Project PANOPTES


 

PANOPTES History and Science Mission


A Brief History of PANOPTES

PANOPTES got its start in late 2010 as a small hobby project of Olivier Guyon. He started experimenting with automated DSLR-based systems to take pretty pictures and rapidly began to wonder about the potential scientific capabilities of such a system. A DSLR provides a large pixel count at a comparatively low price when compared to “scientific grade” CCDs commonly used on small telescopes which makes it appealing for collecting data at low cost, but there were obstacles to overcome.

Automation

Olivier constructed a prototype automated camera system, but could only obtain limited data from home due to poor weather. Olivier, like many of the original PANOPTES team members lives in Hilo, Hawaii – the rainiest city in the United States. Some areas of Hilo average over 175 inches (4.56 meters) of rain per year.

At the time, another eventual PANOPTES team member, Josh Walawender, was working on a different small robotic telescope project which was located at the Mauna Loa Observatory (MLO) run by NOAA on the slopes of Mauna Loa. With his help, the original test unit of PANOPTES was hosted by MLO and infrastructure (power and networking) was provided through the Variable Young Stellar Objects Survey (VYSOS) project.

By this time others had joined the project and formed the initial PANOPTES team. Based on experience at Mauna Loa, the hardware of PANOPTES was tested and refined. We were able to demonstrate that low cost hardware could be made to operate reliably under adverse weather conditions even without a protective dome or roof to protect the equipment.

Image Processing

In addition to the operational testing, Olivier also developed an algorithm to obtain accurate photometry from the DSLR images. This was challenging because DSLRs have a Bayer color filter array in front of the detector which gives them the ability to obtain color images. This filter array places a small red, green, or blue filter over each pixel such that each 2x2 group of four pixels has one red sensitive pixel, two green sensitive pixels, and one blue sensitive pixel. The values for the other two colors are then interpolated for each pixel. For example, for a green pixel, the green value is measured and the red and blue values are interpolated from the neighboring red and blue pixels adjacent to the green pixel.

This color pixel array normally makes determining the brightness of a star difficult because of the star’s small size. The majority of the light for in focus stars falls on to a single pixel, so that star will appear different brightness depending on whether it happened to fall on a red, green, or blue pixel.

To get around this, Olivier’s algorithm generates a comparison stellar image based on selecting other stars in the image which fall on the same position in the RGGB color filter pattern and which are similar in shape (and other properties) to the star being measured. By watching how stars with the same “errors” due to the color filter array (and other effects) change in brightness due to those “errors” over many images, we can compare their behavior to the star being measured and see if it has brightened or faded during a particular image sequence.

By averaging the results of many observations from many cameras, we can obtain a very accurate measurement of any changes in a star’s brightness over time. This opens up an interesting scientific area for exploration with PANOTPES: exoplanet transits. With a large number of PANOPTES units scattered over the world, we can obtain high quality, continuous monitoring of moderately bright stars (roughly 9th to 12th magnitude) for exoplanet transits.

Science

The science goal of PANOPTES is to survey moderately bright stars (roughly 9th to 12th magnitude) to look for exoplanet transits. These stars are much brighter than the typical star studied by NASA’s Kepler mission and thus are likely to be closer to Earth which should make those planets easier to separate from their star.

Thanks to its highly distributed network of units, PANOPTES can provide nearly continuous observations using units around the globe, and observe large parts of the sky. This flexibility complements professional transit surveys that use a smaller number of telescopes, such as Kepler or TESS.

PANOPTES will find new exoplanets that can be studied in detail with larger telescopes using techniques such as radial velocity, transit spectroscopy, and direct imaging. In fact, some of the PANOPTES team also work on extreme adaptive optics systems such as the SCExAO project at Subaru Telescope which can be used to image nearby planetary systems, so the link between these two interests is that PANOPTES can find targets that extreme AO systems (like SCExAO) can then follow up on.