Introduction

Object density in Low Earth Orbit (LEO) is burgeoning, increasing the need for time-critical space domain awareness (SDA) information. Numerica’s cathemeral (meaning “through the 24-hour day”) optical systems are a new tool in the arsenal to meet this growing need. In 2021, Numerica conducted sensor access and data quality simulations to investigate how a global network of low-cost cathemeral optical telescopes (active 24/7) could augment radar systems to enhance tracking and custody of a subset of resident space objects (RSOs) in LEO. The information covered in this blog has been adapted from a research paper put together by the team at Numerica and published online in the AMOS Technical Paper Library.

Advancements in daytime optical tracking

Numerica used their existing cathemeral telescope network (CTN) – “Aquila” – to test the hypothesis that a system of cathemeral telescopes could complement existing radar networks. This daytime-and-nighttime-capable optical tracking system was originally designed in 2018 for geosynchronous orbit (GEO) observation at ranges exceeding 36,000 km. In 2020, Numerica enhanced its systems to track at the high angular rates required to support LEO observation. Five Aquila systems are currently fielded across the United States, Australia and Spain. A sixth Aquila system is being deployed to Africa in early 2022.

Aquila sensors use custom optics, high-speed shortwave infrared (SWIR) cameras and advanced algorithms to cut through noise from the daytime sky background. Aquila also uses a custom baffle to shade the optical mirrors from direct sunlight and to stop unwanted (stray) light from reaching the camera. At the time of writing, Aquila sensors have produced over 250,000 daytime observations on nearly 1000 RSOs, including over 650 RSOs in LEO.

Simulation of hypothetical sensor networks

To assess the benefit of an even larger network of cathemeral telescopes, the team at Numerica simulated a 16-site telescope network against real-world LEO object populations using existing Aquila data to formulate a realistic model of telescope performance. Numerica also simulated a four-site radar network, a common phenomenology in LEO SDA, for comparison individually and as a joint network.

Several technical constraints were modeled to determine whether an object in the sky was observable, including weather outages, horizon limits, detectability limits, solar exclusion and satellite pass duration. The team constrained each cathemeral telescope to view a single object during a collection period, though the constraints for the simulated radar network were more relaxed – each radar was able to view all objects in the sky above 20° elevation simultaneously (an acknowledged oversimplification).

CTN augmentation impact on key tracking metrics

Numerica assessed RSO time since last observation (TSLO) based on the sensor network configurations. TSLO is a different metric than the commonly used “gap time,” but it is a metric the team believes to be more appropriate for SDA timeliness requirements. While gap times measure subsequent observation delays, TSLO measures the time since last observation from a uniformly sampled set of times over an extended period.

The team computed distributions of TSLO from a 16-site CTN, a four-site radar network and a combination of these networks against 250 LEO RSOs. The resulting means/medians in minutes from these simulations were 112/66 (radar network), 68/49 (CTN) and 41/29 (CTN and radar network). The median case was close between the CTN and radar network (25% reduction in TSLO by CTN), although the radar network TSLO distribution had a distinguished tail skewing the means, while the combined network showed significant improvement over each independent network.

The team at Numerica further explored CTN benefits towards maintaining custody of a smaller set of objects to determine the maximum advantages of a 16-site CTN network when the risk of overlapping RSO passes at specific site is low. For less than 50 objects maintained, the radar network provided an average 110-minute mean TSLO, the CTN provided an average 50-minute mean TSLO and the combined network provided an average 33-minute mean TSLO. This test showed that a joint optical/radar network could provide an average half-hour TSLO on a select subset of RSOs.

Additionally, Numerica went on to verify that lower TSLO values can translate into increased RSO orbital state accuracy, even in a period of RSO inactivity. The team took observations against KOMPSAT-5, a South Korean ILRS satellite in dusk-dawn orbit and ran them through an unscented Kalman filter which fuses range/doppler and angle/angle measurements from both radar and optical systems.

This produced orbital state and covariance information over a week-long period of time. Like the TSLO metric previously stated, the team measured tracking position error at random time points during the simulation, but unlike TSLO, the results focused on the final three days of the seven-day simulation (signifying steady-state performance). As expected, the tracking position error rises when KOMPSAT-5 went unseen and was more pronounced in the radar network scenario.

Finally, the team analyzed the cumulative distribution function (CDF) of the tracking position errors across the steady-state period of the simulation and found that a CTN can provide reduced position uncertainty. The average error for the three cases was 34.4 m for the four-site radar network, 25.2 m for the 16-site CTN and 18.7 m for the combined network. In other words, the combined network showed an average decrease in tracking position error of 46% and 26% compared to the two individual sensor networks alone.

Conclusion

Everything discussed above leads to the simple fact that a cathemeral telescope network working in concert with a radar network (or independently from a radar network) could bring significant improvements to SDA timeliness in LEO.

Numerica’s simulations suggest that with this setup, one would see a significant reduction in both “time since last observation” and “tracking state uncertainty.” These simulations have also shown that radar systems and optical systems can work together to provide enhanced SDA at LEO, especially if the optical systems are cathemeral.

Object density in Low Earth Orbit (LEO) is burgeoning, increasing the need for time-critical space domain awareness (SDA) information. Numerica’s cathemeral (meaning “through the 24-hour day”) optical systems are a new tool in the arsenal to meet this growing need. In 2021, Numerica conducted sensor access and data quality simulations to investigate how a global network of low-cost cathemeral optical telescopes (active 24/7) could augment radar systems to enhance tracking and custody of a subset of resident space objects (RSOs) in LEO. The information covered in this blog has been adapted from a research paper put together by the team at Numerica and published online in the AMOS Technical Paper Library.