Thirty Meter Telescope

Astronomy's next generation observatory.



TMT is a major research facility, and represents, along with the Atacama Large Millimeter Array (ALMA), a level of financial investment in astronomical observatories on the ground well beyond what has been done before.

Accordingly, considerable attention is being given to maximizing the science return from the TMT Observatory. The approaches that have been adopted include: enabling efficient science observations as early as possible; supporting a suite of broadly capable cutting-edge science instruments; providing a range of support capabilities for astronomers for observation planning, operation and data verification.

These steps are being taken to ensure that the facility is operated efficiently and effectively, commensurate with the large investment of the TMT partners. The uniqueness of TMT data amplifies the importance of such datasets and that it be archived and made available to the international science community (following a typical proprietary period of 18 months). In addition to optimizing the science return of TMT, the need for education and public outreach is a priority for TMT.

The key features and modes expected for TMT at first light are summarized below, as are the current plans for implementation of added capabilities within the first ~5 years. While the details are still under discussion, and will remain so in some cases as new partners join the Project, a complete and agreed upon Operations Plan will be required by the start of construction (expected in 2012-2013).


TMT will be a fully integrated adaptive optics (AO) telescope with a Nasmyth-mounted suite of powerful instruments. The ability to switch rapidly (~10 min) between different instruments gives observational flexibility to respond to changes in scientific requirements (e.g., time critical observations; multi-instrument observations) and weather conditions. The baseline observatory and operational modes are being designed to match these capabilities, giving maximum science return and to enable direct input to the observational program by the Principal Investigator.

Current experience suggests that optimal use of the telescope is enabled through having both “classical” and “service” modes available for the TMT Partner astronomical community. Thus TMT observing will be possible through either the classical approach involving direct control by a PI and her or his team, or through a service mode, both utilizing an efficient queue-based observing approach for optimal efficiency.

“Classical operation” by a Principal Investigator (PI) and their team:

  • direct control and real-time decision making
  • PI and team operates the instruments & executes the observations
  • assisted by a TMT Support Astronomer
  • PI plans observations utilizing various queued science operation tools
  • observing assistants will operate telescope & AO/LGS systems
  • time allocated in blocks of half or integer nights

“Service Queue” by TMT Support Astronomer (SA)

  • SA carries out observations on behalf of a number of PIs
  • utilizes science operation tools to plan an efficient sequence of observing blocks
  • makes informed real-time decisions to match current observing conditions
  • assisted by a suite of data management tools
  • optimization process used for handling of observing blocks & instrument configurations

While it is a challenge to project the required mix of classical and service observing, the baseline being used for planning is that TMT is expected to steady-state approach 50 percent in each mode in the steady state after about 5 years of operations. The current plan is for largely classical observing immediately after first light, transitioning to a ~50-50 mix smoothly over the first 5 years of early operations.

TMT will have a distributed observatory operation which makes good use of high-speed communications networks for remote observing and science operations, as well as science data management and data distribution.

Extensive use of remote observing and eavesdropping capabilities is expected to be used in both classical and service queue modes. Most observers will not be in close proximity to the telescope and will utilize remote observing facilities.

In service queue observing, an eavesdropping mode can be utilized by PIs to monitor observations being done on their behalf, to evaluate the data, and to interact with support astronomer performing their service observing.

Consideration of the capabilities enabled by this and advances in broadband technology led to consideration of two innovative new observing modes.

These could be baselined in the operations plan as ways to improve the overall science return and the quality of the science data output:

Dynamic classical

In this mode classical observers are awarded observing time and asked to nominate time windows when they might be available to perform their observations. They are then contacted at short notice (~one day) if availability matches the expected observing conditions and instrument readiness. This exploits the advantages of both classical observing and scheduling flexibility to remotely conduct classical queue observing when best matched to conditions, and when the required dataset does not well-match the 0.5 or 1 night scheduling blocks.

Interactive service

During service observing, an automated system alerts the PI that his/her observation is being executed, then enables a direct voice/video connection to control room. The PI can then communicate with the service SA carrying out their observations to interact during target acquisition & view/assess incoming data.

Target of Opportunity (ToO) procedures are integrated into all observing modes. A rapid response mode can be triggered using automated process to exploit TMT’s abilities for fast acquisition and switching (~10 min) between multiple instruments. Time critical and synoptic observations can also be integrated into the observing sequence in either classical or service modes.

On-line data processing tools and automated quick-look evaluation data pipelining will be provided. These will be provided by the science instrument teams as part of their “quick-look” data verification and data analysis software deliverables, and can be utilized by observers in either classical or service modes. At the start of operations science data pipeline processing is expected to be limited to the quick look capabilities, but the underlying data archive structure and data management process would allow upgrades to a full pipeline data processing system if required (and funded). The data is expected to be Virtual Observatory (VO) compliant. Data archiving during early operations will offer basic search and retrieval capabilities and TMT science support staff will assist users in securing data access. An upgrade is possible to provide more extensive services similar to centers like Spitzer and HST with interfaces into the Virtual Observatory.

Each partner maintains ownership of data obtained during the time scheduled for its scientists for a nominal proprietary period of 18 months.

TMT is then committed to release science data to the international science community

Observing time would be allocated according to the indexed contributions by partners to the construction of the Observatory and its science instruments, and to the annual operation and instrument development costs.

Each partner will have a Time Allocation Committee (TAC) which ranks the proposals from their observers and recommends initial time assignments fitting within partner’s total observing time. These will then merged via a observatory-led scheduling process.

Since TMT will be a complex system for infrequent users, a robust user support system will be made available with knowledgeable science operations staff to answer questions, offer advice, follow-up on issues and concerns, and provide prompt feedback and responses.

Education and public outreach is a priority for TMT and will be an integrated element of the headquarters operation. This activity will provide assistance to partner EPO activities.

A possible North American Operations Support Center and TMT Science Center is being considered which would manage a number of user support functions such as: remote observing, observation scheduling and execution support; data delivery and processing support, as well as providing added education and public outreach activities.