All-Sky Camera
Our All Sky Camera views the entire sky with a fisheye lens. It is designed to be deployed overnight. To cope with the extreme cold of Alaska we’ve built a special insulated housing with a built in heater. The camera is controlled with a Raspberry Pi to take time lapse images. The Pi is also equipped with two flux gate magnetometers, and a thermometer. If you’re looking for a quicker introduction to the system, check out our blog post dedicated to the All Sky camera: All Sky All The Time!. Below is a list of components:
- Canon EOS R Mirrorless full-frame camera
- Rokinon 8mm f/3.5 fisheye lens
- Raspberry Pi Model 3B+
- 2x Anker 20,000 mAh battery packs
- 2x Adafruit ADS1115 4-channel 16-bit analog to digital converters
- 3x TMP36 temperature sensors
- 2x Texas Instruments DRV425 flux gate magnetometers
- 3x red/blue LEDs
- 5cm x 10cm electric heating pad controlled by relay
The camera is controlled by a relay which triggers the shutter using it’s mini input jack. This allows the Raspberry Pi to trigger the camera shutter at a fixed interval. The camera is set to record RAW image files, allowing for maximum flexibility in postprocessing. The downside is this generates a large amount of data, with an 8 hour recording period where images are taken every ten seconds corresponding to about 100 gigabytes of raw image data. These images are then processed in Adobe Lightroom, down sampled to UHD 4k resolution, and assembled as a time lapse in Adobe Premiere.
Magnetometry and temperature data are recorded using two dedicated analog to digital converters (ADC) for each sensor type. The low-channel count high bit depth ADS1115 was selected because of the high resolution needed to resolve auroral perturbations to Earth’s magnetic field. The ADS1115s were interfaced with the Raspberry Pi via a single I2C bus.
The DRV425 flux gate magnetometer was selected because of its low cost and high sensitivity. The DRV425 has an analog output, meaning its measured magnetic field value is encoded in an output voltage which must be digitized in order for it to be recorded by the Pi. We purchased a pair of evaluation module boards which came with a default dynamic range of +/- 500 micro Tesla. Given that Earth’s magnetic field is of order 50 micro Tesla and auroral perturbations are on the order of 100s of nano Tesla or less, we needed to have the highest bit depth (and thus sensitivity to changes in voltage) possible. With the ADS1115 and the DRV425 evaluation module (part number DRV425EVM), we achieved a resolution of ~13 nano Tesla / least significant bit, sufficient for the basic magnetometry measurements we sought to make.
The TMP36 was selected because of its low cost and large operating range (works down to -40 Celsius!) which given Alaskan weather we felt we had a good chance of making full use of. We placed three temperature sensors in the system, one internal near the batteries and circuit boards, one up next to the camera, and one just outside the bottom plate of the enclosure. The internal temperature sensor near the batteries and electronics was used to decide when to acuate the electric heater. The external temperature sensor we found to be difficult to trust and seemed to consistently overestimate the external temperature, which we think is a result of heat conduction down the leads to the sensor. This would likely have been solved by placing the sensor farther outside of the enclosure.
The enclosure of the All Sky camera consists of an 8″ diameter cardboard concrete form tube 14″ in length, toped with two wooden end plates which consist of two 3/4″ thick pieces of wood glued together, one whose diameter is slightly larger than the outer diameter of the tube, and one whose is slightly smaller than the internal diameter of the tube, creating a small lip which allows the end cap to friction fit in the cardboard form tube without sliding all the way in. The top plate has an approximately 4″ hole drilled in it, over which 4″ and then 6″ diameter acrylic domes are mounted to create extra insulation for the lens.
We chiseled a 1/4″ slot in the base at a ~7″ chord to permit the fitting of a 1/4″ thick plywood sheet which would run the 14″ length of the cardboard tube. We then reinforced this plate with a pair of wooden ribs along the back of the plate, creating a firm platform on which to mount the camera and electronics. We chiseled a short 1/4″ slot near the top of the board and through one of the ribs which permitted a 1/4″-20 bolt to be fitted to secure the camera to the base plate. We mounted the Raspberry Pi and support electronics board using standard acrylic PCB standoffs. The batter compartment was crafted using leftover plastic housing from the battery packaging and some paracord.
The enclosure was given a coat of white paint, and a pair of 3/8″ foam “coozies” were made to add extra insulation around the cylindrical body of the device. With the enclosure complete, the system was ready to go.
Radios
The Aurora have several different radio emissions associated with them. We have developed several radio recievers to study these
Spectro-binoculars
The light we see when we look at the aurora comes from atmospheric molecules several hundreds of kilometers in altitude. High-energy particles from the solar wind trace Earth’s magnetic field lines toward the ground, smashing into our atmospheric gasses along the way. Each of these collisions can temporarily excite a molecule’s electron to a higher energy state. As the molecular electrons return to their ground state they release the photons we see during an auroral show.
Much information about our upper atmosphere’s gasses is coded in the aurora’s visible light spectrum. Each unique wavelength of light observed is a fingerprint for a different atmospheric gas. The green light seen in most types of aurora has a wavelength of 557.7nm, indicative of the presence of O2. The blue hue that accompanies the green structures occurs at 427.8nm, signaling N2. And the red shading of the aurora’s upper regions occurs at 630nm, further corroborating the presence of O2. Beyond the presence of individual spectral lines, line strength and the distribution of the spectrum near those lines reveals the kinetics of the atmospheric gases. The width of the lines tells us about density and temperature, and their mean position tells us the speed and direction of the air currents.
To capture the spectrum of the aurora we employ an Ocean Optics USB2000+ spectrometer. This instrument receives light via an optical fiber output port, aligns it onto a diffraction grating, and records light intensity at numerous diffraction angles. Each diffraction angle corresponds to a specific wavelength of light. In place of an optical fiber we collimate light into the instrument via one arm of a pair of binoculars. By mounting this pair of binoculars on a tripod, we can look at a target via one arm and record its visible spectrum via the other. We use a Raspberry Pi 3.0 to talk to the spectrometer, telling it what exposure time to use and how frequently to take measurements.
