Why is the VLT very efficient

1
Why is the VLT very efficient?

• Fernando Comerón
• On behalf of Francesca Primas and Martino
  Romaniello

2
A great science machine in full production

• Individual operation of 4 VLT Unit Telescopes
  (8.2m) and VLTI facility
• 13 science instruments, including 2 VLTI
  instruments (more to come)
• Support provided by a staff of over 80 people
  directly related to science operations (Europe
  and Chile), plus software developers
• Over 2,500 hours of science observations
  performed each semester
• Technical downtime below 3
• Weather downtime below 7

3
A great science machine in full production

• 702 VLT/VLTI runs approved in the current
  semester
• 507 individual PIs in the current semester (169
  of them new)
• Nearly two refereed science papers using VLT data
  every day.
• Thanks to excellent facilities and
  instrumentation, highly competent staff, and an
  advanced end-to-end operations model running over
  60 in Service Mode.

4
Service Mode essentials

• In Service Mode, time is allocated to programmes
  to be executed, not in time slots assigned long
  in advance, but spread over the observing
  semester as the conditions are optimal for each
  programme.
• The VLT operations model was designed from the
  beginning to include a large fraction (at least
  50) of Service Mode observing recognized as a
  key element to its scientific productivity and
  efficiency
• Flexible scheduling adapts to the external
  conditions
• Best observing conditions go to projects that
  really need them
• Top rated projects get the highest execution
  priority
• Some science cases require execution under a
  sustained period of stable excellent conditions.
• Long monitoring series (months, years) easily
  manageable
• Target of Opportunity observations less
  disruptive

5
Service Mode essentials

• And other byproducts
• Time available to projects that can make use of
  worse-than-average conditions
• Continue science with another instrument in case
  of instrument malfunction
• Optimize calibrations by sharing them among
  several programs
• Population of the archive with data obtained
  under well defined procedures

The community has increasingly accepted and liked
it!
6
End-to-end Service Mode in a nutshell

• Sophisiticated long-term scheduling taking into
  account programme needs, statistical expectation
  on the conditions to be encountered, and many
  other factors.

• Collection, optimization, and validation of fully
  pre-defined observations
• Science operations policies that enhance the
  overall efficiency of the observatory, even if
  they may occasionally impact specific programmes.
• Many heterogeneous factors taken into account in
  the short-term schedule.
• Standard calibration plan.

• Archival of all science, calibration, and
  environmental data

7
Dealing with a changing atmosphere in VLT Service
Mode

• Not one single set of constraints
• Most instruments operate with natural seeing, but
  some (NACO, SINFONI, CRIRES, VLTI) use Adaptive
  Optics
• Laser Guide Star performance especially sensitive
  to transparency
• Wavelength range covered from 0.3 to 20 microns
  relevant constraints are different

8
The promise of MCAO

• Multi-Conjugate Adaptive Optics successfully
  demonstrated last year at the VLT.
• Not regularly offered, but impressive results
  have triggered interest (and demand!) by the
  community
• Illustrates new constraints on operations
  scheduling

9
Dealing with a changing atmosphere in VLT Service
Mode

• The current shortcomings
• No prediction of the evolution of the conditions
  is systematically factored in the short term
  scheduling decisions
• Short-term scheduling mainly based on current
  conditions and some guesswork based on historical
  knowledge
• Limiting the length of individual observations to
  allow flexibility to changing conditions
  (increasing overheads)
• Improved forecasting by better understanding
  relation between focal-plane image quality and
  DIMM measurements

10
Dealing with a changing atmosphere in VLT Service
Mode

• The uncertainty in the short-term evolution of
  the atmosphere leads to two short-term scheduling
  strategies
• Risky starting the execution of observations
  with constraints very close to the current
  conditions
• Pros makes use of the best possible conditions
  and gives chance to the most demanding programmes
• Cons decreases efficiency by requiring frequent
  repetition of observations
• Conservative starting the execution of
  observations that allow for substantial
  deterioration of conditions
• Pros greater likelihood that the observation is
  completed within constraints
• Cons most demanding programmes left with little
  chances of execution, results below the best that
  the telescope can give
• A sizeable fraction of observations (20) need
  to be repeated because conditions went outside
  constraints during the execution an important,
  hidden source of inefficiency!

11

• Can atmosphere prediction tools further help VLT
  operations?
• In classical mode, weather prediction tools are a
  nice-to-have. In Service Mode, they are a boost
  of the efficiency
• More robust short-term scheduling, better
  fulfillment of the goals of Service Mode
• Demands to increase as new instrumentation comes
  on line (SPHERE)
• Probably the next major step in increasing the
  eficiency of the VLT

Yes, absolutely!