential users of any data that are deemed questionable.

Project Database:

a special, usually non-EOL repository for data which is typically established as a result of a large-scale, national or multi-national, cooperative experiment. It may include Production Dataset(s) generated and collected by a Research Project.

Project Data Center:

The group that administers and maintains a Project Database

Research Platform:

a specific EOL instrumentation system or data-acquisition platform (e.g., a radar, aircraft, sounding system).

Research Project:

a program of scientific study or a field experiment comprising one or more data-acquisition episodes utilizing one or more of EOL's Research Platforms.

Scientific Investigators:

The general class of scientific participants, either for a given Research Project, or for the Community at large; this is a super-set, and includes Platform PIs as well as all Research Project Investigators.