We give here a brief overview of the design specification for the telescope, as driven by the science case. You may also be interested in our SPIE paper on the subject, which is available here on arxiv.org.
The next generation of transient surveys such as LSST will detect transients with typical magnitudes which are beyond the capabilities of 2-metre class telescopes for spectroscopic follow-up. The single visit depth of LSST will be approximately 24.5 in the r-band, so many transients will be beyond the capabilities of even 4-metre class telescopes. However, the pressure on 8-metre telescope time is high and will still be high in 2020+, so 4-metre class facilities will need to do the bulk of the follow-up work, and this work will be limited by the facilities available, rather than the number of transient detections (this is true even today). Recently the Public ESO Spectroscopic Survey of Transient Objects has shown that dedicating large amounts of 4-metre time to spectroscopic follow-up of transients can be extremely scientifically productive.
A robotic telescope with a 4 metre aperture would be a powerful tool for the study of supernovae, since the flexibility of robotic scheduling allows observing cadences to be optimised for each individual object. As well as supernovae, we are also interested in more exotic classes of transient. Experience has shown that the response time of a robotic telescope makes it a particularly powerful tool for the study of rapidly fading transients, such as the afterglows of Gamma-Ray Bursts. The typical response time of the Liverpool Telescope to a Swift trigger; including slew time, mirror settling time, instrumental overhead etc., is between one and two minutes. If this response time can be reduced, the rapid decay rate of the afterglow means it can be observed when it is considerably brighter: response time commonly trumps aperture, in this particular case.
Our concept for LT2 is therefore for an extremely agile, fast slewing telescope. The aim is that from receipt of a trigger from a survey facility, we can be on target and taking data within 20-30 seconds. Triggers will come from proposed missions such as CTA and SVOM, which promise to publish real-time alerts for rapid follow-up of high energy transients. Additionally, a key goal of the time domain community in the coming decades will be to detect the fading electromagnetic counterpart of merging neutron star binaries detected by aLIGO via their gravitational wave emission. The response time of LT2 should enable us to make a key contribution to this programme. As well as fast-fading transients, this capability makes other types of observations possible. Consider for example, a programme in which there is a need to obtain short exposures for a large number of objects dispersed around the sky. The overheads of a conventional telescope would make such a programme too inefficient to be feasible.
The key for fast slewing is to minimise the moment of inertia, and so a fast slewing telescope will be a lightweight telescope, with a thin primary mirror and perhaps a novel choice of materials for the structure. The weight of the instrument payload will also be a consideration (see below), and acquisition time must not be limited by the design of the enclosure. Our initial optical design studies advocate a standard Ritchey-Chrétein layout, with a final focal ratio in the range f/6.5 - f/10 and a primary mirror focal ratio of f/1.5 or faster. This fast primary allows for a compact design in which the separation between the primary and secondary mirrors is similar to that of the LT. The weight of the primary mirror is also a key consideration. We estimate a thin meniscus, 4-metre primary would weigh approximately 5500kg. To reduce this we have been considering constructing the mirror out of thin hexagonal segments. Previously, segmented mirror systems have been the domain of telescopes larger than 8 metres in aperture, necessitated by the capacities of furnaces and polishing machines. However, the E-ELT will require 798 segments for its primary mirror, and so such segments will need to be produced on an industrial scale. This has the potential to simplify procurement of segmented mirror systems for smaller projects. One of the centres currently engaged in developing 1.4m prototype hexagonal segments for E-ELT is Glyndwr Innovations Ltd., based in St. Asaph, North Wales - less than 40 miles from Liverpool. We are working with Glyndwr to investigate the feasibility of using a segmented mirror on a telescope of 4-metre class. Our initial findings suggest a mirror consisting of 6 or 18 segments would be suitable, and would have a total weight in the region of 1400 - 2700kg.
The Liverpool Telescope benefits from a diverse instrument suite, from the main optical imaging camera IO:O (shortly to be complemented by an infrared component) to more specialised imagers (RISE, IO:THOR), to the RINGO line of fast-readout imaging polarimeters, to the FRODOSpec fibre-fed optical spectrograph. Since the telescope is operated entirely robotically all instruments are mounted simultaneously and instrument changes are automatic, with an overhead of a few minutes. This provides a powerful and flexible capability. For example, when a target-of-opportunity observation of a fast-fading object like a Gamma-Ray Burst afterglow is triggered, the sequence of observations is automatically optimised to the target and conditions. An initial image is taken, from which the brightness of the afterglow is measured by the software and the subsequent observing sequence; including the choice of instrumentation, exposure times etc., is automatically decided upon. The aim is to make the same decisions that a human observer would make.
The ability of the telescope to operate a variety of instruments simultaneously brings an additional benefit: it has allowed us to inclue some instruments which are highly specialised in purpose. The fast imaging camera RISE, designed for the study of transiting exoplanets, is one example. Such specialised instruments tend to be cheaper, but more importantly they have a shorter development timescale. This has enabled us to rapidly bring new instrumentation to the telescope in order to pursue new scientific directions as they have arisen, and thus keep the Liverpool Telescope at the cutting edge of time domain astrophysics.
The majority of Liverpool Telescope data has been obtained using its imaging cameras. However, with the rise of the all-sky 'synoptic' surveys the need for photometry from smaller telescopes will be much reduced. It will not disappear entirely - for example there will still be a pressing need for high-cadence photometry of the large numbers of objects which vary on short timescales (seconds or less). However, our discussions with other members of the time domain community have shown that the most pressing need, both now and in the next decade, will be for intermediate resolution (R ~ few thousand) spectroscopy. Our intention therefore is that the main instrument for LT2 will be a spectrograph. The wavelength range will cover both optical and the near-infrared, enabling us to follow extra-Galactic transients such as supernovae at higher redshifts. The design of the spectrograph is to be decided: traditionally a long-slit design maximises throughput and the successful SPRAT spectrograph on the LT has shown that reliable automated acquisition onto a slit is possible for a robotic telescope. However, an IFU design (as is used for FRODOSpec) potentially enables more rapid acquisition and feeds into the LT2 design paradigm.
We recognise the diverse instrument suite of the Liverpool Telescope is one of its core strengths, and this capability is something we would very much like to maintain in LT2. The practicalities of this is something we are discussing at this early design stage: there may be compromises to be made between a fast-slewing facility and the weight of additional instruments. However, there may be ways of satisfying both requirements, such as mounting all instruments at Nasmyth focal stations, for example.
The contribution of instrumentation is potentially one way in which a group might join the project: such collaborations were effective when developing the instrumentation suite for the Liverpool Telescope. We are particularly keen to exploit new technologies that will enable us to meet our goal of an extremely rapid response. EMCCDs for example are now a fairly mature technology for astronomical observations, and enable very short exposures that are not limited by readout noise. Looking into the future, technologies such as Kinetic Inductance Detectors are particularly interesting for our purposes, since they record both temporal and spectral information from individual photons.
If you have any ideas for LT2 instrumentation we would be delighted to hear from you.