The Red-line Emission Geospace Observatory (REGO) All-Sky Imaging System

Aurora KEO SCIENTIFIC has teamed up with The University of Calgary's Auroral Imaging Group to make available to the community at large their RED-LINE EMISSION GEOSPACE OBSERVATORY (REGO) all-sky imaging system. Although originally targeted specifically at the 630 nm emission, the design is fundamentally optimized across a wide range of wavelengths and may therefore be applied to other wavelengths as well, from 427.8 nm and into the NIR (please contact Keo Scientific with your specific wavelength need(s)). Note that due to the unusually tight integration between custom optics and proprietary thin-film filter technology, this imager cannot be equipped with a filter-wheel; it is a highly optimized single-wavelength imager. For multispectral work, multiple REGO imagers may be co-deployed (or our Keo Sentry line of filterwheel-based imagers may be used wherever non-coincident exposures are acceptable).

The REGO all-sky airglow/auroral imager system is a fully independent, easily fielded, single-wavelength remote imaging System that was developed under a Canada Foundation for Innovation grant at the University of Calgary in 2014. The REGO system was specifically designed to operate alongside other Canadian auroral instrumentation at remote sites and provide key information about redline (630nm) aurora from the polar cap to subauroral locations. Now operated in part by a Canadian Space Agency contribution agreement, the array is currently (2018) in its fourth successful imaging season. The complete REGO electrooptical system design was transferred to Keo Scientific Ltd. for commercialization under a special Licence Agreement in August of 2017.

Please contact Keo Scientific for pricing and lead-time information.


Time resolution: 3 seconds (for a 2 second exposure), for auroral work
Spatial coverage: All sky view
Spatial resolution: 512x512x16 (capable of 2048x2048x16)
Sensitivity: 10 R/dn/s


Each REGO Observatory Package is fully self-contained — in that it requires only mains power, a network connection and a building with an installed observation dome — and weighs in at less than 100 lbs (45 kg). Each fully integrated System consists of the following hardware components:

  • All-sky (fisheye) optical column with internal custom filter/optics leveraging proprietary thin-film technology. 630 nm (2 nm FWHM) is the default wavelength but other filters are available upon request
  • Thermoelectrically cooled, large-format (27.6 x 27.6 mm) back-illuminated CCD sensor w/integrated shutter
  • Custom kinematic dome mount
  • Computer (1U rackmount, server grade)
  • GPS device, for time-keeping (accurate to 1 ms via NTP)
  • Network-accessible power distribution unit (PDU)
  • Uninterruptible power supply (UPS)
  • Temperature sensors for monitoring/logging ambient environment
  • Bright light sensor with remotely configurable threshold (logs ambient light levels, Arduino based)
  • Network router
  • External hard drives (backup + transfer drives)
  • Shock Case (8U). All equipment other than the temperature sensors, bright light sensor, and allsky imager are contained inside a rackmount-friendly and easily shipped “shock case”.


The Observatory operates autonomously once the software has been configured for the specific site (latitude, longitude) and observing schedule/mode. Imager settings include settings such as exposure time, binning, gain, vertical and horizontal clock speeds, cooler temperature, etc. The cadence of the REGO system is only limited by the readout speed of the CCD sensor, which is on the order of one second for a typical configuration. For a three second cadence, there is thus typically a 2s exposure, suitable for Auroral work at 630 nm. The image capture is initiated on a modulo of 3s within a minute with 200ms leniency for other CPU background processing causing blocking. In the same way, a schedule can be set up for Airglow work, which generally requires much longer exposure times. Such deterministic imaging ensures the synchronization of the system with other instruments, and also between other REGO systems.

REGO is configured to provide the following core set of data products:

  • Stream0: Raw frames, 20 frames stacked into a single file. The stream0 archive is the core dataset containing the raw images as recorded from the detector. Each file contains one minute of data with appropriate file and frame metadata.
  • Stream1: Thumbnails and keogram slices for real-time transmission. Thumbnails are created at a regular cadence from the full frame data. The thumbnail size, cadence and compression are controlled via the imaging software.
  • Stream2: Keograms, montages, averages. Keograms, montages, and hourly averages are created on site from the full frame data. These are used to help accurately evaluate the operational state of the imaging systems on a daily basis.


The REGO systems are fully calibrated prior to delivery. The final deliverables are files that contain the imager "flat field" and the number of "rayleighs per DN" for a standard 1s exposure. The procedure for correcting REGO image data and converting raw pixel values to units of rayleighs is found here.


The University of Calgary has a Data Portal for browsing the data from their array of REGO systems.

Note how the Open-Closed Boundary is readily visible in this REGO image mosaic, with auroral streamers seen in this time-series of REGO images.

A strong and long-lasting red-line pulsating auroral event that occurred during the St. Patrick’s Day geomagnetic storm is analyzed in this JGR paper, which has an accompanying movie provided in the Supporting Information section.


The REGO System was developed by the University of Calgary Auroral Imaging Group team under Canada Foundation for Innovation funding. Keo Scientific Ltd. wishes to especially thank Drs. Eric Donovan and Emma Spanswick for their kind cooperation during this Technology Transfer, a process facilitated by Innovate Calgary. Sample REGO redline data courtesy of Dr. Donovan et al.

Relevant paper in J. Geophys. Res. Space Physics, 121, 7988–8012, doi:10.1002/ 2016JA022901: