Director's Message - Exploring Astronomy's Time Domain

Figure 1. This illustration (presented during the 1995 SPIE conference on astronomical instrumentation) is an early view of the planned Gemini cassegrain instrument cluster. A total of five beam feeds are provided around a rigid cube containing a highly articulated fold mirror and acquisition and guiding system. The design includes a calibration unit and an adaptive optics module that can feed any instrument. Electronics enclosures using air-liquid heat exchangers remove heat from the telescope environment. An infrared optimized up-looking port is preserved as well. This marvel of engineering serves as the “central nervous system” for each Gemini telescope and is fundamental to the success of our Target of Opportunity program.

By Doug Simons

Perhaps the only thing predictable about nature is its unpredictability. Of course, many natural events are highly predictable, such as the ebbing of tides and changing of seasons. But, as one tunnels deeper into natural phenomena, the vast majority of the universe appears unpredictable simply because it is so poorly understood. The splendor of the universe stems from not only its beauty but its capricious violence. Understanding natural chaotic phenomena is a field in which Gemini’s incredibly creative research community is growing adept, using Gemini as an observation platform. Back when Gemini was originally designed, however, we frankly never predicted these giant portals on the universe would be so effective at exploring astronomy’s time domain. Several recent observations beautifully illustrate how engineering decisions made long ago can sometimes, for one reason or another, actually work together for a completely different purpose today. The result is a powerful new research tool for our community. This could be a lesson worth noting as the next generation of giant telescopes is designed, and similar tough design decisions are made to help them remain within budget and seemingly limit their capabilities. The “moral to the story” is to not underestimate the discovery potential of an inventive community, powerful telescopes, and a target-rich sky to take astronomy in unexpected directions.

Like many observatories, the Gemini Observatory we know today bears, at best, a limited resemblance to how it was first envisioned roughly two decades ago. A tremendous amount of effort went into finding an optimal system performance within the Gemini 8-meter Telescope Project’s finite budget. Some of the effects of this design process are obvious, others less so. A key early decision eliminated the Nasmyth foci at Gemini, leaving the telescopes with unconventional focal stations compared to our contemporary facilities at Keck, Subaru, and the Very Large Telescope. Having instruments function exclusively in a Cassegrain environment, compared to a more benign (and spacious) Nasmyth focus, effectively transferred complexity and risk from the telescopes to Gemini’s instruments—all of which must fit within a 2,000-kilogram mass limit (900 kilograms for the adaptive optics (AO) systems) and limited space envelope, and function under any gravity vector.

Recognizing these limitations, and out of concern for the notorious service and maintenance problems that frequent instrument changes cause, Gemini’s engineers developed an innovative multi-instrument cluster capable of rapid remote reconfiguration while the instruments remain on the telescope for months at a time. It featured an AO system that could direct a corrected beam into any instrument mounted on the telescope. Merged with a clever calibration system (which replicates the telescope’s beam), an acquisition and guidance (A&G) system (featuring three wavefront sensors), and on-instrument wavefront sensors (built into most instruments), the cassegrain cluster at Gemini packs an enormous range of capabilities and sophistication into a relatively small space. Figure 1 shows the early (~ 1995) Cassegrain focal station concept at Gemini. Interestingly, this design was well underway before serious consideration was being given to running a queue-based science operation model, and the term “Target of Opportunity” (ToO) was seldom, if ever, heard in the bustling Gemini Project office while this system was being designed. Ironically, being forced to consider “out of the box” solutions, due in large part to the highly cost-constrained environment in which Gemini was designed, bore the “fruit” hanging on the back of each Gemini telescope we enjoy today.

Another important design distinction at Gemini that is crucial to our ToO program’s success is, of course, the use of a queue-based operations model. Even in this case, though, the rationale behind adopting a queue-based model was only weakly linked to ToO’s at its genesis. With telescopes that were engineered to be site-limited in performance, and having several instruments available concurrently, the dynamic range of the facility’s sensitivity and possibility of matching observing programs with changing weather conditions to maximize Gemini’s scientific product is what really drove the queue concept behind the scenes. Finally, the last key ingredient in Gemini’s ToO system is simply the fact that Gemini—with twin telescopes—has access to the entire sky; again, this originated in large part to provide ground-based back-up for spacebased observations.

The advantages in running a ToO program with 8-meter telescopes on either side of the equator and separated by ~10,000 kilometers was not a strong driver at the time. In the end, it was a unique combination of clever engineering, global coverage, and queue-based science operations that made Gemini the “ToO machine” that is so powerful today. I wish those of us on the original project team could claim credit for having the far-reaching scientific insight about research trends in the 21st century and using that insight to deliberately design an observatory that is so effective for ToO observations. However, the truth is—we just got lucky.

The ToO system now in place makes it possible to receive a signal at either Gemini telescope via the Internet from a principal investigator (PI) anywhere in the world. That signal, when handled by the night staff, allows them to seamlessly interrupt say, a Near- Infrared Imager (NIRI) queue observation, slew the telescope to the opposite side of the sky, configure the Gemini Multi-Object Spectrograph (GMOS) for single-slit spectroscopy, and begin collecting photons from a ToO source. It often takes as little as 15 minutes for a ToO trigger to yield science photons. Another 5-10 minutes after the observation is complete, the anxious PI can retrieve the data from the Canadian Astronomy Data Centre (CADC) and evaluate it immediately, perhaps leading to another trigger using a different instrument. It was this system that made it possible to observe the weather on Titan using Altair and NIRI when photometry from smaller telescopes indicated that interesting meteorology was emerging on Saturn’s largest satellite. It was this system that captured the oldest photons ever recorded, sans those from the Cosmic Microwave Background, from a gamma-ray burst that is a staggering z ~ 8.2 away. Finally, it was this system that made it possible to record fantastic mid-infrared images of the still-hot impact site on Jupiter after an unseen comet plowed into Jupiter’s stormy atmosphere. We await with excitement the many targets of opportunity that will come within range of Gemini in the years ahead. In the longer term, we foresee a rich ToO discovery horizon at Gemini because there are few large facilities better positioned to study the fast-changing targets that will surely be revealed by the next generation of synoptic survey telescopes.