Preliminary/Draft Under Discussion

Final Report of the Survey Working Group

July 20-21, 1998

National Optical Astronomy Observatories, Tucson, Arizona

 Working Group Members

Stephen Strom, Chair

University of Massachusetts

Todd Boroson

National Optical Astronomy Observatories - Science Operations

James Cordes

Cornell University

Kenneth Janes

Boston University

Marc Postman

Space Telescope Science Institute

Tom Soifer

California Institute of Technology

Nick Suntzeff

National Optical Astronomy Observatories - CTIO

Tony Tyson

Bell Laboratories

Dennis Zaritsky

University of California, Santa Cruz - Lick Observatory



The rapid evolution of large format optical and infrared detectors, combined with the growing availability of high speed, low-cost computing and storage media enable astronomical research qualitatively different from that feasible just a few years ago: surveys of large samples of objects spanning a wide range of physical characteristics and evolutionary states. Such surveys provide both the basis for discovering new phenomena and patterns, as well as the statistically well-understood target lists, and ancillary catalogs of acquisition and wavefront sensing stars essential for planning programs that require the unique capabilities of large O/IR, radio, and space-based telescopes.

The role of surveys in the era of large ground-based telescopes was explored in some depth in the course of a workshop sponsored by NOAO and held in September 1997. More than 50 scientists with interests spanning solar system science to galaxy formation, identified a suite of specific surveys critical to advancing research in their fields. Members of the workshop recommended that NOAO develop plans

  1. to make extant facilities available for survey use;
  2. to develop instrumentation and new facilities optimized for carrying out surveys;
  3. to develop capabilities pipeline processing, archiving, archive-querying, and archive mining tools to facilitate survey science and make the results of multiple survey programs available to a broad community of potential users.

They noted as well that opportunities to propose surveys, and policies that ensure public availability of survey data products, are both essential to enabling all members of the US community to plan competitive science for new generation, large telescopes.

In response to the first of these recommendations from the survey workshop and to a parallel recommendation from the joint CTIO/KPNO User's committee, NOAO convened a "survey working group" to examine how NOAO might enable astronomers to propose and carry out surveys using NOAO facilities.


Charge to the Survey Working Group

The working group met on 20 and 21 July, 1998 in Tucson and was asked to respond to the following charge:

  1. develop a framework for proposing, evaluating and scheduling surveys using NOAO facilities;
  2. within this framework, discuss experiments/processes which will refine NOAO's understanding of the capabilities required to carry out surveys efficiently (e.g. telescope scheduling philosophy; mechanisms for developing pipeline processing tools), and to create widely accessible "living databases";
  3. recommend a near-term process to evaluate and schedule surveys proposed to enable efficient access to Gemini during early operations. More specifically, the group was asked to develop:


Conclusions of the Survey Working Group

The consensus conclusions of the SWG are summarized below.


Definition of a Survey

As a working definition, a survey should:

The characteristics of rich scientific content and high value to a broad community distinguish "surveys", so defined, from specialized long-term research programs. To be scheduled on NOAO facilities, a survey should:

While we anticipate that most surveys will make use of instrumentation normally available on NOAO facilities, groups should not be discouraged from proposing surveys, which make use of novel "user-supplied" instrumentation. In such cases however, survey reviewers must be convinced that the proposing team has developed credible plans to serve a larger community, and to minimize impact on other NOAO operations.


Survey Deliverables

Teams proposing surveys should be prepared to provide specific deliverables that are defined by the goals of the survey and the style of the survey team. Among the most important deliverables expected are:

The survey working group recommends encouraging proposing teams to take advantage of extant NOAO pipelines for processing and archiving images and/or spectra, to ensure timely availability of uniformly reduced data. Key survey output (images, spectra, catalogs, etc.) should be placed in a long-lived archive (e.g., at STScI), accessible with well-developed search tools. It is important as well to archive "raw" data (or sets of data, which enable recovery of "raw" data if needed) to allow for future data-mining using new algorithms. These "raw" datasets need not be available via the standard, "on-line" archive.


Balance of Public Needs and Survey Team Incentives

The survey working group recommends that NOAO invest up to 20% of the time available on all facilities to carry out surveys judged to be highly meritorious by the TAC process described below. We believe that the benefits of this approach will outweigh the loss of time to individual research programs, primarily because surveys enable new classes of science, can serve large numbers of individual users, and are important tools in enabling a broad community to prepare effective proposals to make use of the new generation telescopes.

We note that NOAO is a publicly-funded institution, and given the proposed significant investment of time, scheduled surveys must meet the high standard of serving a broad community of scientists (as noted above). The expectation is that survey "deliverables" will be freely available after a minimal proprietary period so that a large number of users can benefit from the database in a timely manner.

At the same time, we recognize that the survey team will invest significant time and resources in producing the survey and should there receive concomitant return for its efforts. We therefore recommend that a survey team be permitted to propose a schedule of release dates which reflect a proprietary period of up to 12 months following the ingest of each dataset into the survey archive.

The survey working group notes the major long-term advantages of developing strong cooperative working relationships between the survey team and NOAO. Of especial importance is development of a "knowledge base" (embodied by NOAO staff) which over time should make each new survey more efficient and cost-effective.

Practical considerations in planning the survey should make such cooperative relationships attractive to both the survey teams and NOAO. In particular, NOAO could provide:

In turn, the survey team can provide the labor necessary to carry out the survey (students; post-docs); funding; new analysis methods.

Although the magnitude of survey efforts may vary, we imagine that most projects scheduled will involve teams of 3-10 individuals with steady-state "level of effort" of between 2 and 5 FTEs/year.


Evaluating and Scheduling Survey Proposals

We recommend the following process for evaluating survey proposals:

As now done by the "discipline panel" chairs for extragalactic and galactic science, the chair of the "survey panel" will advocate survey proposals to the "super TAC".

The initial investment of time in surveys should not exceed 20% on extant NOAO telescopes, but could be much larger on HET and MMT (see below).

In order to work out operational details and to ensure an appropriate level of interaction with NOAO staff and other survey groups we propose:

In order to ensure that survey teams are making effective use of their time and providing "deliverables" on schedule, the survey working group suggests an annual review of survey progress. As part of the review, each survey group would meet individually with the "survey panel" to make a formal presentation of "progress against plan". The survey panel would then provide feedback to the survey team, including recommending schedule recovery plans or restructuring of the survey if necessary, additional time if warranted, or termination if appropriate.

We also believe it important to publish updates regarding survey progress on the NOAO Web pages and in the NOAO newsletter. We recommend that survey proposals be reviewed annually as part of the spring TAC process. This will enable successful proposing groups to prepare proposals for funding (if needed) in time for review by the NSF and many NASA programs.

We also recommend that the overall effectiveness of the survey program and the balance of survey/individual investigator time be subject to review every 3 years by appropriate NOAO committees (Users Committee; Observatories Council).


Proposals for Surveys

Proposals for long-term allocation of large blocks of time must necessarily be more comprehensive than individual investigator proposals. The survey working group recommends that the proposal content be as follows:

Scientific Case:

Technical and Management Approaches:

Ancillary Benefits:


Recommendations Regarding Facilities Available for Surveys

There was thus strong sentiment to reserve MMT time exclusively for survey work when these instruments become available. The survey working group recommends that NOAO recognize the importance of making investments in:


Capturing Survey Experience

As noted above, the SWG considers it essential that in the process of supporting the activities of survey teams, that NOAO develop over time a suite of survey capabilities and a knowledge base of "do's and don'ts" for carrying out surveys efficiently. We have specifically recommended:

We strongly urge that over time, NOAO and the community develop a culture of collaboration on survey efforts. Elements of this culture would include:

Posted: 22Sep98