Knowing where a satellite is answers only part of the space domain awareness question. Programs also want to know what an object is doing: whether it is stable or tumbling, how fast it spins, and whether its behavior has changed since last week. For objects too far away to resolve, the main ground-based tool is the light curve, a series of brightness measurements over time.
This post is for SDA operators, research groups and sensor builders adding light curve photometry to their telescopes. It covers the camera choices that matter most, using Finger Lakes Instrumentation (FLI) Kepler cameras and our High Speed Filter Wheel as worked examples. For finding objects in the first place, see our post on GEO belt surveys.
As a satellite or rocket body rotates, its surfaces catch and lose sunlight, and the brightness seen from the ground rises and falls. Šilha and colleagues at the University of Bern used this to extract apparent rotation periods for more than 400 debris objects in LEO, GEO, GNSS and highly elliptical orbits (Advances in Space Research, 2018). Stabilized satellites leave signatures too. Hall's 2013 AMOS paper describes solar panel glints from stabilized GEO satellites, when a panel reflects sunlight toward the observer almost like a mirror. Linares and co-authors showed that light curves combined with angles data can help estimate an object's shape and attitude (Journal of Guidance, Control, and Dynamics, 2014).
All of these methods depend on frequent samples, good time tags and consistent photometry.
Two numbers set the cadence: how long each exposure lasts and how long the gap is before the next one. A tumbling rocket body can change brightness within seconds, and a glint can be over quickly. A long exposure averages that detail away, and a long gap between frames misses it. As a rule from sampling theory, you need at least two samples across the fastest feature you want to keep, and more is safer.
Readout time is often the limit. Schildknecht and colleagues, writing about light curves of GEO debris at AMOS in 2008, noted that the finite readout time of a classical CCD capped their sampling. They read small subframes around the target to sample faster, and kept the object centered in the subframe as it moved. Kepler cameras are designed for fast, continuous, short-exposure imaging, which suits long runs of short frames. If your software reads a subframe, ask our sales team how your model and software handle it.
A light curve is only as good as its time tags. Rotation periods come from folding many measurements onto one cycle, and glint times are compared with the predicted sun-object-observer geometry. Timing errors smear both.
Kepler sensors use a rolling shutter with a global-reset mode. In rolling mode, each row starts and ends its exposure slightly later than the row above it, so the time of a measurement depends on where the object sits on the sensor. Global-reset mode starts every row's exposure at the same moment, which keeps start times aligned across the frame. Either way, know which mode your pipeline assumes and record it with the data.
For the clock itself, a GPS timestamp option is listed for Kepler and Aurora cameras; contact FLI for current status on your model. Kepler cameras also accept hardware triggers through the SDK, so an external timing system can start exposures. If several cameras must observe together, tight synchronization needs an external sync or trigger cable driving all of them. Our post on synchronizing multiple cameras covers the details.
A glint can be far brighter than the steady reflection around it, and the faint parts of a tumble may sit close to the sky background. If the glint saturates, its peak and shape are lost. If the exposure is cut short to protect the peak, the faint parts sink into the noise.
Kepler cameras address this with a 16-bit high dynamic range image that combines two 12-bit readouts at different gains, with less than 1% non-linearity. Linearity matters as much as range here, because the light curve is a ratio of brightness over time. To check gain, noise and linearity on your own unit, see our post on camera acceptance testing. A production test summary is available for each camera by serial number.
Colors can hint at surface materials. Schildknecht's group took repeated sequences through V, R and I filters, with exposures of a few seconds and about 15 seconds between frames. Brightness often changed a lot within those 15 seconds, so they formed color indices from consecutive frames and averaged them to keep the tumble from posing as color.
The shorter the gap between filters, the less the object changes in between. Our High Speed Filter Wheel lists adjacent filter moves of 23 to 30 ms depending on model, 90 ms for the longest move, and a stored sequence of up to 50 moves that a trigger input steps through. That lets a camera and wheel cycle colors with little dead time. Our post on fast filter switching explains how to budget exposure, readout and filter-change time.
Most programs calibrate satellite brightness against stars in the same field, so the field must contain enough reference stars at each pointing. A wide field helps. The KL4040FI and KL4040BI have 4096 × 4096 pixels of 9 µm on a 52.1 mm diagonal, and the KL6060FI has 6144 × 6144 pixels of 10 µm on an 86.9 mm diagonal. Size the pixel scale to your telescope so the target stays well sampled. Our post on pixel size and sampling covers the math. Our post on time-series photometry covers stability and calibration in more depth.
Kepler cameras are FLI's line for space domain awareness, built for continuous 24-hour use, including both day and night imaging. FI and BI versions are offered for each format. FLI provides an ASCOM driver for Kepler cameras, and the Kepler SDK supports Windows and Linux for custom pipelines. User-replaceable shutter, fans, and digital board make Kepler ideal for remote installations. For running a fleet of sites, see our post on SDA sensor networks, and for daylight work, our post on daytime satellite tracking.
Send our sales team your telescope, target orbits, cadence and timing needs, and filter plan, and we will suggest the standard Kepler and filter wheel configuration that fits. If a program needs a modest change to a standard camera, we are glad to discuss it. Request a quote or email sales@flicamera.com.
FLI cameras are Made in USA, and our sensor chambers are assembled in our 1,000+ sq ft ISO Class 7 clean room. See how on our in-house manufacturing page.